Quality of TB/HIV Data Reported by Health Facilities in Uganda

Findings from Data Quality Assessment, August 2015

Report Prepared by

Monitoring and Evaluation of Emergency Plan Progress – Phase II (MEEPP II)

January 2016
Suggested Citation:
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from Data Quality Assessment,
August 2015. Unpublished manuscript.

The Monitoring and Evaluation of Emergency Plan Progress – Phase II (MEEPP II) project and this report
are made possible by the generous support of the American people through the United States Agency
for International Development (USAID) and President’s Emergency Plan for AIDS Relief (PEPFAR) under
the terms of USAID/Social & Scientific Systems (SSS) Contract No. AID-617-C-10-00008. The contents do
not necessarily reflect the views of PEPFAR or any United States government agency
TABLE OF CONTENTS
LIST OF FIGURES ........................................................................................................................................... iii
LIST OF TABLES .............................................................................................................................................. v
ACRONYMS .................................................................................................................................................. vi
ACKNOWLEDGEMENTS ............................................................................................................................... vii
EXECUTIVE SUMMARY .................................................................................................................................. 1
1. INTRODUCTION ..................................................................................................................................... 5
1.1 Background ................................................................................................................................... 5
1.2 Collaborative TB/HIV Services ...................................................................................................... 5
1.3 Challenges in Data Quality ............................................................................................................ 6
1.4 Rationale for the Data Quality Assessment .................................................................................. 6
1.5 Objectives ..................................................................................................................................... 7
2. DATA AND METHODS............................................................................................................................ 8
2.1 Design ........................................................................................................................................... 8
2.2 Study Sites .................................................................................................................................... 8
2.3 Sample .......................................................................................................................................... 8
2.4 TB Indicators Assessed.................................................................................................................. 9
2.5 Data Sources ............................................................................................................................... 11
2.6 Data Collection Tool ................................................................................................................... 11
2.7 Training of Data Collection Teams .............................................................................................. 14
2.8 Pilot Testing of Tools and Field Procedures................................................................................ 14
2.9 Field Data Collection Procedures................................................................................................ 15
2.10 Data Management ...................................................................................................................... 15
2.10.1 Data Entry ........................................................................................................................ 15
2.10.2 Data Cleaning ................................................................................................................... 16
2.10.3 Data Analysis.................................................................................................................... 16
3. RESULTS............................................................................................................................................... 19
3.1 Descriptive Characteristics of DQA Participating Facilities......................................................... 19
3.2 Objective 1a: Assessment of District M&E Systems ................................................................... 21
3.2.1 Functions and Capabilities of Districts to Handle HIV/TB Information ............................. 22
3.2.2 Understanding of Indicator Definitions and Reporting Guidelines ................................... 25
3.2.3 Availability of Data Collection Tools and Reporting Forms for TB/HIV Data ..................... 26
3.2.4 TB/HIV Data Management Processes in Districts .............................................................. 28
3.2.5 Use of Data for Decision Making at District Level ............................................................. 30

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 i
 3.3 Objective 1b: Assessment of Facility M&E Systems ................................................................... 33
3.3.1 Status of Facility M&E Systems for TB/HIV Services.......................................................... 33
3.3.2 Function and Capabilities of Facility M&E Structure for HIV/TB Information ................... 34
3.3.3 Understanding of Indicator Definitions and Reporting Guidelines ................................... 36
3.3.4 Availability of Data Collection and Reporting Tools .......................................................... 40
3.3.5 Data Management Processes in Health Facilities .............................................................. 42
3.3.6 Use of Data for Decision Making in Health Facilities ......................................................... 45
3.4 Objective 2: Verification of Sources of Reported TB/HIV Data .................................................. 48
3.4.1 Availability of Data Sources ............................................................................................... 48
3.4.2 Completeness of Primary Data Sources ............................................................................ 48
3.5 Objective 3: Validation of Reported TB/HIV Data ...................................................................... 51
3.5.1 Characteristics of the DHIS 2 Data ..................................................................................... 51
3.5.2 Overall Comparison of Totals ............................................................................................ 51
3.5.3 Facility Level Analysis......................................................................................................... 56
3.5.4 Site Level Comparison of Validated Counts versus Figures in the Facility Copy of the HMIS
106a Report ................................................................................................................................ 59
3.6 Objective 4: Challenges Faced in Collecting, Recording and Reporting Data at Sites ................ 60
4. DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS .................................................................. 63
5. REFERENCES ........................................................................................................................................ 66
6. ANNEXES ............................................................................................................................................. 67
Annex 1: Districts and Facilities in the TB/HIV DQA, by IP.................................................................. 67
Annex 2: District TB/HIV Data Verification and System Assessment Tool.......................................... 73
Annex 3a: Health Facility TB/HIV Data Verification and Validation Tool............................................ 79
Annex 3b: Health Facility TB/HIV System Assessment Tool ............................................................... 83
Annex 3c: General Observations and Notable Good M&E Practices at Site ...................................... 89
Annex 3d: Site Feedback Form ........................................................................................................... 91
Annex 3e: Site Recommendations and Action Plans .......................................................................... 95
Annex 4: DQAI Team Members and Central Coordinating Team ....................................................... 97
Annex 5: District Rating by M&E Component Assessed for TB/HIV System..................................... 101
Annex 6: Deviation in DHIS 2 vs. Manual Recounts for Seven Variables .......................................... 105

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 ii
LIST OF FIGURES
Figure 2.1. Districts Where DQA Conducted, August 2015.......................................................................... 8
Figure 3.1. Overall Rating of District Level M&E Systems for TB/HIV Data, August 2015 .......................... 21
Figure 3.2. Summary District-level M&E Systems Assessment for TB/HIV Data by M&E Component,
August 2015 ................................................................................................................................................ 22
Figure 3.3. Overall Rating of District M&E Structures, Functions, and Capabilities for Management of
TB/HIV Data, August 2015 .......................................................................................................................... 22
Figure 3.4. M&E Structures, Functions, and Capabilities of Districts for Management of TB/HIV Data,
August 2015 ................................................................................................................................................ 23
Figure 3.5. Overall Rating of Understanding of Reporting Guidelines and Indicator Definitions, August
2015 ............................................................................................................................................................ 25
Figure 3.6. District Knowledge of Reporting Guidelines and Indicator Definitions, August 2015 ............. 25
Figure 3.7. Overall Rating of Availability of TB/HIV Data Collection and Reporting Tools, August 2015 ... 26
Figure 3.8. Availability of TB/HIV Data Collection and Reporting Tools Among Districts, August 2 ........... 27
Figure 3.9. Overall Rating of Data Management Processes for TB/HIV Data, August 2015 ....................... 28
Figure 3.10. Data Management Processes for TB/HIV Data Among Districts, August 2015 ...................... 29
Figure 3.11. Overall Rating of Data Use for Decision Making Among Districts, August 2015 .................... 30
Figure 3.12. Data Use for Decision Making Among Districts, August 2015 ................................................ 31
Figure 3.13. Overall Performance of All 106 Facilities for Five Components of M&E Systems Assessed,
August 2015 ................................................................................................................................................ 33
Figure 3.14. Facility M&E Systems Assessment by Component, August 2015 ........................................... 34
Figure 3.15. Overall Rating of Facility M&E Structures, Functions, and Capabilities for Management of
TB/HIV Data, August 2015 .......................................................................................................................... 34
Figure 3.16. Facility M&E Structures, Functions, and Capabilities for Management of TB/HIV Data, August
2015 ............................................................................................................................................................ 36
Figure 3.17. Proportion of Facilities Meeting All Requirements of Capabilities and Functions for TB/HIV
Data Management, August 2015 ................................................................................................................ 37
Figure 3.18. Overall Rating of Understanding of Reporting Guidelines and Indicator Definitions among
Facilities Assessed, August 2015 ................................................................................................................. 37
Figure 3.19. Understanding of Reporting Guidelines and Indicator Definitions among Facilities Assessed,
August 2015 ................................................................................................................................................ 38
Figure 3.20. Proportion of Facilities Meeting All Requirements for Understanding of Reporting Guidelines
and Indicator Definitions, August 2015 ...................................................................................................... 39
Figure 3.21. Overall Rating of Availability of TB/HIV Data Collection and Reporting Tools among Facilities,
August 2015 ................................................................................................................................................ 40
Figure 3.22. Availability of TB/HIV Data Collection and Reporting Tools At Facilities, August 2015 .......... 40

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 iii
Figure 3.23. Proportion of Facilities Meeting all Requirements for Availability of Data Collection and
Reporting Tools, August 2015 ..................................................................................................................... 42
Figure 3.24. Overall Rating of Data Management Processes Among Facilities, August 2015 .................... 42
Figure 3.25. TB/HIV Data Management Processes Among Facilities, August 2015 .................................... 43
Figure 3.26. Proportion of Facilities Meeting All Requirements for Data Management Processes, August
2015 ............................................................................................................................................................ 44
Figure 3.27. Overall Rating of Data Use for Decision Making in Facilities, August 2015 ............................ 45
Figure 3.28. Use of Data for Decision Making in Facilities, August 2015 .................................................... 46
Figure 3.29. Proportion of Facilities, by Level, Meeting All Requirements for Use of TB/HIV Services Data
for Decision Making, August 2015 .............................................................................................................. 47
Figure 3.30. Summary of Deviation of Totals from DHIS 2 and Facility Copy of HMIS 106a Report from
Manual Recount, August 2015.................................................................................................................... 53
Figure 3.31. Comparison of Manual Recount Totals to DHIS 2 and Facility Copy of HMIS 106a Report for
Variables for TB/HIV Indicator 1, August 2015 ........................................................................................... 54
Figure 3.32. Comparison of Manual Recount Totals to DHIS 2 and Facility Copy of HMIS 106a Report for
Four TB Variables, August 2015 .................................................................................................................. 55
Figure 3.33. Site Level Comparison of Validated Counts and Totals Reported in DHIS 2 for Seven
Variables, August 2015 ............................................................................................................................... 57
Figure 3.34. Site Level Comparison of Validated Counts and Totals Reported in Facility Copy of HMIS
106a Report for Seven Variables, August 2015 ......................................................................................... 59

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 iv
LIST OF TABLES
Table 2.1. Comparison of Indicators Assessed in DQA: All Reporting Facilities vs. DQA Sample, August
2015 .............................................................................................................................................................. 9
Table 2.2. TB/HIV Indicators Reported at SAPR 2015 ................................................................................. 10
Table 2.3. Data Sources for TB/HIV Data Assessed in DQA by Variable, August 2015 .............................. 11
Table 3.1. Description of Facilities Participating in DQA, August 2015 ...................................................... 19
Table 3.2. District M&E Functions and Capacities for Handling TB/HIV Data, August 2015 ...................... 24
Table 3.3. Availability of Data Collection and Reporting Tools at District, August 2015 ........................... 27
Table 3.4. Modalities of Followup on Late Submission or Inconsistent Reports, August 2015 .................. 30
Table 3.5. Use of TB/HIV Data for Decision Making by Districts, August 2015 .......................................... 31
Table 3.6. Staff Compiling and Reviewing Registers and Reports before Submission, August 2015......... 35
Table 3.7. Modalities of Communication of Changes to Previously Submitted Reports, August 2015 ...... 38
Table 3.8. Reported Availability of TB/HIV Tools by Type, August 2015 .................................................... 41
Table 3.9. Detailed Findings on Availability of Data Collection and Reporting Tools, August 2015 ........... 41
Table 3.10. Number of Facilities, by Level, with “Yes Completely” Responses to Requirements Assessed
under Data Management Processes, August 2015 ..................................................................................... 44
Table 3.11. Staff that Led Data Analysis and Interpretation, August 2015................................................ 46
Table 3.12. Completeness of Data Sources for the DQA in Facilities, August 2015 .................................. 48
Table 3.13. Reasons for Incomplete TB/HIV Data Collection and Reporting Forms, August 2015 ............. 49
Table 3.14. Availability and Completeness of Data Sources across Levels of Facilities, August 2015 ........ 50
Table 3.15. Percentage Deviation of Manual Recount Totals against DHIS 2 Facility Copy of HMIS 106a
Report Totals, August 2015......................................................................................................................... 52
Table 3.16. Percentage Deviation of Manual Recount Totals from District TB Register, August 2015 .... 52
Table 3.17. Reasons for Disparity between Manual Recount and DHIS 2 and Facility Copy of HMIS 106a
Report for Number of Clients in Pre-ART Care, August 2015 ..................................................................... 54
Table 3.18. Reasons for Disparity between Manual Recount and DHIS 2 for Number of Clients in ART
Care, August, 2015 ...................................................................................................................................... 54
Table 3.19. Reasons for Disparity between Manual Recount and DHIS 2 for Number of HIV-positive TB
Cases that Started or Continued ART During the Period for the SAPR 2015, August 2015 ........................ 55
Table 3.20. Deviation of Manual Recounts from Reported Totals by Level of Facility, August 2015 ......... 56
Table 3.21. Site Validated Data vs. DHIS 2, HMIS 106a Report at Site, and District TB Register, Overall
Level of Adequacy, All Indicators, August 2015 .......................................................................................... 58
Table 3.22. Facilities with DHIS 2 Totals within Range of Validated Count, by Number of Variables,
August 2015 ................................................................................................................................................ 59

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 v
ACRONYMS
ACP AIDS Control Programme
APR Annual Program Results
ART Antiretroviral Therapy
CQI Continuous Quality Improvement
DHO District Health Officer
DHT District Health Team
DHIS 2 District Health Information System version 2
DoD U.S. Department of Defense
DQA Data Quality Assessment
DTLS District TB and Leprosy Supervisor
HC Health Centre
HIV Human Immunodeficiency Virus
HMIS Health Management Information System
IDI Infectious Disease Institute
IM Implementing Mechanisms
INH Isoniazid
IP Implementing Partner
LTFU Loss To Follow Up
MDR TB Multi Drug Resistant Tuberculosis
MEEPP Monitoring and Evaluation of Emergency Plan Progress Project
MJAP Makerere University Joint AIDS Programme
MoH Ministry of Health
MUSPH Makerere University School of Public Health
NTLP National TB and Leprosy Programme
OGAC Office of the Global AIDS Coordinator
PEPFAR President’s Emergency Plan for AIDS Relief
PLHIV People Living with HIV
RC Resource Centre
RDQA Routine Data Quality Assessment
RPMT Regional Performance Monitoring Teams
RRH Regional Referral Hospitals
RTBLFP Regional TB Leprosy Focal Person
SAPR Semi-annual Program Results
SOP Standard Operating Procedure
SUSTAIN Strengthening Uganda’s Systems for Treating AIDS Nationally
TB Tuberculosis
UPMB Uganda Protestant Medical Bureau
US United States

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 vi
ACKNOWLEDGEMENTS
The central coordinating team and all investigators of the TB/HIV data quality assessment (DQA)
acknowledge with thanks the contribution, input, and efforts of all stakeholders and individuals that
singularly and collectively facilitated the successful conduct of the activity. Staff of the Monitoring and
Evaluation of Emergency Plan Progress Project (MEEPP) and Ministry of Health (MoH) contributed to the
concept design, preparatory activities, training and supervision of field data collection staff, data
analysis, report writing, review of various documents including the protocol, data collection tools, and
draft versions of the report. Implementing Partners, District Health Teams, Regional Performance
Monitoring Teams, MoH, MEEPP and staff of various United States Government agencies took part in
data collection including the grueling manual recounts of facility level TB/HIV data, as well as preliminary
analysis and debriefing of facility teams. The quality of the data in this report is a reflection of the
dedicated efforts of field teams. We are most appreciative of all their efforts.

We are most grateful to management of the various districts, health facilities, and their staff for their
time, cooperation, and willingness to make records available for the assessment. We also appreciate the
countless hours they spend each and every day providing critical TB/HIV/AIDS and other health services
to the deserving population.

Special thanks to Dr. George Upenytho from National TB and Leprosy Programme and Dr. Norah
Namuwenge, the DQA team lead, who worked tirelessly to support all processes of the DQA until
compilation of the final report. Finally, a big thank you to the MEEPP staff who worked late into the
night during various phases of the activity to ensure the project ran smoothly.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 vii
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 viii
EXECUTIVE SUMMARY
Context: For the past three decades, the dual epidemic of tuberculosis (TB) and human
immunodeficiency virus (HIV) has presented unprecedented challenges to Uganda’s health care system
with approximately 1.5 million people living with HIV at the end of 2014, and approximately 100,000
new HIV infections and 32,000 acquired immune deficiency syndrome (AIDS)-related deaths during the
year. In Uganda, approximately 60,000 people developed TB and about 4,100 died of the disease in 2013
(WHO report, 2014). During the same period, the health system detected and notified 47,650 cases. Of
these, 2,760 people died of TB according to National Tuberculosis and Leprosy Programme (NTLP)
reports. Tuberculosis is the leading opportunistic infection among people living with HIV (PLHIV) and a
leading cause of AIDS-related deaths. About 50% of people with TB are co-infected with HIV, and HIV
prevalence is seven times higher among TB patients than in the general population (49% versus 7.3%).
The Ministry of Health (MoH) and its partners have rolled out strategies for integration of TB/HIV
prevention and treatment, comprising screening and treatment of HIV and TB in both TB and HIV clinical
settings and isoniazid (INH) prophylaxis among PLHIV.

Monitoring and evaluation of TB/HIV collaboration is integrated into TB/HIV monitoring systems with
indicators for tracking outputs of collaborative activities included in the Health Management Information
System (HMIS). The indicators are tracked using standard HMIS recording and reporting tools and
systems. However, to assure that quality data is reported through the reporting system, a periodic data
quality assessment (DQA) of routinely reported data is conducted. DQA is also a key requirement for
data submitted by the President’s Emergency Plan for AIDS Relief (PEPFAR) to the Office of the Global
AIDS Coordinator (OGAC). For the Semi-annual Program Results (SAPR) 2015, TB/HIV indicators were
selected for DQA in part because TB/HIV data quality was last assessed in 2011.

Objectives: The main objective of the DQA was to improve data management systems for TB/HIV
programs in Uganda. The specific objectives were to: 1) assess M&E systems for TB/HIV programs at
service delivery and district levels with PEPFAR-supported implementing mechanisms (IMs); 2) verify
sources for TB/HIV data reported at SAPR 2015; 3) validate TB/HIV data reported at SAPR 2015; and 4)
identify challenges faced by facilities in collecting, recording, collating, and reporting TB/HIV data.

Data and Methods: A cross-sectional assessment was conducted in a representative sample of 106
health facilities selected from 1,007 facilities countrywide that reported TB/HIV data for SAPR 2015.
Mixed sampling methods were used, including stratified random sampling based on Regional
Performance Monitoring (RPM) health regions, and purposive sampling to maximize representation.
Final adjustments were made for logistical and security reasons.

The following three TB/HIV indicators reported at SAPR 2015 were assessed: 1) Percentage of PLHIV in
HIV clinical care settings who were screened for TB symptoms at the last clinical visit; 2) Percentage of
registered new and relapsed TB cases with documented HIV status; and 3) Percentage of HIV-positive
new and relapsed registered TB cases who started or continued antiretroviral therapy (ART) during TB
treatment.

The assessment used semi-structured MS Excel-based questionnaires adopted from the Measure
Evaluation routine DQA (RDQA) tool. Eleven trained multi-agency field teams comprising representatives
from MEEPP, MoH (NTLP, Resource Centre (RC), AIDS Control Programme (ACP)), Track TB, Regional
Performance Monitoring Teams (RPMT), implementing partners (IPs), and District Health Team (DHT), as
well as facility staff, were formed and deployed for data collection. Data were collected from district
offices and health facility staff where teams assessed systems capacity and functionality, and reviewed

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 1
data sources at sites to verify availability and completeness. Data reported in SAPR 2015 were validated
through manual recount of the outputs for the seven variables conducted jointly by field teams and
facility staff for the three TB/HIV indicators assessed.

Preliminary analyses at the site level, and analysis of pooled data at the central level, were conducted.
Automated preliminary analysis at sites utilized dashboards which displayed frequency distributions of
summary statistics, with color codes highlighting performance levels to aid feedback to facilities by field
teams. At the central level, pooled data were analyzed and qualitative data was grouped under themes,
coded, and frequencies generated. Excel and SPSS packages were used. Scores for the various questions
from the M&E systems assessment were averaged by component, with the resulting score per
component ranging from 0 to 3. Scores were categorized as follows:

 less than 2.25 denoted inadequate systems

 2.25–2.74 denoted acceptable systems
 2.75–3.00 denoted adequate systems
Reported data were validated by joint manual recount from the primary data collection tools and
registers, and compared with data downloaded from the District Health Information System version 2
(DHIS 2) database a day before field work commenced. The deviation of facility data in DHIS 2 from the
validated joint manual recount was assessed for all seven variables, for each facility and for pooled
totals for all facilities. The degree of variation for each variable was calculated as follows:

DHIS 2 Figure – Joint Recount Figure

Percentage difference for each variable = x 100%
Joint Recount Figure

The degree of variation reported was characterized as follows:

 0 to ±5% deviation of verified counts was deemed to be adequate

 ±5% to ±10% deviation was deemed within acceptable error limits
 greater than ±10% deviation was deemed inadequate as it represents significant over- or under-
reporting
Results: The quality of the TB/HIV data reported was assessed at 106 facilities comprising 20 hospitals,
28 Health Centre (HC) IVs, and 58 HC IIIs in 61 districts. Depending on the variable, 23–44% of facilities
grossly over-reported figures in the DHIS 2 database by 10% compared to validated figures at the facility,
while 18–38% under-reported these variables by 10%. Only 21–47% of the facilities reported data in
DHIS 2 that were adequate (0 to ±5% deviation) when compared to verified figures. Specifically:

 The number of PLHIV in clinical care settings that were screened for TB was grossly over-reported by
31% of facilities and under-reported by 22%. Reporting by 30% of facilities was adequate.
 About 32% and 38% of facilities grossly over-reported and under-reported the number of adults and
children in pre-ART care, respectively. Reporting by 23% of facilities was adequate.
 The number of clients actively enrolled on ART was grossly over-reported by 34% of facilities and
grossly under-reported by 18%. Reporting by 36% of facilities was adequate.
 Similarly, the number of new and relapsed TB cases with documented HIV status was adequately
reported by 47% of facilities, grossly over-reported by 35%, and under-reported by 29%.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 2
 Nearly half of facilities (44%) grossly over-reported the number of registered or relapsed TB cases,
with one-fifth (21%) reporting adequate figures. About 24% of facilities grossly under-reported it.
 The number of HIV-positive clients that started or continued ART (and its denominator, i.e., the
number of registered TB cases with documented HIV-positive status) were adequately reported by
36% and 29% of facilities, respectively. However, nearly all remaining facilities either grossly over-
reported or under-reported the figures for the two variables.
 For all variables, the proportion of facilities that reported figures with moderate or acceptable
deviation (±5% to ±10%) was small.

No facility had an exact match between the total joint validated recount and the totals reported in the
DHIS 2 database for all the seven variables reviewed. In some instances, however, the over-reporting
and under-reporting by facilities tended to balance each other out. Specifically:

 For two variables (the number PLHIV screened for TB in clinical settings, and the number of new and
relapsed TB cases with documented HIV serostatus) the totals were within ±5% deviation (78,184 in
DHIS 2 versus 78,883 from manual recount and 2,462 versus 2,578, respectively).
 For a third variable, i.e. the number of registered TB cases with documented HIV-positive status, the
deviation of the totals was within the acceptable range of ±10%.
 For the remaining variables, the deviations of the total exceeded the ±10% limit, with the deviation
for one variable (the number of HIV-positive adults and children in pre-ART care) deviating by 23%.

Overall, facilities performed better on reporting variables recorded and reported through the TB
information system than those reported through the HIV/AIDS chronic care information system. Similar
trends were found when comparing manually validated data with that recorded in the facility copy of
the HMIS 106a report and with the totals in the pooled data. The inaccuracies in the reported facility
data reflected the fairly weak information management systems for TB/HIV services data at the district
and facility level, as noted in the facility and district M&E systems assessment.

Overall, 42% of facilities and 39% of districts were found to have adequate M&E systems (average scores
of 2.75–3.0). Six districts (Adjumani, Butaleja, Buyende, Dokolo, Hoima, and Kanungu) either had M&E
systems that were only partly proficient or not proficient at all. Facility and district systems for the
following four components of M&E systems were just about average: 1) availability of data collection
and reporting tools (66% of districts and 62% of facilities scored adequately against the requirements
with scores of 2.75–3.0); 2) understanding of reporting guidelines and monitoring indicators (53% of
districts and 60% of facilities scored adequately against the requirements); 3) capacity to handle TB/HIV
services data (46% of districts and 58% of facilities scored adequately); and 4) data use for decision
making (approximately 51% of districts and 34% of facilities scored adequately against the
requirements). However, facility and district systems were particularly weak on the data management
processes component (where only 8% and 34%, respectively, scored adequately against the
requirements).

All sampled facilities, with the exception of two, had the primary data collection and reporting tools. The
completeness of the tools was slightly sub-optimal, especially for HIV/AIDS tools where in over one-third
of facilities, both the pre-ART and ART registers were not up to date. However, TB/HIV data covering the
period October 2014–March 2015 were more available (i.e., 91% and 95% of facilities had complete data
for chronic HIV care and for TB, respectively), though in some instances, field teams resorted to other
non-standard sources of data where standard registers were incomplete.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 3
Conclusions and Recommendations: The quality of TB/HIV services data reported by facilities to the
central level was largely inadequate with gross over-reporting and under-reporting by the majority of
facilities. The proportion of facilities that reported data that was within the acceptable range of the
verified figures was very small. However, because of the balancing out effect of over-reporting and
under-reporting, in some instances, the totals reported by the national level may tend to mask the gross
deviations at the facility level. The systems for collecting and reporting TB/HIV services data in the
health system revealed a mixed pattern with information systems in most facilities and districts in
urgent need of remediation in one or more areas, particularly data management processes. The
deficiencies in reported TB/HIV data are likely to constrain accurate measurement and monitoring of
these key program indicators.

These findings make a compelling case for a broad range of innovative approaches tailored to specific
needs of facilities to improve data management processes in order to accurately inform program
planning and evaluation. These may include expanded use of electronic medical records systems, new
ways to promote data use within facilities and districts, and creative ways to address the severe human
resource crisis. Increased data use can, in turn, lead to the demand for better quality data. Other areas
to be strengthened include feedback mechanisms for quality of data reported, support supervision of
facilities and districts, and the elaboration of guidelines and standard procedures for a broad range of
data management practices. Facilities identified during this assessment as having specific constraints
and challenges with their TB/HIV services need tailored remediation in the short-term to implement the
corrective actions.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 4
1. INTRODUCTION

1.1 Background
For nearly three decades, Uganda has braved a severe and generalized dual epidemic of HIV/AIDS and
tuberculosis (TB) across the country. The Uganda 2011 AIDS Indicator Survey (AIS) revealed an increase
in HIV prevalence among adults ages 15–49 from 6.4% in 2004/05 to 7.3% in 2011. This increase
occurred in several regions of the country, and even doubled in some regions such as West Nile. It is
estimated that by the end of 2014, there were approximately 1.5 million people living with HIV (PLHIV)
in the country. Furthermore, the country continues to experience a high rate of new HIV infections,
estimated at approximately 100,000 during 2014. The number of new HIV infections was consistently
higher than that of the number of individuals enrolled in the lifelong antiretroviral treatment (ART) until
2013 when new ART enrolment exceeded new HIV infections for the first time. HIV/AIDS mortality in
Uganda may have declined in recent years, but AIDS continues to claim the lives of many people, with
approximately 32,000 individuals estimated to have died from it in 2014.

The scourge of the HIV epidemic in the country has been aggravated by the TB epidemic. As part of this
dual epidemic, TB has been the main underlying causes of deaths in AIDS; in 2012, approximately 37% of
autopsied deaths among PLHIV was attributed to TB.1 Co-infection by TB and HIV is high. Approximately
50% of TB patients are co-infected with HIV which is seven times higher than in the general population
where it is 7.3%. For several decades, TB has been the leading opportunistic killer of AIDS patients.
Though TB/HIV co-infection has declined from 54% in 2011 to 49% in 2013, an estimated 1.4 to 7% of
adults, and up to 9.5% of children living with HIV had prevalent TB.2 Integrated adjunctive interventions
such as isoniazid (INH) preventive therapy, early detection of TB in HIV-infected individuals, and
expanding ART treatment to all TB/HIV co-infected individuals has implications for the incidence,
prevalence, and mortality of TB in the country.

World Health Organization (WHO) estimates indicate that TB prevalence and incidence rates in Uganda
have declined from 50,492 and 624 per 100,000 population in 1990 to 13,175 and 179 per 100,000
population in 2012,3 respectively. Accurate estimates of TB prevalence or mortality are not available due
to weaknesses in the surveillance system. (However, accurate data on TB prevalence are expected from
the TB Prevalence Survey that was conducted in 2014.) Data on TB and HIV treatment outcomes among
co-infected individuals are not available, either. Treatment outcomes for TB in Uganda were previously
not disaggregated by HIV status and, therefore, the proportion of TB/HIV co-infected individuals that
died is unknown. This disaggregation has been recently introduced for the cohort beginning January
2014, and data will be available when treatment outcomes for this and subsequent cohorts are reported.

1.2 Collaborative TB/HIV Services

In view of the burden of the dual TB and HIV epidemics in Uganda, the Ministry of Health adopted WHO-
recommended strategies for integration of TB/HIV prevention and treatment. National guidelines and
strategic plans for TB/HIV integration were developed and rolled out. Under this approach, active case
finding of TB and HIV among PLHIV in AIDS and TB care and treatment facilities was introduced, along

1
Cox JA et al, 2012
2
NTLP Annual Report, 2012/2013
3
Global TB Report, 2013

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 5
with capacity building for service providers. TB/HIV co-infected patients commence treatment in the
same facility where the diagnosis is made since TB and HIV service providers have been trained in
treatment of both conditions. Under recent national ART guidelines, all TB/HIV co-infected individuals
are automatically eligible for ART. Prevention of TB and HIV are also integrated. Furthermore, in order to
reduce incidence of TB among PLHIV, the Ministry of Health (MoH) recently introduced and is rolling out
isoniazid prophylaxis among PLHIV when active TB has been excluded. These guidelines are being rolled
out in a phased approach.

To monitor TB/HIV integration efforts, monitoring strategies, guidelines, and indicators have been
introduced under the TB and HIV programs in the country. Indicators of TB/HIV integration are now
captured under the national M&E strategy and some are incorporated in routine facility patient records
based on standard tools and monitoring indicators. Output indicators of treatment programs are
reported using standard Health Management Information System (HMIS) tools. The indicators from the
health system are currently reported through the MoH District Health Information System version 2
(DHIS 2) online reporting system.

The National TB and Leprosy Programme (NTLP) also runs a parallel reporting system for TB indicators
which also incorporates some of the TB/HIV indicators. Under this information system, TB registers are
maintained and regularly updated at district and regional level by the District TB and Leprosy Supervisor
(DTLS) and Regional TB and Leprosy Focal Person (RTBLFP), respectively.

1.3 Challenges in Data Quality

Although there has been recent expansion in coverage and uptake of TB and HIV/AIDS care and
treatment services, challenges remain in maintaining and improving the quality of the services that
patients receive. In addition, issues with data quality constrain the ability to measure and monitor key
program indicators at facility, district, and national levels. These challenges include the paperwork
burden; limited use and availability of electronic patient monitoring systems; weaknesses in data
recording, completeness, and accuracy; weak patient monitoring; frequent stock outs of standard data
collection and reporting tools; inadequate human resources (HR) in numbers and skills training, etc. In
the past, health facilities were burdened by requirements to use a patchwork of different recording and
reporting forms and electronic information systems by different IPs. This has been addressed in recent
years through harmonization led by MoH working closely with the President’s Emergency Plan for AIDS
Relief (PEPFAR) program. However, the roll out of standard HMIS forms is often not harmonized, and
sometimes tools are rolled out without adequate training.

1.4 Rationale for the Data Quality Assessment

The PEPFAR implementing mechanisms (IMs) supporting HIV care services are expected to meet
required standards for documenting and reporting program outputs and/or outcomes semi-annually.
Reporting provides a rapid assessment of the overall performance of IMs and allows them to self-audit
their own records and processes, helping program managers evaluate their strengths and weaknesses.

PEPFAR-funded countries are all required to regularly conduct data quality assessments (DQAs) as one
of the strategies for ensuring high quality data for informing program planning, monitoring, and
performance management. The DQAs help identify data quality issues for which corrective measures
can be quickly instituted. PEPFAR mandatory DQAs are to be done at least once every 3 years for each of
the selected PEPFAR indicators. To meet this requirement, Monitoring and Evaluation of Emergency

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 6
Plan Progress (MEEPP) project, as the monitoring and evaluation (M&E) IM, performs DQAs following
PEPFAR semi-annual and annual reporting periods.

In May 2015, PEPFAR Uganda submitted the FY 2015 Semi-annual Program Results (SAPR 2015) to the
Office of the Global AIDS Coordinator (OGAC), and a DQA was performed following the reporting period.
The SAPR 2015 DQA focused on TB/HIV indicators because: 1) the last TB/HIV DQA was conducted in
2011; 2) during the 2015 SAPR reporting, several data quality issues relating to TB/HIV indicators were
identified, and, as these indicators form approximately 13% of the indicators required for OGAC
reporting at APR 2015, if not addressed, could impact negatively on reported results; and 3) TB/HIV co-
infection is a common cause of morbidity that needs to be accurately reported to inform programming.
In light of this, it was hoped that the DQA would assess and help improve the systems in place for
monitoring national TB/HIV programs.

1.5 Objectives
The main objective of the DQA exercise was to improve data management systems for TB/HIV programs
in Uganda. The specific objectives were to:

 Assess M&E systems for TB/HIV programs at service delivery and district levels with PEPFAR-
supported IMs
 Verify data sources for TB/HIV data reported at SAPR 2015
 Validate TB/HIV data reported at SAPR 2015
 Identify challenges faced in collecting, recording, collating, and reporting data at sites

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 7
2. DATA AND METHODS

2.1 Design
A cross-sectional rapid assessment was conducted nationwide using both qualitative and quantitative
methods of data collection.

2.2 Study Sites

The target population included all districts and service outlets supported by PEPFAR that reported
HIV/TB data at SAPR 2015 (for the period October 2014–March 2015). A total of 1,007 facilities in 111
districts provided TB/HIV data.

2.3 Sample
Mixed sampling methods were used to select the DQA sample. Initially, a stratified random sampling
approach was used based on the MoH regional performance monitoring health zones as sampling strata.
To ensure maximum representation, purposive sampling was also done to supplement the random
sample. Final adjustments were
made to the sample for
logistical and security reasons.
Through these approaches, a
representative sample of 106
facilities was selected. The
facilities were drawn from 63 of
the 112 districts (56%) in the
country (Figure 2.1). Annex 1
shows the districts and facilities
selected for the DQA.

The sample of 106 facilities was

representative of 1,007
facilities. Table 2.1 shows how
the three indicators were
distributed among the facilities.
The percentage of HIV-infected
individuals in clinical care
settings reported as screened
for TB symptoms at the last
clinic visit during the period at
the 1,007 facilities was 79%,
while the percentage of
registered new and relapsed TB
patients that had documented
HIV status was 85%. For the Figure 2.1. Districts Where DQA Conducted, August 2015
same period, the percentage of
HIV-positive new and relapsed registered TB cases on ART during TB treatment was 79%.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 8
For the sample of 106 facilities assessed in the DQA (Table 2.1), data on these indicators was
downloaded from DHIS 2 just before the DQA field visits started. These data were found to be
comparable to data reported by all facilities at SAPR. The percentage of HIV-infected individuals in
clinical care settings that was reported as screened for TB symptoms at the last clinic visit during this
period was 80%. The percentage of registered new and relapsed TB patients that had documented HIV
status was 88%; while the percentage of HIV-positive new and relapsed registered TB cases on ART
during TB treatment was 79%. Thus, the selected sample was broadly representative of all 1,007
facilities that reported TB/HIV data at SAPR.

Table 2.1. Comparison of Indicators Assessed in DQA: All Reporting Facilities vs. DQA Sample,
August 2015
Indicator Value Derived From
SAPR 2015 DHIS 2 at DQA
Indicator (N=1,007) Onset (N=106)
1. Percentage of HIV-infected individuals in clinical care settings
that were reported as screened for TB symptoms at the last
clinic visit during this period 79 80
2. Percentage of registered new and relapsed TB patients that has
documented HIV status 85 88
3. Percentage of HIV-positive new and relapsed registered TB
cases on ART during TB treatment 79 79

2.4 TB Indicators Assessed

For this DQA, three TB/HIV indicators reported at SAPR were assessed. The seven variables shown below
were collected to generate these indicators. Details about the indicators and variables are provided in
Table 2.2.

 Percentage of PLHIV in HIV clinical care settings who were screened for TB symptoms at the last
clinical visit based on the following variables:
1. The number of PLHIV who were screened for TB symptoms at the last clinical visit to an HIV care
facility during the reporting period
2. Number of HIV-positive adults and children who received at least one of the following during
the reporting period: clinical assessment (WHO staging), CD-4 count, or viral load assessment in
pre-ART care
3. Number of HIV-positive adults and children who received at least one of the following during
the reporting period: clinical assessment (WHO staging), CD-4 count, or viral assessment load in
ART care

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 9
 Percentage of registered new and relapsed TB cases with documented HIV status based on the
following variables:
4. Number of registered new and relapsed TB cases with documented HIV status, during the
reporting period
5. Total number of registered new and relapsed TB cases during the reporting period
 Percentage of HIV-positive new and relapsed registered TB cases on ART during TB treatment, based
on the following variables:
6. The number of registered TB cases with documented HIV-positive status who start or continue
ART during the reporting period
7. The number of registered TB cases with documented HIV-positive status during the reporting
period

Table 2.2. TB/HIV Indicators Reported at SAPR 2015

Indicator
Code Indicator Data Variables Constituting Indicator Female Male
Number of PLHIV who were screened for TB
symptoms at the last clinical visit to an HIV care
facility during the reporting period
Percentage of Numerator
PLHIV in HIV Age <15 years
clinical care
Age 15+ years
who were
TB_SCREEN Number of HIV-positive adults and children who
screened for
received at least one of the following during the
TB symptoms
reporting period: clinical assessment (WHO
at the last
Denominator staging) OR CD4 count OR viral load
clinical visit
# Active Pre-ART
# Current on ART
Percentage of Number of registered new and relapsed TB
registered cases with documented HIV status, during the
new and Numerator reporting period
TB_STAT relapsed TB HIV status: Positive
cases with HIV status: Negative
documented Total number of registered new and relapsed TB
HIV status Denominator
cases, during the reporting period
Percentage of Number of registered TB cases with
HIV-positive Numerator documented HIV-positive status who start or
new and continue ART during the reporting period
relapsed
TB_ART
registered TB
cases on ART Denominator Number of registered TB cases with
during TB documented HIV-positive status during the
treatment reporting period

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 10
2.5 Data Sources
The above stated data elements are recorded as part of routine clinical management in standard data
collection tools at facilities, and summarized periodically in standardized reporting forms. Variations
may occur in facilities depending on availability of tools. However, with harmonization of information
systems in recent years, the latter may be less prevalent. Where electronic medical records systems are
used, the information may be maintained in standardized electronic databases.

Table 2.3 summarizes the standard data sources for the seven variables at the site level.

Table 2.3. Data Sources for TB/HIV Data Assessed in DQA by Variable, August 2015
Variable Data Sources
1. The number of PLHIV who were screened for TB symptoms at the
last clinical visit to an HIV care facility during the reporting period Pre-ART and ART register
2. Number of HIV-positive adults and children who received at least
one of the following during the reporting period: clinical assessment
(WHO staging) OR CD4 count OR viral load — Pre-ART care Pre-ART register
3. Number of HIV-positive adults and children who received at least
one of the following during the reporting period: clinical assessment
(WHO staging) OR CD4 count OR viral load — ART care ART register
4. Number of registered new and relapsed TB cases with documented
HIV status, during the reporting period Unit and District TB registers
5. Total number of registered new and relapsed TB cases, during the
reporting period Unit and District TB registers
6. Number of registered TB cases with documented HIV-positive status
who start or continue ART during the reporting period Unit and District TB registers
7. Number of registered TB cases with documented HIV-positive status
during the reporting period Unit and District TB registers

2.6 Data Collection Tool

The Routine Data Quality Assessment (RDQA) tool developed by Measure Evaluation
(http://www.cpc.unc.edu/measure/resources/tools/monitoring-evaluation-systems/data-quality-
assurance-tools/data-quality-assurance-tools) was customized for the TB/ HIV indicators (Annex 2). The
RDQA tool is an electronic Excel tool that can be adapted to the technical area for review. The tool is
comprised of data verification, validation, and M&E systems assessment sheets and dashboards. The
data entry screen and forms are comprised of:

 Facility identification variables (district, sub-county, facility name and level to facilitate
disaggregated analysis by IP, district, and facility level).
 Verification of data sources: For each of the seven variables constituting the three indicators,
information on availability of data sources, completeness of data sources, and availability of data for
the period under review was entered (as Yes/No) along with relevant comments.
 Data validation sheet: For each variable, data was entered separately for each of the two quarters
under review (i.e., October–December 2014, and January–March 2015). The data entered
comprised the latest information downloaded from DHIS 2 on July 26, 2015, i.e., a day before
commencement of field work and the joint manual recount by field teams. Calculation of the

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 11
 disparity between the two values was automated and displayed with relevant color codes (indicating
no disparity, acceptable disparity, and excessive disparity). Relevant comments, especially on the
disparity between the data, were also entered. This was also done for cross-check 1 in the tool,
where data in the site level HMIS 106a report copy retained at the facility was compared with the
manual recount for each variable. Comments were also recorded in cases of significant disparities.
Cross-check 2 in the tool involved data from the district TB register being compared with the manual
recount and disparity calculated based on similar premises.
 Five components were assessed for M&E systems at the district and facility levels: 1) M&E structure,
functions and capabilities to handle HIV/TB data; 2) understanding of indicator definitions and
reporting guidelines; 3) availability of data collection and reporting forms for TB/HIV services; 4)
data management processes; and 5) use of data for decision making. The requirements under each
component were comprised of:
 District M&E System Assessment
– M&E structure, functions, and capabilities to handle TB/HIV data assessed based on the
requirements of: 1)availability of staff who regularly review the quality of TB/HIV data; 2)
whether there is a staff designated to review quarterly reports before submission to the
next level; 3) whether a team approach is followed so that even in the absence of
designated staff, reports are still submitted; 4) whether the district provides regular
feedback to facilities on quality of their reports; 5) whether regular feedback is provided to
district from national level; 6) whether the district conducts regular support supervision to
facilities on TB/HIV data; and 7) whether the district receives regular support supervision of
TB/HIV integration from the national level.
– Understanding of indicator definitions and reporting guidelines assessed based on the
requirements of: 1) understanding what should be recorded in source documents; 2)
understanding what should be included in the quarterly report (HMIS 106a TB/Leprosy
section); 3) understanding how the report should be submitted; 4) understanding to whom
the reports should be submitted; 5) understanding when the reports are due; and 6)
availability of written guidelines or standard operating procedures (SOPs) on standardized
recording and reporting of TB/HIV data.
– Availability of data collection and reporting forms for TB/HIV services assessed based on the
requirements of: 1) availability of the national TB register for reporting at this level; 2)
whether standard TB/HIV reporting forms are consistently used at health facilities; 3)
whether the multiple IPs in the district supporting TB/HIV services use national tools and
follow the same reporting timelines; 4) availability of sufficient quantities of blank registers
at the district; 5) whether the district consistently monitors availability of standard recording
and reporting tools at facilities including forecasting needs.
– Data management processes assessed based on the requirements of: 1) whether the district
has computerized TB/HIV data and reports; 2) if so, whether they have quality control
procedures in place for entry of data from paper forms and registers to the computer
database; 3) whether the district routinely creates back-up files for the TB/HIV data; 4) if so,
whether the date of the latest backup was appropriate; 5) whether the district has a written
procedure to address late, incomplete, inaccurate, and missing reports including following
up with service points on data quality issues; 6) whether the district has ever uncovered
errors in the reports from health facilities; 7) whether the district has a written policy on
how long source documents and reporting forms should to be retained; and 8) whether the

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 12
 district has a written policy on how program documents (e.g., source documents and
reporting forms) should be archived (e.g., filing cabinets, storage rooms, etc.).
– Use of data for decision making assessed based on the requirements of: 1) evidence of
TB/HIV data use at the district in the form of graphs, maps, charts, etc.; 2) availability of staff
at the district who take the lead in analysis and interpretation of TB/HIV data; 3) timely
dissemination of TB/HIV data to inform programmatic decisions; and 4) programmatic
decisions taken by the district based on analyzed TB/HIV data/results.
 Facility M&E System Assessment
– M&E structure, functions and capabilities to handle HIV/TB data assessed based on the
requirements of: 1) assignment of staff to regularly review the quality of TB/HIV data; 2)
whether a specific staff is designated to review the quarterly reports before submission to
the next level; 3) whether a team approach is followed so that even in absence of
designated staff, reports are still submitted; 4) whether the facility receives regular
feedback on quality of their reports; 5) whether the facility receives regular support
supervision for TB/HIV data; and 6) whether staff that handle TB/HIV data have been
trained/oriented in data management.
– Understanding of indicator definitions and reporting guidelines assessed based on the
requirements of: 1) understanding what to record in source documents; 2) understanding
what to include in the quarterly report (HMIS 106a TB/LP section); 3) understanding to
whom the report should be submitted; 4) understanding when reports are due; 5)
understanding how to effect changes to a previously submitted report; 6) availability of
written instructions for recording and reporting TB/HIV data; 7) understanding of variables
for the indicator “percent of PLHIV screened for TB”; 8) understanding of the variables for
the indicator “percent of TB cases with documented HIV status”; and 9) understanding of
the variables for the indicator “percent HIV-positive TB cases on ART.”
– Availability of data collection and reporting forms for TB/HIV services assessed based on the
requirements of: 1) whether the health facility has the national HMIS forms/tools to be
used at their reporting level; 2) if yes, whether the standard forms/tools are consistently
used at the health facility; 3) if there are multiple organizations that are supporting TB/HIV
services at the site, whether they use the national reporting forms and follow the same
reporting timelines; 4) whether there are sufficient stocks of blank primary data collection
tools /registers and summary HMIS forms at the facility; 5) whether the site has a regular
refill program of the TB/HIV tools when they are used up; 6) whether the primary data
collection tools/registers have all relevant variables needed to compile the HMIS 106a
reports.
– Data management processes assessed based on: 1) whether the established patient
recording and reporting system allows tracking of unique individuals within and across
service delivery points to avoid double counting at the site; 2) whether there are data
quality controls in place for ensuring compilation of accurate quarterly (HMIS 106a) reports
(e.g., controls for detection of data inconsistencies, incomplete/incorrect TB/HIV reporting,
missing data, transcription errors); 3) whether facilities have computerized data and reports;
4) if so, whether there are quality controls in place for when data from registers/paper-
based forms are entered into a computer to ensure the accuracy of data entry (e.g., edit
and/or logic checks, post-data entry verification, etc.; 5) whether confidentiality of patients
information is maintained according to national guidelines; 6) whether there are SOPs/job
aids/guidelines that describe how site documents should be handled and archived clearly

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 13
 written and accessible to all staff; 7) if yes, whether the staff were aware and using the
SOPs/job aides/guidelines; 8) whether the data archiving/storage system at the facility is
adequate; and 9) whether the HMIS 106a report is submitted on time.
– Use of data for decision making assessed based on the requirements of: 1) whether there is
evidence of TB/HIV data use at the facility; 2) whether TB/HIV services staff have been
trained in data analysis and interpretation; 3) whether there is staff member at the facility
who takes the lead in analysis and interpretation of TB/HIV data; 4) whether analyzed
data/results are presented/disseminated to other information system stakeholders in a
timely manner so that the information can be used to inform decisions; and 5) whether
there programmatic decisions taken by the facility based on analyzed TB/HIV data or results.
For the systems assessment, for all variables under the five components of the systems assessed, the
tools provided for entry of the following pre-coded responses: Yes completely, Partly, Not at all, or Not
applicable. In addition, where appropriate, relevant comments were entered as text.

The data verification and validation tool (Annexes 3a, 3b, and 3c) was used to abstract detailed
information that aided further analysis of each indicator variable by age and sex and included:

 A feedback form (Annex 3d) aided summarization of findings of the DQA for use by field teams to
provide preliminary feedback to facility staff. The feedback form highlighted good practices and
areas needing improvement.
 An action plan template (Annex 3e) aided field teams working with facility staff and management to
identify four main weakness for which action plans for improvement were developed.

Copies of the feedback form and action plan were left at the facility.

2.7 Training of Data Collection Teams

Multi-agency and multi-disciplinary field teams were made up of staff from MEEPP, PEPFAR-funded IMs,
Track TB, MoH (NTLP, AIDS Control Programme (ACP), and Resource Centre (RC)), Regional Performance
Monitoring Teams (RPMTs), DTLS, and health facility staff. The teams were trained in a 2-day workshop
conducted by MEEPP, MoH, CDC/USAID, and PEPFAR staff. A total of 65 individuals were trained and
distributed into 11 field teams. Didactic sessions were held on the first day of the training, followed by a
1-day pilot test assessment in five facilities in Mukono district and the office of the District Health Officer
(DHO). On the third day, lessons learned from the pilot visits were shared, and SOPs and data collection
instruments were reviewed. Schedules and plans for data collection were also developed. The training
and pilot was conducted during July 22–24, 2015.

2.8 Pilot Testing of Tools and Field Procedures

All data collection tools and methods were pretested in five facilities in Mukono district that were not
part of the main sample, i.e., Mukono CoU Health Centre (HC), Mukono HC IV, Nakifuma HC IV,
Nagalama hospital, and Goma HC III; as well as the DHO’s office. All trained field workers took part in the
pilot testing. Following the pre-test, experiences were shared, and appropriate modifications were made
to survey tools and deployment plans based on lessons from the pilot.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 14
2.9 Field Data Collection Procedures
Data for the DQA were collected through three complementary and sequential activities, and were
entered directly in the Excel tool at each facility:

1. At the facility and district levels, M&E systems were reviewed through semi-quantitative approaches
involving structured questionnaire-guided, in-depth interviews with a team of facility staff.
2. In the second part, verification of data sources was conducted through document review to
ascertain availability and completeness of primary data collection sources for TB/HIV data in the
facilities, primarily the pre-ART and ART registers, and the Health Unit TB registers.
3. Data validation was conducted through manual recount of outputs for the seven variables
conducted jointly by field teams and facility staff to determine the level of accuracy or the degree to
which onsite validated counts represented the SAPR 2015 results.
Additional data were collected through on-site observations and review of primary and secondary data
sources at the sites. Interviews were conducted with facility staff and district staff implementing TB/HIV
activities.

Visits to facilities for data collection began immediately following training. A total of 11 teams (Annex 4)
were deployed, one team to each of the regions. District teams and health facilities were notified about
the DQA by IPs prior to the site visits. The assessment at facilities began with introductory discussions
with the district and facility staff about the DQA processes. At the district, the team lead introduced the
objectives of the assessment to the DHO, Biostatistician, and DTLS. The DQA team then conducted the
district assessment and abstracted the SAPR 2015 for the facilities selected from that district before
proceeding to the facilities. The abstracted data was used in cross-check 2 described above. The DQA
team was joined by the DTLS during its facility visits. At the health facilities, team leads introduced the
objectives of the assessment to facility management teams, TB and HIV clinic in-charges, and facility
staff.

After data collection, the field team synthesized and summarized the findings onto site feedback forms
that were used along with various dashboards to provide feedback to facilities, districts, and IPs. During
feedback, field teams highlighted achievements and weakness and made recommendations to the
facility and IPs, and together with facility teams, developed action plans. Mukono district was excluded
from the district-level assessment because it was part of the pilot. It was not possible to conduct the
district M&E assessments in the Butambala or Kiryandongo districts, nor in the Makindye Division in
Kampala. Efforts to get in touch with the DTLS officers in these districts were not successful. Thus,
district M&E assessment was conducted in 61 out of the 65 districts visited.

2.10 Data Management

2.10.1 Data Entry
Data were entered in an electronic Excel database in real time at point of collection with hard copy
backup. The Excel database enabled entry of numerical and text information as appropriate, and had
dropdown lists to facilitate entry of structured responses. The pre-coded options facilitated accurate
data entry.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 15
2.10.2 Data Cleaning
Copies of completed templates were submitted to the central office daily for preliminary review.
Inconsistencies detected were referred back to field teams for correction. At the central level, data from
all sampled facilities were pooled in Excel before analysis. Specific composite sheets were created for
each theme of the assessment including verification and data validation, systems assessment, main
weaknesses, and recommendations. The key variables for each facility were linked and transferred to
the composite sheets where each row identified a facility and the columns represented variables.
Further cleaning was performed on the pooled data before data analysis.

2.10.3 Data Analysis

Two levels of data analysis were conducted: preliminary analysis at the site level, and analysis of pooled
data at the central level.

Preliminary analysis at the site level was automated to provide summary information to aid feedback to
facilities. Pre-programmed dashboards in Excel spreadsheets displayed summary statistics with
automated color codes indicating areas of strong or weak performance. The composite score for each of
the five components of the M&E system, as well as for the system overall, were automatically calculated
as percentages and displayed on the dashboard. In addition, formulas to calculate summary statistics,
such as deviation of the reported from verified counts and composite scores, were built into the
spreadsheets.

The percentage deviation for quantitative data was derived as follows:

Percentage Result in Report Under Review – Joint Recount Figure

difference for each = x 100
variable Joint Recount Figure

The color coding of the cells displayed green when there was minimal disparity (<+5%) between
reported and validated count, yellow when there was an acceptable level of disparity (+5 to +10%), and
red when disparity exceed +10%. Preliminary site-level analysis aided preliminary feedback and
identification of areas in need of remediation.

Analysis of pooled data at the central level was done with a focus on national, district, and facility level
issues. The quantitative data were supplemented with qualitative data to get a better understanding of
issues at the various levels. Thematic analysis was conducted on the qualitative data to generate
common themes. Likert scale responses were transformed into quantitative data: Responses of “yes
completely, partly, not at all, and not applicable” were assigned scores of 3, 2, 1, and 0, respectively and
averaged. The score per component ranged from 0–3 and was categorized as follows:

 Less than 2.25 denoted inadequate systems

 2.25–2.74 denoted acceptable systems
 2.75–3.00 denoted adequate systems
For qualitative data, simple descriptive statistics were used to generate frequency distributions, and
data were summarized in graphs or charts. Where possible, cross-tabulations were done to analyze data
by site level.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 16
Analysis of pooled data was conducted by facility, by facility level, and by district. For Kampala district,
analysis was done by division because, in this district, each division functions as an independent district
in the health system. For site level analysis, key findings were summarized in a table reflecting the
performance for each site, with appropriate color codes that provide at-a-glance site performance, as
well as key areas that performed well or needed improvement. Data analysis was conducted using both
Excel and SPSS.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 17
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 18
3. RESULTS

3.1 Descriptive Characteristics of DQA Participating Facilities

The 106 facilities comprised 20 (18.9%) hospitals (including two regional referral hospitals), 28 (26.4%)
HC IVs (including one specialized clinic [TASO]); and 58 (54.7%) HC IIIs. The distribution of the facilities
by level, region, and IP is summarized in Table 3.1.

Of the 11 RPMT health regions in the country, Central/Kampala region had 21 facilities in the sample,
followed by Mbarara Region (16), Mbale (12), and Jinja and Masaka each with 10 facilities. Hoima region
had three facilities.

The facilities in the sample were receiving direct support from 17 IMs, with 19 facilities in the sample
supported by Baylor College of Medicine, followed by STAR SW and Mildmay, each with 12 facilities.

Kalangala DHO HCT and Care Program, and the U.S. Department of Defense (DoD)/UPDF each had one
facility in the sample, while Makerere University Joint AIDS Programme (MJAP), Makerere University
School of Public Health (MUSPH), and Strengthening Uganda’s Systems for Treating AIDS Nationally
(SUSTAIN) each had two as shown.

Table 3.1. Description of Facilities Participating in DQA, August 2015

Number of
Facilities Percent
Characteristic N=106 %
Facility Level
General hospitals 20 18.9
HC IV 28 26.4
HC III 58 54.7
Supporting Implementing Partner
ASSIST 11 10.4
Baylor 19 17.9
CEM/ Uganda Private Health Support Program 3 2.8
DOD/UPDF 1 0.9
STAR SW 12 11.3
Infectious Disease Institute (IDI) 4 3.8
STAR EC 7 6.6
Kalangala DHO/HCT & Care 1 0.9
Mildmay 12 11.3
STAR E 8 7.5
MJAP 2 1.9
MUSPH 2 1.9
TASO 7 6.6
Uganda Episcopal Conference 6 5.7
Uganda Protestant Medical Bureau (UPMB) 3 2.8
SUSTAIN 2 1.9
Walter Reed 6 5.7
Regional Performance Monitoring Health Region

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 19
Table 3.1. Description of Facilities Participating in DQA, August 2015
Number of
Facilities Percent
Characteristic N=106 %
Arua 9 8.5
Fortportal 7 6.6
Gulu 7 6.6
Hoima 3 2.8
Jinja 10 9.4
Central/Kampala 21 19.8
Lira 5 4.7
Masaka 10 9.4
Mbale 12 11.3
Mbarara 16 15.1
Soroti 6 5.7

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 20
3.2 Objective 1a: Assessment of District M&E Systems
Summary Findings
39% of districts had adequate M&E systems.
46% of districts had adequate functions and capabilities to handle HIV/TB data.
53% of districts had adequate understanding of indicator definitions and reporting guidelines.
66% of districts had adequate supply of data collection tools and reporting forms for TB/HIV data.
67% of districts had inadequate TB/HIV data management processes.
51% of districts had adequate data use for decision making.

Ratings of all districts by M&E component assessed are shown in Annex 5. Of the 61 districts, including
two divisions in Kampala, M&E systems in 39% were found to be adequate, and 21% were found to be
inadequate (Figure 3.1). Six districts had systems that were only partly adequate or not adequate at all.
These districts were Adjumani, Butaleja, Buyende, Dokolo, Hoima, and Kanungu.

Further analysis of the individual

21% sites M&E systems assessment
rated components revealed differences as
Inadequate 39% sites indicated in Figure 3.2.
(Score: rated
<2.25) Adequate Overall, districts had adequate
(Score: understanding of reporting
2.75 - 3.00) guidelines and indicator definitions
as well as data collection and
reporting tools. M&E structures,
functions, and capabilities to handle
39% sites TB/HIV data and use of data for
rated decision making were rated
Acceptable acceptable. Among the five
(Score: components of the systems, data
2.25-2.74) management processes constituted
the area of weakest performance.
Figure 3.2 shows the performance of
Figure 3.1. Overall Rating of District Level M&E Systems for district M&E systems based on each
TB/HIV Data, August 2015 of the five components assessed.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 21
 Understanding
Understanding of
of indicator
Indicator definitions
Definitions and
and 3
Reporting Guidelines guidelines
reporting

Availability of
Availability of Data-collection
data collection Tools
tools and
and 2.8
reportingForms
Reporting formsfor
forTB/HIV
TB/HIVsevices
services

M&E
M&E structure,
Structure, functions,
Functions andand capabilities
Capabilities to 2.7
handletoHIV/TB
handleinformation
HIV/TB information

Use of data for decision making 2.5

Use of data for decision making

Data management processes

Data Management Processes 2.1

0 1 2 3

Figure 3.2. Summary District-level M&E Systems Assessment for TB/HIV Data by
M&E Component, August 2015

3.2.1 Functions and Capabilities of Districts to Handle HIV/TB Information

Overall, about 46% of districts had structures, adequate functions, and capabilities for handling TB/HIV
data. On the other hand, for 51% of the districts, the functions and capabilities to handle HIV/TB data
were only partially proficient and 3% sites
will need remediation. In a few rated
districts, the functions and Inadequate
capabilities to handle HIV/AIDS (Score:
information are inadequate and in <2.25)
need of urgent remediation (Figure 46% sites
3.3). rated
Adequate
The proportion of districts that were 51% sites (Score: 2.75
proficient in the areas assessed rated - 3.00)
under “Functions and Capabilities of Acceptable
District Structures to Handle TB/HIV (Score: 2.25-
2.74)
Data” is summarized in Figure 3.4.

With the exception of one district

(Jinja), all districts had staff
designated to review the quality of Figure 3.3. Overall Rating of District M&E Structures, Functions,
TB/HIV data. In these districts, the and Capabilities for Management of TB/HIV Data, August 2015
DTLS moves to TB/HIV facilities and
updates the district TB registers quarterly. Quarterly reports are submitted to district biostatisticians.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 22
Of the 61 districts, 85% had staff designated to review district TB/HIV reports (HMIS 106a) before
submission. Three districts (Busia, Buyende, and Pader) reported having a staff member assigned part-
time to reviewing reports, while Buvuma, Rubaga Division, Kisoro, Mubende, Nakasongola, Sironko, and
Wakiso districts didn’t have any staff assigned to reviewing reports.

Yes completely Partly Not at all NA

Receive regular support supervion from National Level

Regular feedback from National Level

Conduct regularly supervision

Provide feedback to facilities on quality of reports

Team approach to report submission

Designated staff to review report

Saff to review data quality

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.4. M&E Structures, Functions, and Capabilities of Districts for Management of
TB/HIV Data, August 2015

Approximately 67% of the 61 districts had a team approach to submission of HMIS 106a reports to the
extent that even in the absence of a designated officer, timely submission of report would not be
affected. The staff involved in ensuring that the report is submitted are indicated in Table 3.2.

Of the 61 districts, 56 (91.8%) regularly provided feedback to facilities on the quality of their TB/HIV
reports. In most cases, feedback was provided quarterly as indicated by 36 of the 41 (88%) districts that
provided responses to the frequency of feedback. Monthly feedback was cited by four districts, and
weekly by one district. The feedback took several forms including quarterly performance review
meetings, quarterly data review meetings, and supervision visits.

The frequency of the various forms of feedback is shown in Table 3.2. Feedback was mainly provided by
the DTLS (cited in 50% of districts) and the district biostatistician (cited by 34% of districts). Others
included the HMIS Focal Person (FP), District Health Educator, Records Officer, and Health Sub-District
TB FP. The five districts that didn’t provide regular feedback were Kamuli, Kaliro, Kibaale, Nebbi, and
Zombo. Nebbi and Zombo indicated that they used to conduct quarterly performance review meetings
but stopped due to lack of funds.

Slightly fewer districts (53) provided regular support supervision to TB/HIV services in facilities.

Most (80%) of the 44 districts that indicated the date of the last supervision visit to the health facility
indicated that it had been within 1 month of the DQA (between June and August 2015). Six districts had
last conducted supervision visits 2 months prior to the DQA, and three districts had done so 3 months
before the DQA. Fourteen of the 23 districts indicated that supervision was conducted quarterly, while
nine conducted monthly supervisions for TB services.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 23
The supervision was either targeted or integrated as was the case in 17 districts. In most districts (16 of
22, or 72%), TB-related supervision was conducted by the DTLS. The RPMT, NTLP, IP, HIV FP and the
assistant DHO are the other officers
that were involved in support Table 3.2. District M&E Functions and Capacities for Handling
supervision of TB services. The seven TB/HIV Data, August 2015
districts that didn’t provide support No. of
supervision of TB/HIV to facilities Staff Districts
were Bundibugyo, Dokolo, Gulu, Staff designated to review quarterly report before submission*
Rubaga Division, Kibaale, Mubende, RTLFP 17
and Nebbi, largely on account of lack DTLS 23
of funds as was pointed out in five District biostatistician 16
districts. During regional meetings 3
HMIS FP 7
Three-quarters of the districts (46)
DHO 1
indicated that they receive regular
No response 5
feedback from the national level on
Team approach to report submission
the quality of their TB/HIV reports.
Team work by other member of District Health Team
The feedback is mainly provided
(DHT) including DTLS, District biostatistician, HMIS
through quarterly regional
FP, DLFP, records assistant 25
performance review meetings as
reported by 26 of the 31 districts Sub-county TB FP 10
that indicated the frequency of the Health facility staff 2
feedback. Among the districts that DTLS predecessor 1
reported receiving regular feedback, No response 2
most (25) had done so within 1 Format/forum of feedback to facilities on quality of reports*
month of the DQA (June–July 2015), During quarterly district review meeting
six had received it 2–3 months prior (performance and data) 33
to the DQA (April–May), while five During site visits (data collection, supervision) 19
districts last received feedback 4–6 Phone call/SMS/MTRAC 13
months prior to the DQA. In most Email 1
instances, feedback on the quality of Format/forum of receiving feedback on quality of reports*
reports was provided by the NTLP During quarterly district review meeting
(cited by 13 districts), Regional (performance and data) 26
coordination offices (RTLFP, RPMT; Phone call/SMS/MTRAC 3
11 districts), RC (8 districts), and IPs During site visits (data collection, supervision) 1
(in six districts). It was worth noting Manual report 1
that ACP has not been providing District receives regular supervision from*
regular feedback to the districts and Regional teams (RTLFP, RPMT) 14
this was highlighted in the NTLP 10
qualitative data from Luuka, Other MoH officers( QA, EPI, Unclassified MoH) 9
Mayuge, Jinja, Busia, and Butaleja: IPs 5
“Feedback on the side of HIV care UAC 2
has not been very regular. Also *Some districts gave more than one response
Resource Center stopped one year
ago.” Another caption from Busia district indicated that “IP, Resource Centre, NTLP gives feedback. ACP
does not give feedback especially when there is poor performance in TB/HIV.”

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 24
A sizeable number of districts didn’t receive regular feedback from any of the above institutions.

Only 34 (just over half of the districts assessed) received regular support supervision for TB/HIV services
from the national or regional level. Support supervision was often provided by the RTLFP, RPMTs, and
central MoH (NTLP) as shown in Table 3.1. The majority of districts (22) that received regular support
supervision had done so within 1 month of the DQA (June–August 2015). Four received supervision
during March and April 2015, while two did so during the first 2 months of 2015. Two districts
supervised in 2015 could not
specify the date of the visit. 2% sites
rated
3.2.2 Understanding of Inadequate
(Score:
Indicator Definitions and
<2.25)
Reporting Guidelines
Figure 3.5 shows the overall rating 47% sites
53% sites
of the districts on their rated
rated
Acceptable
understanding of reporting Adequate
(Score: 2.25-
guidelines and TB/HIV indicator 2.74)
(Score: 2.75
definitions. Over 90% of the - 3.00)
districts have adequate or
acceptable understanding of
indicator definitions and reporting
guidelines. Figure 3.5. Overall Rating of Understanding of Reporting
Guidelines and Indicator Definitions, August 2015
Figure 3.6 shows the performance
of districts for each requirement assessed under “Understanding of Reporting Guidelines and Indicator
Definitions.”

Only six districts (Kibaale, Busia, Butaleja, Maracha, Soroti, and Tororo) did not have adequate
understanding of what should be recorded in the source documents.

Yes completely Partly Not at all NA

District specific Guidelines on recording and reporting TB

data

Understanding Reporting Guidelines; due date for report

Understanding Reporting Guidelines; where to submit the

report
Understanding Reporting Guidelines; how to submit the
report
Understanding what should be included in the Quarterly
report
Understanding what should be recorded in the source
documents

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.6. District Knowledge of Reporting Guidelines and Indicator Definitions, August 2015

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 25
Similarly, only four districts (Apac, Dokolo, Shema, and Kibale) failed to demonstrate proper
understanding of what should be included in the HMIS 106a report (TB section). In these districts, the
DTLS appeared to know what should be included in the HMIS but the biostatistician and HMIS FPs did
not.

All districts except Dokolo had good understanding of how reports should be submitted. All districts
except two (Dokolo and Maracha) were knowledgeable as to where quarterly reports were to be
submitted. Three districts (Dokolo, Kyegegwa, and Maracha) did not know the due date for submission
of TB/HIV reports.

Availability of district-specific instructions to ensure standardized recording or reporting of TB/HIV data

was the requirement that manifested most weaknesses under this component. Only 25 districts were
compliant with this requirement. The remaining districts were using instructions in the HMIS manual,
guidelines within the forms, or cover pages of the registers. Some districts considered these instructions
adequate.

3.2.3 Availability of Data Collection Tools and Reporting Forms for TB/HIV Data
Districts performed fairy well on this component of the assessment, with two-thirds (66%) or 40 districts
rated adequate (Figure 3.7). Five
districts were rated inadequate
and need urgent remediation. 8% sites
rated
Examination of the how districts Inadequate
performed on the various (Score:
requirements revealed that <2.25)
having a copy of the national TB 26% sites
register for reporting was the rated
area of best performance, while Acceptable
having sufficient stocks of blank (Score: 2.25-
registers available in the district 2.74)
was the weakest (Figure 3.8).

Only one district (Mbale) did not 66% sites

have the national TB register for rated
Adequate
reporting at their level.
(Score: 2.75 -
However, the reasons why were 3.00)
not clear or consistent. Two
other districts (Busia and Figure 3.7. Overall Rating of Availability of TB/HIV Data Collection
Butaleja) had some registers but and Reporting Tools, August 2015
not enough for all the facilities,
especially HC IIs.

Rubaga division and Pallisa were the only two districts not consistently using the district TB register and
other standardized TB/HIV tools. For instance, in Pallisa district, it was noted that one TB client from
Apopong HC III had not been registered in the district TB register.

In the districts of Jinja, Hoima, and Rubaga Division of Kampala, the multiple IPs supporting TB/HIV
services in the district were not consistently using national recording and reporting tools for TB/HIV
services or following national reporting timelines. In Hoima, the IP that supported services in the

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 26
refugee resettlement camps in Kyangwali used parallel reporting and recording forms. In Jinja district,
there were several IPs using their own specific tools to collect data from the facilities, though data is
collected by the partner staff, not facility staff. In Rubaga Division of Kampala, the IP had different
reporting formats and their reporting timeline was mid-month.

Yes completely Partly Not at all NA

District
Districtmonitors
moinitorsconsistent
consistentavailability
availabilityofoftools
toolsatat
facility
facilitylevel
level
Sufficient quantity of blank registers available at the
district

Multiple partners use standard National tools

Standard forms consistently used at facility

Have copy of National TB register for reporting

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.8. Availability of TB/HIV Data Collection and Reporting Tools Among Districts, August 2015

It was noted that most of the registers were printed by IPs and that new forms and registers were not
yet sufficiently rolled out. The other reasons for lack of sufficient quantities of registers are summarized
in Table 3.3.

For effective information

system management, Table 3.3. Availability of Data Collection and Reporting Tools at District,
districts should August 2015
consistently monitor No. of
availability of sufficient Reasons for Insufficient HIV/TB Tools Districts
stocks of data
Inadequate or no stock of tools 16
collection/reporting forms
No forecasting 8
at facilities to ensure their
Has old tools of TB register 8
continuous availability
Orders not honored/delayed 5
and to forecast need as
No blank copies of district TB register 5
appropriate. This
requirement was also
assessed in the DQA, where 45 districts were found to be consistently compliant, nine partially
compliant, and seven not compliant at all. The districts that did not comply with this requirement had a
range of reasons for not doing so, including relying on IPs to supply them, forecasting being undertaken
by the IPs, the IP delivering the tools directly to facilities and bypassing districts, orders being delayed,
and some districts ordering whenever need arises.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 27
3.2.4 TB/HIV Data Management Processes in Districts
As part of the M&E system assessment, district TB/HIV data management processes were also assessed.
Overall, districts scored poorly on this component with over 60% of the districts having inadequate
processes in place (Figure 3.9). Under this component, several parameters were assessed (Figure 3.10).
Districts performed best on the parameter of uncovering errors in facility reports. The areas of most
weaknesses were in backup of computer databases, written procedures for retention of source
documents, and written
procedures for archiving
of source documents. 8% sites rated
Adequate
About one-half (32) of (Score: 2.75 -
the districts indicated 3.00)
that they had 25% sites
computerized TB/HIV rated
Acceptable
data and reports.
(Score: 2.25-
However, most of these 2.74)
(29 of 32) were actually
referring to the DHIS 2.
Overall, about 44 (72%)
districts indicated that 67% sites
they were using DHIS 2. rated
Responses included 29 Inadequate
of 32 that indicated “yes (Score: <2.25)
completely,” 5 of 5 that
indicated “partly,” and Figure 3.9. Overall Rating of Data Management Processes for TB/HIV
10 of 24 that indicated Data, August 2015
“not at all” to the question of availability of computerized TB/HIV data. Of the remaining (3/32) districts
that indicated that they had computerized TB/HIV data, one was using an Excel format while the other
two did not indicate the format. It appears the question was largely misinterpreted to mean that the
DHIS 2 reporting system was part of the computerization of the TB/HIV data by all but one of the 30
districts that responded “yes completely.” On the other hand, most of the districts that indicated that
they did not have computerized TB/HIV data could have interpreted the question of availability of
computerized TB/HIV data to mean other databases besides DHIS 2. This would largely imply that with
the exception of one district, none of the districts have computerized TB/HIV data beyond the DHIS 2.

Among the 37 districts that indicated complete or partial computerization of TB/HIV data, 33 had quality
controls in place to use when data from paper forms were entered into a computer to ensure accuracy
of data entry. Data quality checks mentioned included checks built into the DHIS 2 application (24
districts), post-data entry verification by members of the DHT (eight districts total, with the DHT in two,
biostatistician alone in four, and biostatistician and HMIS FP in two). In one district, the DTLS verified the
data before entry. In Lwengo, there were no quality controls, and there was no response in two districts.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 28
 Yes completely Partly Not at all NA

District has written procedures on archiving of source

documents
District has written procedures on rentention of source
documents

District ever uncovered errors from facilities

District has written procedures to adress late,

incomplete and missing reports, including following…

If Yes, Is
is backup
back upappropriate
appropriate

District routinely creates back-up files for TB/HIV data

Have quality controls for data entry from paper forms

to computer

District has computerized TB/HIV data and reports

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.10. Data Management Processes for TB/HIV Data Among Districts, August 2015

Furthermore, among the 37 districts with complete or partially computerized data, 18 indicated that
they had backed up their data (14 completely, four partially). Backup media used included online, paper
hard copies, drop box, email, and flash discs.

For the 18 districts that backed up their data either completely or partly, 11 had an up-to-date backup,
while four districts had backups that were only partially up to date. Among the 11 districts that indicated
that they had an up-to-date backup, six had done so within 1 month to the DQA, one within 3 months,
and there was no response from three districts.

Also assessed as part of this component was having a written procedure to address late, incomplete,
inaccurate, and missing reports, including following-up with service delivery sites on data quality issues.
Nearly half of the districts did not have any such procedures, 24 had, and seven districts only partly so.
For districts that did, the avenues or modalities included phone calls and SMS text messaging to facilities
when errors were found in DHIS 2 or reports not submitted, sending circulars, during quarterly
performance review meetings, and monthly timeliness charts (Table 3.4).

An overwhelming majority of districts indicated that they have uncovered errors in district reports. Only
Kamuli district had never done so.

Most districts (43) lacked a written policy that stated how long source documents and reporting forms
should be retained. Only 13 fully complied with this requirement, while three only partially did so.
However, even for those that had such policy, they were not necessarily district-specific. For instance,
most districts used the HMIS manual, or borrowed guidelines on documentation practices from the
District Registry.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 29
Having a written policy that describes how program documents (e.g., source documents and reporting
forms) should be archived (e.g., filing cabinets, storage rooms) was also assessed. However, this was
found lacking in 40 districts. It was noted that even where no expressed policy was in place, some
districts had a proper archiving system. Conversely, in some districts with such policy, some files and
registers were scattered in a disorganized way.

Table 3.4. Modalities of Followup on Late Submission or Inconsistent

3.2.5 Use of Data for Reports, August 2015
Decision Making at Follow-up Modality No. of Districts
District Level Phone call and/or SMS including sending messages through
Under this MTRACK 15
component, the Circulars and warning letters 6
performance of EQA guidelines 3
districts was average, Chart displays, registry, and checklists 3
with 31 (51%) of the Discussed in meeting including DHT and quarterly review
districts meetings 3
demonstrating overall DHO follow up 2
adequate data use. Supervision 2
Just under one-third Feedback on quality of report given 2
(31%) were rated Use motorcycle to go to site 1
inadequate and in Not clarified 5
need of urgent Note: Some districts employed more than one mechanism to address late and inconsistent
remediation. This is report
summarized in Figure 3.11.

Under the various requirements for this component, strong performance was reflected in the area of
staff assignment to lead analysis of TB/HIV data, while the weakest was in the areas of evidence of use
of TB/HIV data in the district (Figure 3.12).

In just under 50% of districts,

there was evidence of use of data 31% sites
for program action. This evidence rated
Inadequate
was in the form of graphs,
(Score:
presentations at quarterly review
<2.25)
and planning meetings, facilities
quantifying and ordering TB/HIV 51% sites
medicines and other commodities rated
based on the data, and Adequate
participation on radio talk shows (Score: 2.75
on multi-drug resistant (MDR) TB, - 3.00)
for example, as summarized in 18% sites
rated
Table 3.5. In 22 districts, there was
Acceptable
no evidence of data use at all, (Score:
while 11 districts only had partial 2.25-2.74)
evidence of use of the data.

Figure 3.11. Overall Rating of Data Use for Decision Making

Among Districts, August 2015

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 30
 Yes completely Partly Not at all NA

Were there programatic decisions taken based on data

Is analysed data diseminated timely to stakeholders

Assigned staff to lead analysis of TB/HIV data

Assigned staff to develop TB/HIV charts, graphs etc

Evidence of use of data in the district

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.12. Data Use for Decision Making Among Districts, August 2015

Thirty-six districts indicated that to facilitate data use for informed decision making, they had specifically
assigned an officer to develop graphs and other information products for use in dissemination. The staff
members assigned included district biostatistician, DTLS, HIV FP, and HMIS FP as summarized in Table
3.5. However, five districts, comprised of Kabale, Bundibugyo, Kalungu, Rakai, and Masaka had not
assigned any staff to lead this responsibility.

A specific staff member was assigned to lead the analysis of TB/HIV data in 49 districts, and partly so in
three. These included 38 of the 39 districts with some evidence of data use and 11 others with no
evidence of data use. In seven districts, the district biostatistician led the analysis, while the TB and HIV
FP led the interpretation. There was no lead person assigned at all in 10 districts: Bundibugyo, Butaleja,
Kabale, Kasese, Luwero, Masaka, Mubende, Nakasongola, Kalungu, and Rakai. Staff assigned to lead the
analysis are listed in Table 3.5.

Timely dissemination of analyzed TB/HIV data to stakeholders to inform program decisions was
conducted in 44 districts (75%), and partially so in five others. However, in seven districts (Kasese,
Kyegegwa, Mubende, Sironko, Soroti, Masaka, and Rakai), this was not done at all. The format of
dissemination included display of data tables and graphs in facilities, presentation at quarterly review
and planning meetings, and reports to health unit management (Table 3.5).

Table 3.5. Use of TB/HIV Data for Decision Making by Districts, August 2015
Component No. of Districts
Evidence of use of TB/HIV data in the district 29
Evidence of data use seen in districts 20
Graphs and charts on the wall 13
Forecasting and orders of TB/HIV commodities 1
Radio talk shows 1
Designing strategies such as TB camps, lake shore outreaches 4
Staff assigned to develop information products for dissemination
District biostatistician 25
DTLS 13
HMIS FP 8

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 31
 Table 3.5. Use of TB/HIV Data for Decision Making by Districts, August 2015
Component No. of Districts
HIV FP 1
Not indicated 1
Staff assigned to lead analysis of TB/HIV data
District biostatistician 39
DTLS 18
HMIS FP 8
HIV FP 9
Not indicated 2
Format of timely dissemination of TB/HIV data for use by stakeholders
Data tables in facilities 4
Presentations at quarterly review meetings 26
Presented to DHT beyond quarterly review meetings 15
Presented to IPs beyond quarterly review meetings 13
Other district meetings including DHMT, DAC, DEC, District Council, District
Technical Planning Meeting, HOD 15
Report to facility management beyond quarterly review meetings 8
Progress reports to CAO, DHO, etc. 5
Presented in regional pediatric conference 1
Recent programmatic decisions informed by analysis and dissemination of TB/HIV data
Quantification and forecasting of pharmaceutical needs 6
Informed HR allocation 8
Continuous quality improvement projects (CQI) 1
Targeted interventions for data quality, e.g., training, mentorship, and
supervision 2
Patient tracing including, MDR patients, Loss to Follow Up (LTFU), referrals 5
Strategies to improve TB case detection 5
Enhanced EQA 5
Identify hot spots 2
Lobby for accreditation of more facilities 2
Conducted community interventions including targeted HCT outreach, door
to door campaigns and outreaches, TB camp, program to support TB DOTS 5
Training, mentorship, and supervision 3
Improve gene expert utilization 1
In 47 districts, there were examples of program decisions taken that had been informed by the TB/HIV
data disseminated to stakeholders or other evidence of data use at the local level. These included
quantification, ordering, and forecasting of pharmaceutical and other medical supplies for TB/HIV
services, continuous quality improvement (CQI) projects in facilities and districts, and decisions on
allocation of human resources according to burden as summarized in Table 3.5. Some districts also
indicated that they prioritized utilization of scarce resources on support supervision of facilities with the
highest burden or problems such as LTFU or low cure rates. However, 10 districts (Adjumani, Bukedea,
Butaleja, Kanungu, Kasese, Kyegegwa, Mubende, Kalungu, Soroti, and Rakai) had no evidence of use of
TB/HIV data for decision making in their districts, and appear to be in need of urgent remediation. The
reasons underlying this situation were not however immediately clear and may need further
investigation.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 32
3.3 Objective 1b: Assessment of Facility M&E Systems
Summary Findings
42% of facilities were found to have adequate M&E systems.
58% of facilities had adequate M&E structures, functions, and capabilities to handle HIV/TB
information.
66% of facilities had adequate availability of data collection and reporting tools.
60% of facilities had adequate understanding of indicator definitions and reporting guidelines.
17% of facilities had adequate data management processes.
34% of facilities had adequate use of data for programming.

All five components of M&E

24% sites
system capacity were assessed rated
at the facility level. Overall, 42% sites Inadequate
42% of the facilities were rated (Score:
found to have adequate M&E Adequate <2.25)
systems as indicated in Figure (Score: 2.75 -
3.13. 3.00)

3.3.1 Status of Facility M&E 35% sites

Systems for TB/HIV rated
Services Acceptable
(Score: 2.25-
The status of facility M&E 2.74)
systems by component is
summarized in Figure 3.14.
Availability of data collection Figure 3.13. Overall Performance of All 106 Facilities for Five
tools and reporting forms was Components of M&E Systems Assessed, August 2015
rated adequate, while data use for decision making and data management processes were inadequate
and need urgent remediation for most facilities.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 33
 Availability of Data-collection Tools and Reporting 2.8
Forms for TB/HIV sevices

Understanding of Indicator Definitions and Reporting 2.7

Guidelines

M&E Structure, Functions and Capabilities to handle 2.7

HIV/TB information

Use of data for decision making 2.0

Data Management Processes 2.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Figure 3.14. Facility M&E Systems Assessment by Component, August 2015

Detailed findings under the respective components are presented in the following sections.

3.3.2 Function and Capabilities of Facility M&E Structure for HIV/TB Information
The overall status of the facility M&E structures, capabilities, and functions for managing TB/HIV
information is shown in Figure 3.15. Overall, approximately 61 (58%) of the facilities had adequate M&E
structures, functions, and
capabilities to manage TB/HIV data, 5% sites
rated
while another 40 (35%) were rated
Inadequate
acceptable. Facilities performed (Score:
fairly well on the different <2.25)
requirements of this component,
with almost 80% meeting various
requirements, except for having 38% sites
staff trained in collecting TB/HIV rated
data which was the case in just Acceptable
58% sites (Score: 2.25-
about three-quarters of the
rated 2.74)
facilities (Figure 3.16).
Adequate
(Score: 2.75
Nearly 85% of facilities had - 3.00)
sufficient staff assigned
responsibility for recording TB/HIV
services data. However, in eight
facilities (Butezi HC III (Sironko Figure 3.15. Overall Rating of Facility M&E Structures,
district), Gweri HC III (Soroti Functions, and Capabilities for Management of TB/HIV Data,
district), Kashumba HC III (Isingiro August 2015
district), Kidongole HC III (Bukedea
district), Kyere HC II (Serere district), Magale HC IV (Manafwa district), Mukwaya General Hospital
(Makindye division), and Seeta Nazigo HC III (Mukono district)) there was no staff assigned the
responsibility of recording TB/HIV data. Table 3.6 summarizes the types of staff handling the
responsibility at facilities where the role was assigned.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 34
Just under 80% of facilities Table 3.6. Staff Compiling and Reviewing Registers and Reports
indicated that they had a team before Submission, August 2015
approach to data management and No. of Facilities
reporting so that even in absence (N=106)
of some staff, data management Staff n (%)
was not affected; at 14 facilities Staff that compile registers and reports
this was partly the case. However, HIV care/ART register
in nine facilities, notably HMIS officer/medical records assistant/data
Bamugolodde HC III in Nakasogola clerk/data officer 14 (13%)
district, Kirima HC III in Kanungu, HIV FP 12 (11%)
Bweyogerere HC III and Kasangati Clinical staff 5 (5%)
HC IV in Wakiso, Kitayunjwa HC III In-charge 4 (4%)
in Kamuli, Kyangwali HC IV in TB register
Hoima, Mpambwa HC III in Jinja TB FP 39 (37%)
Paramedical staff (dispenser, laboratory technician,
district, Nyakibale Hospital in
lab assistant) 6 (6%)
Rukungiri, and St. Joseph Hospital
Clinical staff
Kitovu in Masaka district, there
Medical records assistant/data officer 4 (4%)
appeared to be no systematic Clinical staff 3 (3%)
mechanism by which data In-charge 3 (3%)
management duties were shared, Other (sub-county health assistant/Porter, TB FP) 2 (2%)
i.e., a team approach. Therefore, in Staff that review the reports before submission
the absence of key staff, there may Facility in-charge or TB/HIV clinic in-charge 70 (66%)
be no staff to handle data Team work 13 (12%)
management and reporting. It was Review by non-site staff (PEPFAR IP, biostatistician,
remarkable that in Buteza HC III in at monthly data review meetings) 12 (11%)
Sironko district, a porter is the one No reviews are made 8 (8%)
who fills in the TB register. Feedback on quality of reports received
District review meetings (performance/data
Four out of five facilities had a staff review), TB Regional meetings (NTLP) 53 (50%)
delegated to review reports prior District (Biostatistician/DHO/DTLS/HSD) or IP 40 (38%)
to submission to the next level, a No regular feedback 10 (10%)
measure necessary to reduce reporting errors. The staff that mainly reviewed the data are shown in
Table 3.6. In two-thirds of the facilities, the facility in-charge or TB/HIV clinic in-charge reviewed the
reports ahead of submission to the next level. However, eight facilities did not have anyone designated
for this responsibility, and in 16 facilities, this was done partly or on ad hoc basis.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 35
 Yes completely Partly Not at all NA

Staff collecting TB/HIV data trained

Receive regular support supervision

Regular Feedback on Reports

Designated staff to review reports

Team Approach to Report Submission

Desginated staff to record TB/HIV data

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.16. Facility M&E Structures, Functions, and Capabilities for Management of TB/HIV Data,
August 2015

A total of 84 out of 106 facilities indicated that they received regular feedback on the quality of reports
that they submit to the district and national level. However, six facilities did not receive any feedback,
while 16 did so partly or on ad hoc basis. Feedback is mainly provided during quarterly review meetings
and from members of the DHT (Table 3.6). Telephone calls are also made in cases of discrepancies.

All facilities except seven reported that they receive regular support supervision from the national or
district level or other relevant organization, based on standard supervision guidelines. Most (85 or 80%)
of the facilities reported having had a support supervisory within 2 months of the DQA. This finding was
consistent with those in the district assessment. However, 6% of the facilities reported having had a
support supervisory visit 3–4 months before the DQA, and a similar proportion 5–8 months before the
DQA.

The majority (95%) of the facilities reported that all or some of the staff responsible for collecting and
reporting TB/HIV data were trained/oriented in data collection and reporting. However, only 17% of
facilities reported having a plan for orienting their new staff in data collection and reporting for HIV/TB
services.

As shown in Figure 3.17, the capabilities and functions of the facility for meeting the requirements for
effective management of TB/HIV data was directly related to the hierarchical level of health facilities. In
fact, hospitals were almost twice as likely to meet the requirements relative to the lower level facilities,
especially HC IIIs.

3.3.3 Understanding of Indicator Definitions and Reporting Guidelines

Understanding indicator definitions and reporting guidelines is key in ensuring quality and timely data
being reported. Among the facilities assessed under this component, almost two-thirds (64 or 60%)
were rated adequate, while five (5%) were inadequate and will need immediate remediation (Figure
3.18).

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 36
 Among the requirements for this
60%
component, facilities performed very well
50% in understanding where reports should be
submitted and reporting timelines; what
40%
should be recorded in the source
30% documents, i.e., registers and patient
charts; having standard guidelines within
20% facilities on how data should be recorded
10% in the source documents; how to
communicate and effect changes to
0% reports previously submitted to the next
Hospitals HC IV HC III All Facilities
level; and what should be included in the
quarterly reports (HMIS 106a).
Figure 3.17. Proportion of Facilities Meeting All Understanding of how to derive indicator
Requirements of Capabilities and Functions for TB/HIV variables was somewhat weak (Figure
Data Management, August 2015 3.19).

The overwhelming majority of facilities

had good understanding of what to record in source documents, but 13 facilities only had partial
understanding of this requirement. Fortunately, no facility indicated that they had no such
understanding.

Over 70% of facilities had good understanding of what variables should be included in quarterly reports,
while 22 had only partial understanding. It was notable, however, that three facilities (Ciforo HC III in
Adjumani district, Makerere Hospital in Kampala Central Division, and Mubende Rehabilitation HC III in
Mubende) did not have adequate understanding of the variables to be included in the quarterly report.

Whereas an overwhelming majority of facilities (100) had complete understanding of where to submit
their quarterly HMIS 106a
reports, four facilities,
5% sites
including Soroti Regional
rated
Inadequate Referral Hospital, only had
(Score: partial understanding. Two
<2.25) facilities, namely Makerere
Hospital and Mubende
Rehabilitation HC III, did not
have any understanding of
35% sites where the reports should be
60% sites rated
rated
submitted.
Acceptable
Adequate (Score: 2.25-
(Score: 2.75 2.74)
- 3.00)

Figure 3.18. Overall Rating of Understanding of Reporting

Guidelines and Indicator Definitions among Facilities Assessed,
August 2015

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 37
 Yes completely Partly Not at all NA

Understanding how to derive variables for indicator "% HIV…

Understanding how to derive variables for indicator "% TB…

Understanding how to derive variables for indicator "PLHIV…

Are written instructions adequate for recording and reporting

How to effect changes in previously submited report

When reports are due

Understanding to whom report should be submitted

Understanding what should be included in quarterly report

Understanding what to record in source documents

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.19. Understanding of Reporting Guidelines and Indicator Definitions among Facilities
Assessed, August 2015

Almost all facilities (93) had good understanding of reporting timelines for the quarterly report.
However, nine facilities only had partial understanding and four, namely Soroti Regional Referral
Hospital, Mubende Rehabilitation HC III, Mpabwa HC III in Jinja district, and Makerere University
Hospital, did not have any understanding of the reporting timelines. Soroti Hospital adhered to
guidelines from the IP, while Makerere University Hospital just sent the reports the next month.

Table 3.7. Modalities of Communication of Changes to Previously Submitted Reports, August 2015
Communication Mode No. of Facilities
Phone communication to biostatistician 32
Physically travel to district biostatistician to effect the changes 10
Adjust and resubmit the report (modality of resubmission not indicated) 9
Direct edit in DHIS 2 6
Phone communication to IP 5
Corrected during monthly review meeting 4
Communicate correction to district or HSD (modality of communication
not specified) 21

Only 80 facilities had complete understanding of how to effect changes to already submitted reports
should the need arise. The main modalities through which this was done are summarized in Table 3.7.
However, it should be noted that the majority did so through a phone communication to the district
biostatistician. The remaining facilities were divided roughly equally into either having no or partial
understanding of how to effect the requisite changes. Reservations about making adjustments in the
DHIS 2 were raised as this would impact on the date of submission as illustrated in this quote from

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 38
Kityerera HC IV: “Updates the DHIS 2 however updating in the DHIS 2 affects the date of report
submission in DHIS 2 and this has affected correcting of errors once detected.”

In four instances, physically delivering the corrected report to the district biostatistician was preceded
by a telephone call to communicate the correction. All the five instances of communicating with or
informing the IP were preceded by communication to the district. In Kasangati and Kitebi HC IVs, even
after making direct edits in DHIS 2, the corrections were submitted to the district so that they could
adjust the hard copy. Among the facilities that indicated that they knew how to communicate and effect
changes to an already submitted report, 10 had never corrected any such error.

About 80 percent of the facilities felt that the current written guidelines were adequate to guide
accurate recording and reporting of TB/HIV data. However, of the remainder, two-thirds felt they were
partially adequate and the remaining third felt that they were not. Some staff indicated:

 “They are adequate but HMIS Manual does not have provisions for indicator definitions.” –
Apopong HC III in Pallisa district
 “For HMIS 106a report there are no instructions to define the indicators.” – Kashumba HC III in
Isingiro district
 “HMIS 106a TB indicators and HIV indicators do not have a written guide.” – Nyamarebe HC III in
Ibanda district
The main weaknesses were found in understanding of TB/HIV indicators, with just over two-thirds of
facilities indicating that they knew how to derive the variables for the indicator on the “percentage of
HIV-positive adults and children screened for TB in clinical settings.” Of the remainder, 11 facilities had
no understanding of how to derive the indicator at all. Several facilities felt that the HMIS manual does
not provide good guidance on TB/HIV indicators.

Nearly 32% of facilities either had no or just partial understanding of how to derive the variables for the
indicator for “percentage of new or relapsed TB cases with documented HIV serostatus.” Among them,
six facilities had no understanding of the
variables necessary to derive the
Met all requirements
indicator. Partly met 1 or more requirements
60% Did not at all meet 1 or more requirements
At the same time, almost one-third had
50%
no or just partial understanding of how
to derive the variables for the indicator 40%
on “percentage of HIV-positive TB cases
30%
that started or continued ART while still
on TB treatment,” with nine facilities 20%
having no understanding of the indicator 10%
at all.
0%
To get a better understanding of the Hospitals HC IV HC III All Facilities
variability across facilities, the analysis
was also disaggregated by facility level. Figure 3.20. Proportion of Facilities Meeting All
On the whole, hospitals performed Requirements for Understanding of Reporting Guidelines
better than facilities at lower levels, and Indicator Definitions, August 2015
though still almost 30% of the hospitals did not meet any of the requirements for this component
(Figure 3.20).

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 39
3.3.4 Availability of Data Collection and Reporting Tools
Facilities performed well on the 5% sites
component on availability of data rated
collection and reporting tools as Inadequate
shown in Figure 3.21. Almost two- (Score:
thirds (62% or 66) of the facilities <2.25)
had adequate HMIS data collection 33% sites
and reporting tools in stock and rated
were consistently using them, the Acceptable
primary data collection tools/ (Score:
registers had all the relevant 2.25-2.74)
variables for compiling the HMIS 62% sites
106a report, and, where available, rated
multiple organizations supporting Adequate
TB/HIV services in facilities use (Score: 2.75
- 3.00)
national data collection and
reporting tools.
Figure 3.21. Overall Rating of Availability of TB/HIV Data
Figure 3.22 shows the performance Collection and Reporting Tools among Facilities, August 2015
of health facilities on the various
requirements of this component. Only 16 facilities indicated partial availability of national HMIS forms
for reporting of TB/HIV data, while 90 had the tools available.

Yes completely Partly Not at all NA

Primary data collection tools/registers have relevant

variables to compile HMIS 106a reports
Facility has regular refill program for TB/HIV tools when
they are used up
Sufficient stocks of blank data collection tools/registers
and HMIS forms available
Multiple organizations implementing TB/HIV services at
the site use national reporting forms and timelines

Standard forms consistently used at the facility

Facility has national HMIS forms to be used for reporting

0% 20% 40% 60% 80% 100%

Figure 3.22. Availability of TB/HIV Data Collection and Reporting Tools At Facilities, August 2015
Sufficient stocks of blank primary data collection tools/registers and summary HMIS forms were
available in approximately 66% of the facilities, and partially so in 26% of facilities.

The proportion of facilities reporting complete stock outs of TB/HIV care tools, insufficient stocks of
blank tools, and facilities with old versions of tools are summarized in Table 3.8. Use of old versions of
tools affected all facilities including those indicating availability of tools.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 40
Over 80% of facilities consistently used the standard HMIS forms for routine reporting to the next level,
while 11% of facilities Table 3.8. Reported Availability of TB/HIV Tools by Type, August 2015
partially used standard
reporting forms. None No Blank Old Versions
Inconsistent use of some Available Copies/Few Cards of Tools
tools resulted from lack Tools n (%) n (%) n (%)
of these tools as HIV care/ART cards 4 (3.8%) 17 (16.0%) 15 (14.2%)
indicated in Table 3.8 Pre-ART register 3 (2.8%) 18 (17.0%) 20 (18.9%)
and was cited in ART register 3 (2.8%) 21 (219.8%) 17 (16.0%)
Budumba HC III, TB cards 6 (5.7%) 14 (13.2%) 10 (9.4%)
Bweyogerere HC III, Unit TB register 3 (2.8%) 15 (14.2%) 17 (16.0%)
Kibaale HC IV, and Presumptive TB register 1 (0.9%) 11 (10.4%) 13 (12.3%)
Pakadha HC III. In such HMIS 106a summary
instances, facilities tool 2 (1.9%) 13 (12.3%) 15 (14.2%)
resorted to use of black books, exercise books, or photocopying. In Kirima HC III, lack of knowledge was
the reason for the inconsistent use of the tools. However, five facilities that were not consistently using
the standard forms included Apopong HC III, Mubende Rehabilitation HC III, Buteza HC III, Mpabwa HC
III, and St. Luke Namaliga HC III. In Mbabwa HC III and Mubende Rehabilitation HC III, this was a result of
lack of knowledge on how the tools should be filled. The reasons why standard tools were not
consistently used are summarized in Table 3.9.

Where there are multiple organizations

implementing TB/HIV services in facilities, Table 3.9. Detailed Findings on Availability of Data
nearly all use the national reporting forms Collection and Reporting Tools, August 2015
and follow the same reporting timelines, No. of
except six facilities where this is partially Reason/Modality Facilities
done, and two facilities where it is not Reasons standard HMIS tools are not consistently used
done at all (Kyangwali refugee Lack of tools 4
resettlement HC IV and Makerere Knowledge gap of how to fill 3
University Hospital which follows IP Availability of alternative tool (Excel, EMRS) 2
guidelines that were different from Reason for inconsistent update not given 7
national guidelines). Modality of refill TB/HIV tools
Rely on DHO’s office and IP 35
A regular refill program of TB/HIV tools Rely on IPs 32
when they are used up was in place in 82 Rely on DHO’s office 20
facilities, and partially so in 16 others. One Rely on MoH supplies and IP 3
of the facilities that indicated partial Diocese 1
availability of a refill program was
Modality of refill not indicated 4
photocopying them locally, while in two
facilities, there was no regular refill program at all. The modalities of refilling TB/HIV tools are shown in
Table 3.9. Only eight facilities had no such refill program. Where there was no refill program, it was
mainly due to the fact that such facilities relied on ad hoc supplies from NMS or DHO.

The primary data collection tools/registers used in facilities had all the relevant questions or variables
needed to compile the HMIS 106a reports in an overwhelming majority of facilities (97), and partially so
in six others. However, three other facilities, namely, Buteza HC III in Sironko district, Mpambwa HC III in
Jinja district, and Nkondo HC III in Buyende district, felt that the data collection tools did not have all the
relevant variables needed to compile the report. Indeed, in these facilities, the report was compiled
from the HIV care/ART card in Buteza HC III, and a non-facility staff member facilitated by the IP

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 41
prepared the report in Nkondo HC. Where the primary data collection tools were found not to have all
relevant variables, it was partly due to torn
or unfilled registers, or that staff were not
trained.
70%
60% Figure 3.23 shows the proportion of
50% facilities by level with adequate (“yes
40% completely”) responses to the various
30% elements assessed under availability of
20% data collection and reporting tools. Again,
the availability of HMIS tools in accordance
10%
with these requirements was twice as
0%
likely in hospitals as compared to lower
Hospitals HC IV HC III All Facilities
level facilities (HC IVs and HC IIIs).

Figure 3.23. Proportion of Facilities Meeting all 3.3.5 Data Management Processes in
Requirements for Availability of Data Collection and Health Facilities
Reporting Tools, August 2015
Just like at the district level, the area of
data management processes manifested
most weaknesses in the assessment of facility M&E systems, where only 18 facilities (17%) had
adequate data management processes as shown in Figure 3.24.

The status of the various requirements of this component are shown in Figure 3.25. Areas of strengths
comprised timely submission of the HMIS 106a report and use of unique identifying number in patient
recording and reporting system to track patients across services in the facility.

The majority of facilities (85%) had an established patient recording and reporting system that allowed
tracking of unique individuals within and across service delivery points to avoid double counting at the
site. However, among the 15 facilities with inadequate patient tracking, disorganized numbering with
clients sharing client numbers, missing client IDs, manual records in high-volume hospitals, and clients
not having unique IDs at all, were some
of the issues noted that could
17% sites
potentially lead to double counting or 34% sites
rated
under counting clients. Adequate rated
(Score: 2.75 Inadequate
Just over 60% of facilities had quality - 3.00) (Score:
control procedures in place to assure <2.25)
quality of the HMIS 106a reports
submitted. Several methods used for
ensuring data quality were cited,
including peer review of completed
reports before submission and 49% sites
ensuring registers/data sources are rated
completed before report compilation, Acceptable
(Score:
site trend analysis, QI team reviews,
2.25-2.74)
and data triangulation.
Figure 3.24. Overall Rating of Data Management Processes
Among Facilities, August 2015

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 42
 Yes completely Partly Not at all NA

HMIS reports submitted timely

Data archiving/ storage system at site adequate

Staff aware and using SOPs/ Job Aids/ Guidelines on data

management
SOPs/ Guidelines on how documents should be handled/
archived are accessible to all staff
Confidentiality of patient information maintained according to
national guidelines
Computerised sites have quality controls for data entry from
paper forms to computer

Facility has computerized TB/HIV data and reports

Quality controls in place for compilation of accurate HMIS

106a reports
Patient recording/ reporting system tracks unique individuals
across service points to avoid double counting

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.25. TB/HIV Data Management Processes Among Facilities, August 2015

While computerized data management alleviates paperwork burden, just under one-third of facilities
had computerized TB/HIV data and reports. These included 14 of the 20 hospitals assessed, six of 28 HC
IVs, and only seven of 58 HC IIIs. At 63 facilities, TB/HIV data were not computerized at all. However, in
23 facilities, this was partially the case, with HIV data mostly computerized under the Open MRS system.

Among facilities with fully or partially computerized TB/HIV data, 25 had quality control procedures in
place for entry of data from paper forms to computer databases to minimize key stroke and other data
entry errors, while 10 others had them partially or on ad hoc basis.

Just over 75% of facilities had procedures in place to safeguard patient confidentiality in their reporting
system that were consistent with national guidelines. Among these facilities, only 62% used these job
aids effectively. Twelve facilities did not have any such procedures, while 15 others were only partially
compliant.

Guidelines or SOPs on how to handle and archive primary data source documents such as registers were
in place in just over one-quarter (28%) of facilities, and partially so in 11 other health facilities. However,
62 facilities did not have any such procedures at all.

Lack of clearly written and accessible SOPs, job aides, or guidelines that described how site documents
should be handled and archived, affected all levels with more than half of the facilities at all levels not
meeting this requirement.

Facility staff were aware of and effectively using the above mentioned SOPs, job aids, and guidelines in
only 24 facilities and partially so in another 15 facilities. In the remaining 66 facilities, the guidelines
were either not available or staff were not aware of and using them.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 43
 Although standard
16.00% guidelines on archiving and
14.00%
handling documents were
lacking in most facilities, a
12.00%
higher proportion of facilities
10.00% (50%) had fairly adequate
8.00% data archiving and storage
6.00% systems in place.
4.00%
Over three-quarters of
2.00%
facilities submitted their
0.00% quarterly reports of TB/HIV
Hospitals HC IV HC III All Facilities
services on time, and 14
facilities were partially doing
so. The 10 facilities that
Figure 3.26. Proportion of Facilities Meeting All Requirements for Data didn’t submit their reports
Management Processes, August 2015 on time were Alenga HC III,
Gweri HC III, Kataraka HC IV,
Kigarale HC III, Kinoni HC IV, Namwiwa HC III, Nganda HC III, St. Joseph Hospital Kitovu, Wekomire HC III,
and Makerere Hospital.

Figure 3.26 shows the proportion of facilities by level with adequate “yes completely” responses to all
the various requirements assessed under the data management component of the M&E system. Again,
hospitals performed a lot better than the lower level facilities. The performance of the various facility
levels by requirement under this component is summarized in Table 3.10. The requirements for which
hospitals performed better than the lower level facilities comprised timely submission of the quarterly
HMIS 106a reports, archiving of documents, confidentiality of patient records, and quality controls in
compilation of reports.

Table 3.10. Number of Facilities, by Level, with “Yes Completely” Responses to Requirements
Assessed under Data Management Processes, August 2015
Hospital HC IV HC III
n (%) n (%) n (%)
Data Management Element Total = 20 Total = 28 Total = 58
Patient recording and reporting system allows tracking of 16 (80.0%) 27 (96.4%) 47 (81.0%)
unique individuals within and across service delivery points
Quality controls in place for ensuring compilation of 16 (80.0%) 20 (71.4%) 30 (51.7%)
accurate quarterly (HMIS 106a) reports
Facility has computerized TB/HIV data and reports 14 (70.0%) 8 (28.6%) 7 (12.1%)
Patient data is maintained according to national 16 (80.0%) 20 (71.4%) 43 (74.1%)
confidentiality guidelines
Site has SOPs/job aids/guidelines that describe how 5 (25.0%) 8 (28.6%) 17 (29.3%)
documents should be handled and archived are clearly
written and accessible to all staff
Site has adequate data archiving/storage system 13 (65.0%) 16 (57.1%) 28 (48.3%)

HMIS 106a report not submitted on time 16 (80.0%) 21 (75.0%) 42 (72.4%)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 44
The four facilities that met all the requirements for data management processes were Laropi HC III,
Nyenga Hospital, Rubaga Hospital, and Ruharo Mission. On the other hand, Budumba HC III and Ngando
HC III did not score adequately on any of the components assessed under data management processes.

3.3.6 Use of Data for Decision Making in Health Facilities

Health facilities performed sub-optimally on use of data for decision making with only about one-third
(34%) or 36 of the facilities found to be proficiently using their data for decision making. More than half
of the facilities (57 or 54%)
were inadequately using
their data for decision
making and are need of 34% sites
immediate remediation. This rated
is summarized in Figure 3.27. Adequate
(Score: 2.75 -
The individual requirements 3.00)
under this component
54% sites
revealed that about 40–50%
rated
of facilities had physical
Inadequate
evidence of use of TB/HIV (Score:
service data in their facilities, <2.25)
41 facilities had no evidence 12% sites
at all, and 14 only partly had rated
such evidence as shown in Acceptable
Figure 3.28. The evidence (Score: 2.25-
took various forms, such as 2.74)
CQI activity, wall tables and
charts of trend graphs Figure 3.27. Overall Rating of Data Use for Decision Making in
summarizing the data, and Facilities, August 2015
use of data for TB/HIV
commodity ordering.

About 47% of facilities had staff providing TB/HIV services sufficiently trained in data analysis and
interpretation, while 24 other facilities had staff just partially trained. However, it is worrying that 32
facilities did not have staff trained to analyze and interpret TB/HIV data. Among facilities that reported
having trained staff, the staff included the in-charge of the facility, the records officer, data entry clerk,
or HMIS FP.

It was amazing to note that even among facilities with staff trained in data analysis and use, one-quarter
(25%) of them had no evidence of data use. At the same time, 26% of the 32 sites that did not have
anyone trained in data analysis had evidence of data analysis and use. Similarly, among the 32 facilities
with no one trained in data analysis, 16% were using their TB/HIV data for making program decisions,
while 31% of the 75 sites that reported having someone trained in analysis and interpretation of data
were not using it for making such decisions beyond the routine MoH reports.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 45
 Yes completely Partly Not at all NA

There
There areare programmatic
programatic decisions
decisions takentaken by facility
by facility based
on based on analyzed
analysed TB/HIV data results
TB/HIV dataresults

Analysed data disseminated to other stakeholders in a

timely manner to inform decisions

There are staff at the site who takes lead in analysis and
interpretation of TB/HIV data

TB/HIV services staff have been trained in data analysis

and interpretation

There is evidence of TB/HIV data use at site

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 3.28. Use of Data for Decision Making in Facilities, August 2015

Almost two-thirds (64%) of the

Table 3.11. Staff that Led Data Analysis and Interpretation,
facilities had a staff member who
August 2015
took the lead (completely or
partially) in analysis and Staff No. (%) of Facilities
interpretation of TB/HIV data. Health Information Assistant/Medical
Facility staff leading the analysis are Records Assistant/M&E Officer 40 (38%)
summarized in Table 3.11. In two Team Approach 8 (8%)
facilities, however, these were non- In-charge 6 (6%)
facility staff and included the DTLS TB Focal Person 2 (2%)
and HMIS FP. Clinicians were less External support (Person from District
likely to take the lead in the analysis (DTLS /HMIS FPs) 2 (2%)
and interpretation of TB/HIV data, Clinicians 2 (2%)
as this was the case in only two
facilities. It was noteworthy that in
all instances where data analysis was conducted by the team, the data were used for programmatic
decisions.

Evidence of program decisions taken by facilities based on analyzed TB/HIV data or results was seen in
46 facilities, and was also partially available in 11 others. Program decisions made by service delivery
staff based on analyzed TB/HIV data or results included following up on LTFU patients, planning
outreach services, setting targets, reorganizing clinic services to ease burden on clients and/or staff
workload, MDR TB client identification and tracking, and management and ordering of TB/HIV
commodities and supplies as well as their procurement.

Figure 3.29 shows the proportion of facilities with adequate (“yes completely”) responses to the various
requirements assessed under the data use component of the M&E system assessment.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 46
Again, like in most of the other components, hospitals were almost twice as likely to be proficient or
compliant on the various requirements compared to lower level health facilities. Data use for decision
making at facility level was particularly weak in HC III facilities.

60.00%

50.00%

40.00%

30.00%

20.00%

10.00%

0.00%
Hospitals HC IV HC III All Facilities

Figure 3.29. Proportion of Facilities, by Level, Meeting All

Requirements for Use of TB/HIV Services Data for
Decision Making, August 2015

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 47
3.4 Objective 2: Verification of Sources of Reported TB/HIV Data
Summary Findings
The majority of facilities had the required data sources for obtaining data.
24% of facilities had incomplete data in the ART and pre-ART registers.
10% of facilities had incomplete data in the TB register.
90% of the required information for the DQA was available in the data sources.

3.4.1 Availability of Data Sources

The standard data sources (pre-ART, ART, and Unit TB paper registers or electronic medical records
systems) for the variables needed to generate the three indicators assessed were found to be available
in almost all the facilities. In fact, data sources for TB variables were available in all facilities. However, in
one facility (Nyondo HC III in Mbale district) there was no ART register. In another facility (Kirima HC III in
Kanungu District) there was no pre-ART register. In these two facilities, the number of clients assessed
for TB and those on pre-ART and ART care could not be validated. Before field teams left these facilities,
they ensured that the registers were obtained from the DHO’s offices or the supporting IP was notified.

3.4.2 Completeness of Primary Data Sources

Whereas the data collection tools were available in most facilities, the level of completeness was less
optimal, especially for HIV/AIDS tools (Table 3.12). In approximately 30% of facilities, data in the Pre-ART
registers were not complete, i.e., essential data fields were not filled out. For ART registers,
approximately 26% had incomplete data.

Table 3.12. Completeness of Data Sources for the DQA in Facilities, August 2015
No. of Facilities Percent
Completeness of Data Sources (N=106) (%)
Data sources were complete (i.e., essential data filled out) for
No. PLHIV screened for TB 76 72%
No. HIV+ adults and children in pre-ART care 73 69%
No. HIV+ adults and children in ART care 79 74%
No. new and relapsed TB cases with documented HIV status 97 92%
Total no. registered new and relapsed TB cases 98 92%
No. HIV+ TB cases that started or continue ART 97 92%
No. registered TB cases with documented HIV-positive status 99 93%
Information was available covering the period under review for
No. PLHIV screened for TB 96 91%
No. HIV+ adults and children in pre-ART care 95 90%
No. HIV+ adults and children in ART care 100 94%
No. new and relapsed TB cases with documented HIV status 101 95%
Total no. registered new and relapsed TB cases 101 95%
No. HIV+ TB cases that started or continue ART 101 95%
No. registered TB cases with documented HIV-positive status 101 95%

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 48
Where the standard data sources were not available, in some instances, alternatives such as patient
chronic HIV/AIDS care/ART cards were used by the field teams during the DQA.

Lack of tools or, where available, inconsistent updating of them, were the main reasons cited for
incomplete TB/HIV data sources (Table 3.13). The ART and pre-ART registers were not regularly updated.
For the Unit TB register, inadequate recording of TB/HIV services included instances where HIV test
results were recorded without a record of HCT services, failure to record ART status of patients on
TB/HIV co-treatment, not indicating transfer-ins, etc. This problem was less common for health unit and
district TB registers, although the reasons for this were not immediately clear. There was no district TB
register in Mbale, while in Butambala district, the team failed to access the district TB register. In Zombo
the district TB register was not updated.

One consequence of these deficiencies cited by respondents during the DQA was under-reporting of
clients in care when their information was not updated (11 facilities). Another likely consequence of
inappropriate updating of the register could be over-reporting of numbers in instances where the
facilities did not indicate transfer-ins in the unit TB register (two facilities). In instances where there was
no age and sex disaggregation, this would constrain appropriate disaggregation of data reported to the
national level (cited in six facilities). On the other hand, where electronic medical records systems
existed and were up to date, lack of update of the manual copies was unlikely to affect reported
numbers (three facilities). However, such facilities needed to ensure that there was adequate regular
backup. Incomplete or lack of HMIS 106a reports at the facility, or District TB register at the district,
meant that Cross-checks 1 and 2 could not be performed.

Incomplete records
were found in facilities Table 3.13. Reasons for Incomplete TB/HIV Data Collection and Reporting
across all levels from Forms, August 2015
HC III to hospitals
Tools Not Available Tools Not Updated
(Table 3.14).
Tool N % N %
Availability of Data for Pre-ART register 1 0.9 33 31.1
the Period October ART Register 1 0.9 36 34.0
2014–March 2015 Unit TB register — — 11 10.4
Oct–Dec 2014 HMIS
Although 106a copy at site (TB
completeness of data section) 5 4.7 4 3.8
sources was sub- Jan-March 2014 HMIS
optimal for some 106a copy at site (HIV
variables, the section) 5 4.7 3 2.8
availability of data for Jan–March 2014 HMIS
the reporting period 106a copy at site (TB
under review (SAPR section) 5 4.7 2 1.9
2015) was better. District TB register 2 1.9 1 0.9
Information for the
seven variables assessed was available in at least 90% of the facilities, with data on TB variables and for
current ART enrollment almost universally available (>95% of facilities) as indicated in Table 3.14. The
better performance was due to the fact even for facilities where the data sources were incomplete as
cited above, the incomplete records did not cover the period for the SAPR 2015. It was not possible to
perform the DQA for Indicator 1, TB screening, in Nyondo HC III (variables 1, 2, and 3), Kalangala HC IV
(variables 1 and 2), and Kinoni HC III (variable 2) because of incomplete or lack of records.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 49
Table 3.14. Availability and Completeness of Data Sources across Levels of Facilities, August 2015
Variables Assessed for Availability and Completeness of Data Sources During the DQA
No. of No. HIV+ No. of
HIV+ adults No. of new Total no. of HIV+ TB No. registered
adults and and and relapsed registered cases that TB cases with
No. PLHIV children in children TB cases with new and started or documented
Facility Level screened Pre-ART in ART documented relapsed TB continue HIV-positive
(n) for TB care care HIV status cases ART status
Data sources were complete (i.e., essential data filled out)
Hospitals (20) 55.0% 55.0% 60.0% 85.0% 90.0% 85.0% 85.0%
HC IV (28) 75.0% 67.9% 78.6% 92.9% 96.4% 89.3% 92.9%
HC III (58) 75.9% 74.1% 77.6% 93.1% 93.1% 94.8% 96.6%
Information was available covering the period under review
Hospitals (20) 85.0% 85.0% 85.0% 90.0% 90.0% 90.0% 90.0%
HC IV (28) 89.3% 89.3% 96.4% 96.4% 96.4% 96.4% 96.4%
HC III (58) 93.1% 91.4% 96.6% 96.6% 96.6% 96.6% 94.8%

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 50
3.5 Objective 3: Validation of Reported TB/HIV Data
Summary Findings
For all seven variables assessed, no exact match between the manual recount and data reported in
DHIS 2 was found.
For two variables (the number of PLHIV screened for TB in clinical settings and number of new and
relapsed TB cases with documented HIV serostatus), the difference between manual recount and
DHIS 2 was within the adequate range of deviation (+5%).
For one variable (number of registered TB cases with documented HIV-positive status), the
difference was within the acceptable range of deviation (+10%).
Facilities performed better with reporting variables that are recorded and reported through the TB
information system compared to the HIV/AIDS chronic care information system.

The TB/HIV data reported in DHIS 2 for each facility were compared with the validated manual recount.
A further cross-check compared the data in the copy of the facility HMIS 106a report against the joint
manual recount. This was done for each variable. Additionally, a cross-check was done comparing facility
TB data with the district TB register for the TB variables (variables 4–7).

3.5.1 Characteristics of the DHIS 2 Data

The number of clients reported for the three indicators varied widely among the facilities. For instance,
the number of clients screened for TB in the facilities ranged from 2 to 7,234 (median 351) reported by
103 facilities, out of a range of 4 to 18,123 clients (median 471) that attended chronic HIV/AIDS care
services in these facilities during this period. At the same time, the number of new and relapsed TB
cases with documented HIV status during the reporting period ranged from 0 to 443 (median 10), out of
a range of 0 to 441 (median 13) registered new and relapsed TB cases in 105 facilities during the same
period. The number of registered TB cases with documented HIV-positive status who started or
continued TB treatment during this reporting period ranged from 0 to 194 (median 4), out of a range of
1 to 300 (median 5) registered TB cases with documented HIV-positive status, reported from 88 out of
the 106 facilities.

3.5.2 Overall Comparison of Totals

Table 3.14 summarizes, for the 106 facilities assessed, the deviation of the totals for each variable
derived from the joint manual recount from the DHIS 2 totals, and the deviation of the total from the
joint manual recount from the totals in the copy of HMIS 106a report at the facility. For each variable,
the deviation in the totals are summarized in Table 3.15.

In all 106 facilities, it was possible to compare the manual site recount against figures in the DHIS 2
database. However, for five facilities (Kalisizo Hospital, St. Joseph Kitovu hospital, Kinoni HC III, Nyondo
HC III, and Kucwiny HC III) there was no facility copy of HMIS 106a report at the site at all. For St. Joseph
Hospital, Kitgum, there was no January–March 2015 HMIS 106a report at the site, while for Masaka
Regional Referral Hospital (RRH), the was no TB section for both quarters. For Ciforo HC III, Metu HC III,
and Bugamba HC IV there was no October–December 2015 report. At these sites, the comparison of
manual recount against the facility copy of the HMIS 2 report was not possible and was therefore
excluded from the analysis for cross-check 1, and explains why the figures for DHIS 2 from the first
analysis differ from the second.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 51
Table 3.15. Percentage Deviation of Manual Recount Totals against DHIS 2 Facility Copy of HMIS 106a
Report Totals, August 2015
Total % Deviation % Deviation
from (DHIS 2 – Total from Total from (Facility Copy of
Joint Total from Joint Joint Facility Copy HMIS 106a –
Manual DHIS 2 Manual Manual of HMIS Joint Manual
Variable Recount Database Recount) Recount* 106a Report Recount)
No. PLHIV screened for TB 78,883 78,184 -0.9% 72,500 78,431 8.2%
No. of PLHIV in pre-ART care 9,542 11,721 22.8% 9,250 14,343 55.1%
No. PLHIV in ART care 75,173 87,779 16.8% 68,463 72,268 5.6%
No. new/relapsed TB cases with
documented HIV status 2,462 2,578 4.7% 1,796 1,914 6.6%
Total of registered new and
relapsed TB cases 2,498 2,927 17.2% 1,835 1,986 8.2%
No. of HIV+ TB cases that started
or continued ART 1,069 925 -13.5% 649 639 -1.5%
No. registered TB cases with
documented HIV+ status 1,284 1,184 -7.8% 785 773 -1.5%
*Total figures under this column differ from those under the column of joint manual recounts used for the comparison with
DHIS 2 figures because facilities that lacked the HMIS 106a copy were excluded.

Although for all seven

Table 3.16. Percentage Deviation of Manual Recount Totals from
variables there was no exact District TB Register, August 2015
match between the total joint
manual recount and the totals Total from % Deviation
from the DHIS 2 database, for Joint Total from (DHIS 2 - Joint
two of the variables, the Manual District TB Manual
Variable Recount* Register Recount)
deviation was within the
acceptable range of +5%. No. new/relapsed TB
These two variables were the cases with documented
number of PLHIV screened for HIV status 2,437 2,329 -4.4%
TB in clinical settings (78,184 Total of registered new
in DHIS 2 database versus and relapsed TB cases 2,473 2,372 -4.1%
78,883 obtained from the joint No. of HIV+ TB cases that
manual recount), and the started or continue ART 1,050 987 -6.0%
number of new and relapsed No. registered TB cases
TB cases with documented HIV with documented HIV+
serostatus (2,462 versus status 1,262 1,093 -13.4%
*Total figures under this column differ from those under the column of joint manual
2,578). For a third variable, the
recounts used for the comparison with DHIS 2 figures because in five districts there
number of registered TB cases was either no data in the district TB register or the team failed to access it; data from
with documented HIV-positive facilities in these districts was excluded.
status, the deviation was
within the moderate range of +10%. For the remaining variables, the deviations in the total exceeded
the +10% limit, with the deviation for one variable (the number of HIV-positive adults and children in
pre-ART care) being approximately 23%.

For the comparison of the joint manual recount totals against the totals derived from the facility copy of
the HMIS 106a report, again there was no variable where the totals matched exactly. However, for two

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 52
variables, the deviation was within the acceptable range (+5%) and, for four variables, the deviation was
within the moderate range of 10%. For one variable (the number of adults and children in pre-ART care),
the deviation was marked at 55% (Figure 3.30 and Table 3.14).

Table 3.16 summaries the deviation of the district TB register from the facility counts for four variables
that are also recorded in the district TB register. With the exception of the variable for the number of
registered TB cases with documented HIV status, the TB variables were within acceptable range.

Overall, facilities performed better at reporting variables that were recorded and reported through the
TB information system compared to the HIV/AIDS chronic care information system.

Table 3.16 summaries the deviation of the district TB register from the facility counts for four variables
that are also recorded in the district TB register. With the exception of the variable for number of
registered TB cases with documented HIV status, the TB variables were within acceptable range.

Overall, facilities performed better at reporting variables that were recorded and reported through the
TB information system compared to the HIV/AIDS chronic care information system.

Indicator 1: The proportion of HIV-positive clients in care that were screened for TB during January–
March 2015
For indicator 1, the proportion of HIV-positive clients in clinical settings that were screened for TB during
January–March 2015, there was only slight discrepancy between the DHIS 2 totals and the totals from
the joint manual recount for the numerator, i.e., the number of clients screened for TB: 78,184 versus
78,883, or a deviation of only –0.9%. The deviation of the manual recount from the totals in the facility
copy of the HMIS 106a report was, however, slightly more marked at 8.2%, Figure 3.31.

For variable 2, the total

number of clients in
pre-ART care, Deviation of DHIS-2 Totals againt Joint Manual Recount
established through Deviation
DeviationofofFacility
Facilitycopy
Copyifof
HMIS
HIMS106a Report
106a ReportTotals againt
Totals Joint
against Manual
Joint Recount
Manual Recount
joint manual recount 60.0%
from facility records
50.0%
(9,542) was over-
reported in the DHIS 2 40.0%
database by 22.8%.
However, when 30.0%
compared to the 20.0%
facility copy of the
HMIS 106a, the 10.0%
deviation was even
0.0%
more at 55.1% (Figure No. of PLHIV No of HIV +ve No. HIV +ve No. of new Total No of No. of HIV +ve No. registered
3.31). The reasons that -10.0% Screened for adults and adults and and relapsed registered TB cases that TB cases with
TB children in children in TB cases with new and started or documented
account for this over- Pre-ART care ART care documented relapsed TB continue ART HIV-positive
reporting in the DHIS 2 -20.0% HIV status cases status

as deduced from the

comments in the data Figure 3.30. Summary of Deviation of Totals from DHIS 2 and Facility Copy
collection tools are of HMIS 106a Report from Manual Recount, August 2015
summarized in Table 3.17 and are consistent with findings of the document review where 25% of the
pre-ART registers were incomplete.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 53
There was also no copy of the
HMIS 106a at the site in four DHIS-2 Manual Recount-1 HMIS 106a Copy Manual Recount -2
facilities and, in another three
100,000
facilities, the HIV section was 90,000
missing from the HMIS 106a 80,000
70,000
report.

No. of Clients
60,000
50,000
The total number of clients on 40,000
30,000
ART in these facilities as 20,000
established through joint 10,000
0
manual recount (75,173), was PLHIV Screened for Adults and Children Adults and Children
over-reported in the DHIS 2 TB in Pre-ART Care in ART Care
Variable / Indicator
database by 16.8% (87,779).
However, the deviation of
totals derived by joint manual
Figure 3.31. Comparison of Manual Recount Totals to DHIS 2 and
recount from the totals from
Facility Copy of HMIS 106a Report for Variables for TB/HIV Indicator
the facility copy of the HMIS
1, August 2015
106a report was slightly less,
5.6% (Figure 3.31). The reasons for this over-reporting in DHIS 2 as deduced from the comments in the
data collection tools are summarized in Table 3.18. The main reasons, consistent with previous findings,
were that the ART register was not being updated, which led to errors in determining the number of
active clients.

Table 3.17. Reasons for Disparity between Manual Recount and DHIS 2 and Facility Copy of
HMIS 106a Report for Number of Clients in Pre-ART Care, August 2015
Reason No. of Facilities
Registers not updated 25
Arithmetic error 13
Using alternate data sources for report 6
Report compiled by IP so facility staff could not explain the discrepancy 4
Lack of knowledge of what to report 4

Indicator 2: Percentage of registered new and relapsed TB cases with documented HIV status
For indicator 2, the proportion of registered new and relapsed TB patients that had documented HIV
status, the comparison of totals from manual recount and DHIS 2 and facility copy of the HMIS 106a
report are summarized in Figure 3.32.

Table 3.18. Reasons for Disparity between Manual Recount and DHIS 2 for Number of Clients in ART
Care, August, 2015
Reason No. of Facilities
Registers not updated 22
Arithmetic error 18
Lack of knowledge of what to count 6
Using alternate data sources for reporting 6
Report compiled by IP so facility staff could not explain the discrepancy 4

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 54
The total number of new or relapsed TB cases with documented HIV status in the DHIS 2 database was
within acceptable range of what was ascertained through manual recount, i.e., 2,578 versus 2,462 for a
deviation of 4.7%. For facilities that had a copy of the HMIS 106a report on site, the totals in the HMIS
106a report (1,914 cases) exceeded the joint manual count by 6.6% (1,796 cases), Figure 3.32.

DHIS-2 Manual Recount-1 HMIS 106a Copy Manual Recount -2

3,500

3,000
No. of Clients

2,500

2,000

1,500

1,000

500

0
No. of TB cases with Total No. of No. of HIV+ve TB No. of registered TB
documented HIV registered and cases started or cases with
status relapsed TB cases continued ART documented HIV +ve
status

TB/HIV Variable

Figure 3.32. Comparison of Manual Recount Totals to DHIS 2 and Facility

Copy of HMIS 106a Report for Four TB Variables, August 2015

The total number of new and relapsed TB cases (variable 5) was over-reported in DHIS 2 compared to
joint manual recount by 17.2% (2,927 versus 2,498). The total for this variable was also slightly over-
reported among facilities that had a facility copy of HMIS 106a at the site compared to the joint count
total by 8.2% (1,986 versus 1,835).

Indicator 3: Percentage of HIV-positive new and relapsed registered TB cases on ART during TB treatment
The total number of HIV-positive TB cases that started or continued ART during the reporting period was
under-reported in DHIS 2 compared to the joint manual recount by 13.5%, (925 versus 1,069 cases).
However, for facilities that had a copy of the HMIS 106a report on site, the manual recount and the
HMIS figures were close, with a
deviation of just -1.5% (649 versus 639). Table 3.19. Reasons for Disparity between Manual Recount
The reason for the discrepancy between and DHIS 2 for Number of HIV-positive TB Cases that
the manual recount and DHIS 2 figures Started or Continued ART During the Period for the SAPR
as recorded in the data collection tools 2015, August 2015
are summarized in Table 3.19. Reason No. of Facilities
Insufficient knowledge of what to report 26
The total registered TB cases with Arithmetic error 14
documented HIV-positive status was Registers not updated 3
also under-reported in DHIS 2 relative to Wrong data source 2
the joint manual recount by 7.8% (1,184
versus 1,284). However, relative to the facility copy of HMIS 106a, this was within the acceptable range
with a deviation of -1.5%).

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 55
 The disparity between totals for all variables derived from the joint manual recounts and the totals in
the DHIS 2 database, as well as totals in the facility copies of HMIS 106a where applicable, disaggregated
by facility level, are summarized in Table 3.20. There was very little variation by facility level on the
reported number of PLHIV that were screened for TB.

Health Centre III facilities over-reported the number of clients in pre-ART care. There also appears to
have been over-reporting of the number of clients on ART in DHIS 2 relative to manual recounts
especially in hospitals and HC III facilities. The number of new or relapsed TB cases with documented HIV
status was over-reported in DHIS 2 relative to the average in HC IIIs and HC IVs, but there was no
disparity at all among hospitals. The number of new and relapsed TB cases was also grossly over-
reported by HC III and IV facilities, while the number of HIV-positive TB cases that started or continued
ART during this period was grossly over-reported by HC IV facilities and under-reported by hospitals. The
total number of HIV-positive TB cases was grossly under-reported by hospitals, while HC III and HC IV
facilities grossly over-reported this variable.

Table 3.20. Deviation of Manual Recounts from Reported Totals by Level of Facility, August 2015
No. of new Total no. of No. of HIV+
No. of No. of and relapsed registered TB cases No. registered
PLHIV PLHIV in No. of TB cases with new and that started TB cases with
screened Pre-ART PLHIV in documented relapsed TB or continue documented
for TB care ART care HIV status cases ART HIV+ status
Disparity between DHIS 2 figures and joint manual count
Hospitals (20) -5.7% 3.2% 23.8% -3.3% 2.0% -25.9% -20.1%
HC IV (28) 1.6% 10.4% 7.3% 11.6% 23.8% 24.8% 22.9%
HC III (58) 5.4% 54.6% 20.3% 29.1% 72.5% -0.8% 16.6%
All Facilities -0.9% 22.8% 16.8% 4.7% 17.2% -13.5% -7.8%
Disparity between facility copy of HMIS 106a and joint manual count
Hospitals -0.8% -0.8% 2.6% -8.5% -6.3% -10.7% -17.4%
HC IV 15.0% 87.6% -0.2% 19.0% 9.8% 11.9% 18.0%
HC III 11.8% 54.4% 28.5% 27.9% 46.4% 1.8% 11.4%
All Facilities 8.2% 55.1% 5.6% 6.6% 8.2% -1.5% -1.5%
Disparity between district TB register and joint manual count
Hospitals N/A N/A N/A -2.1% -16.2% -13.3% -21.0%
HC IV N/A N/A N/A -8.3% 21.4% 10.4% 8.1%
HC III N/A N/A N/A -8.8% 9.7% 13.9% 0.7%
All Facilities N/A N/A N/A -4.4% -4.1% -6.0% -13.4%

3.5.3 Facility Level Analysis

Site level comparison of joint manual recount with total figures reported in DHIS 2, facility copy of HMIS,
and district TB register
At the site level, the deviation of the validated manual recount from the DHIS 2 figures, the facility copy
of HMIS 106a, and, for four TB variables, the TB register are summarized in Table 3.21 and Figure 3.33
which show the proportion of facilities with adequate (0% to +5%), acceptable (+5% to +10%), or gross
deviation (≥ +10%) for each of the seven variables.

For all seven variables, the majority of facilities (23 or 44%) grossly over-reported (deviation by 10%) in
the DHIS 2 database compared to validated figures at the facility. Gross under-reporting compared to

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 56
validated figures was also
common at facilities (18 or No or Acceptable Deviation 0 + 5% Acceptable Under-reporting

38%). The proportion of Acceptable Over-reporting Gross Under-reporting

facilities where the GrossOver-reporting
Gross Over-reporting
reported totals in the DHIS
100%
2 database were within
90% 23
the adequate range (0 to 31 32 34 35 32
80% 44
±5%) of the manually
70%
verified figure ranged from
60% 29 37
23–47%, depending on the 22 18
50% 38 35
variable. 7 4 24
40% 8 5
6 3
1
9
The number of PLHIV in 30% 4 8 3
1
4 4
clinical care settings that 20% 36 47 36
30 29
were screened for HIV was 10% 23 21
grossly over-reported by 0%
31% of facilities, under- No. of PLHIV No of HIV No. HIV +ve No. of new Total No of No. of HIV No.
Screened for +ve adults adults and and relapsed registered +ve TB cases registered
reported by 22%, and TB and children children in TB cases new and that started TB cases
reported adequately at in Pre-ART ART care with relapsed TB or continue with
30% of facilities. Health care documented cases ART documented
HIV status HIV-positive
facilities also grossly over- status
reported (32%) and under-
reported (38%) the
number of adults and Figure 3.33. Site Level Comparison of Validated Counts and Totals
children in pre-ART care, Reported in DHIS 2 for Seven Variables, August 2015
with only 23% of facilities reporting numbers within adequate range of the verified figures.

The number of clients actively enrolled on ART was correctly reported by 36% of facilities with almost a
similar proportion (34%) grossly over-reporting this number. At the same time, 18% of the facilities
grossly under-reported this number. Similarly, the number of new and relapsed TB cases with
documented HIV status was correctly reported by 47% of facilities, grossly over-reported by 35%, and
under-reported by 29%.

Nearly half of facilities (44%) grossly over-reported the total number of registered or relapsed TB cases,
with just over one-fifth (21%) reporting figures that were adequate. About 24% of facilities grossly
under-reported the number of new and relapsed TB cases.

The number of HIV-positive clients that started or continued ART, and its denominator, the number of
registered TB cases with documented HIV-positive status, were correctly reported by 36% and 29% of
the facilities, respectively. However, nearly all remaining facilities either grossly over-reported or under-
reported the figures for the two variables.

For all variables, the proportion of facilities that reported figures with acceptable deviation of +5% to
+10% was small. Overall, there was sub-optimal performance by facilities on all the seven TB/HIV
variables.

The performance of specific facilities against each variable is shown in Annex 6 (with a color coding
scheme highlighting the level of deviation) and summarized in Table 3.21.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 57
Table 3.21. Site Validated Data vs. DHIS 2, HMIS 106a Report at Site, and District TB Register, Overall Level of
Adequacy, All Indicators, August 2015
% Sites with % Sites with % Sites
% Sites Acceptable Acceptable % Sites with with Gross
with Adequate Under- Over- Gross Under- Over- % Sites
Results Reporting Reporting Reporting Reporting with No
(0 to 5%) (-5 to -10%) (5 to 10%) (< -10%) (>10%) Data
Joint manual count versus DHIS 2
No. PLHIV screened for TB 30.2 9.4 7.5 21.7 31.1 0
No. of HIV+ adults and children in
pre-ART care 22.6 3.8 3.8 37.7 32.1 0
No. HIV+ adults and children in ART
care 35.8 5.7 6.6 17.9 34.0 0
No. of new and relapsed TB cases
with documented HIV status 27.4 4.7 3.8 29.2 34.9 0
Total no. of registered new and
relapsed TB cases 20.8 3.8 7.5 23.6 44.3 0
No. of HIV+ TB cases that started or
continue ART 36.8 0.9 2.8 36.8 22.6 0
No. of HIV+ TB cases that started or
continue ART 29.2 0.9 2.8 34.9 32.1 0
Joint manual count versus facility copy of HMIS 106a
No PLHIV screened for TB 31.1 8.5 5.7 25.5 23.6 5.7
No of HIV+ adults and children in pre-
ART care 22.6 4.7 2.8 28.3 35.8 5.7
No. HIV+ adults and children in ART
care 38.7 2.8 6.6 14.2 32.1 5.7
No. of new and relapsed TB cases
with documented HIV status 25.5 2.8 1.9 23.6 40.6 5.7
Total no. of registered new and
relapsed TB cases 33.0 3.8 0.9 17.9 38.7 5.7
No. of HIV+ TB cases that started or
continue ART 37.7 0.0 3.8 26.4 26.4 5.7
No. of HIV+ TB cases that started or
continue ART 30.2 1.9 2.8 27.4 32.1 5.7
Joint manual count versus district TB register
No. of new and relapsed TB cases
with documented HIV status 47.2 5.7 2.8 23.6 17.0 3.8
Total no. of registered new and
relapsed TB cases 48.1 6.6 0.0 28.3 13.2 3.8
No. of HIV+ TB cases that started or
continue ART 51.9 2.8 0.0 27.4 14.2 3.8
No. of HIV+ TB cases that started or
continue ART 47.2 2.8 0.0 34.9 11.3 3.8
It is remarkable that there was not a single facility for which the figures for all seven variables reported
in the DHIS 2 database were within the acceptable margin of the validated counts (Table 3.22). Only one
facility, i.e., Laropi HC III, had adequate figures for six of the seven variables. In fact, for 23 facilities,
there was no variable for which the figures in DHIS 2 database were within the acceptable margin of
validated counts. At the other extreme, only two facilities (Laropi HC III and Kyazanga HC IV) had no
variable reported out of range of the validated count by 10% or more. However, for 23 facilities, all the
figures reported for the seven variables were out of range of the validated count.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 58
3.5.4 Site Level Comparison Table 3.22. Facilities with DHIS 2 Totals within Range of Validated
of Validated Counts versus Count, by Number of Variables, August 2015
Figures in the Facility Copy No. Facilities with DHIS 2 Figures within
of the HMIS 106a Report Number of Variables Range of Validated Count
The site level comparison of the Adequate Acceptable Inadequate
validated counts against figures All 7 variables 0 0 15
in the facility copy of HMIS 106a 6 variables 1 0 16
report is summarized in Figure 5 variables 3 1 17
3.34. Again, depending on the 4 variables 9 2 26
variable, between 24% and 41%
3 variables 14 4 16
of facilities grossly over-
reported (≥ 10%) the numbers 2 variables 29 15 10
in the HMIS 106a copy at the 1 variable 27 45 4
facility compared to the No variables 23 39 2
validated manual recount. Total 106 106 106
Fourteen to 28% of facilities
grossly under-reported values by 10% compared to validated figures. In the case of only 23–39% of
facilities were the figures in the facility copy of HMIS 106a report within acceptable margin of the
validated counts.

The number of No or Acceptable Deviation 0 + 5% Acceptable Under-reporting Acceptable Over-reporting

HIV-positive
Gross Under-reporting GrossOver-reporting
Gross Over-reporting
adults and
100%
children in ART
90% 24
care was the 80% 36 32
39
26
32
41
variable that was 70%
reported by most 60% 26 14 26
27
facilities within 50%
6 28 7 24
18
40% 3 4
0
1
the acceptable 9
3 2
4 3
2
30% 3
range of the 5
20% 39 38
31 33 30
validated count 10% 23 26

(39% of facilities), 0%
No. of PLHIV No of HIV +ve No. HIV +ve No. of new and Total No of No. of HIV +ve No. registered
while the number Screened for TB adults and adults and relapsed TB registered new TB cases that TB cases with
of HIV-positive children in Pre- children in ART cases with and relapsed TB started or documented
ART care care documented cases continue ART HIV-positive
adults and HIV status status
children in pre-
ART care was
Figure 3.34. Site Level Comparison of Validated Counts and Totals Reported in
reported within
Facility Copy of HMIS 106a Report for Seven Variables, August 2015
acceptable limits
of the validated counts by the lowest number of facilities (23%). The variable most over-reported by
facilities compared to the facility copy of the HMIS 106a report was the number of new and relapsed TB
cases with documented HIV status (over-reported by 41% of facilities), while the number of PLHIV in
clinical care settings screened for TB was over-reported by the fewest facilities (24%).

The number of adults and children in pre-ART care was the variable that was under-reported by the
most facilities (28%), and the number of adults and children in ART care was under-reported by the least
facilities (14%).

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 59
3.6 Objective 4: Challenges Faced in Collecting, Recording and Reporting Data
at Sites
In this section we present the main areas presenting threats to the M&E system and needing
improvement, as identified by health workers together with the DQA teams. Important to note is that
these issues affected all levels of health facilities.

1. Human resources issues were noted in a number of facilities. Aspects of human resources for M&E
in facilities included:

a) Inadequate technical capacity, which was the most prevalent, affecting 99 (93%) facilities of all
levels comprising all HC IIIs, all HC IVs, and 13 hospitals. This emerged in the form of knowledge
gaps about variables for deriving TB/HIV
indicators and reporting timelines, Ninety-nine facilities including all the HC IIIs (58) and
inability to generate reports from IVs (28) as well as more than half the hospitals (13)
manual or electronic databases, and had inadequate technical capacity to handle TB HIV
inability to analyze and report data. In data.
some facilities, this situation was
aggravated by non-site staff taking over the reporting function without necessarily empowering
the site staff to own the process, and incompetent mentors with inadequate knowledge and
skills misguiding the facility teams.

b) Human resource issues of leadership, though slightly less common, were noted as a major area
of weakness in 19 (18%) facilities. As a result of this deficiency, some facilities did not have any
staff assigned to certain M&E functions,
including data management, and report
19 facilities including seven HC IIIs, six HC IVs, and six
generation and dissemination. At the
hospitals had human resource management issues.
same time, lack of teamwork was noted
in a number of facilities with only a few
staff empowered to handle certain roles.

c) In a smaller number of facilities (five HC IIIs and one HC IV) there were limited human resources
for TB/HIV. These and other deficiencies in the systems resulted in poor or no data utilization
that was noted in almost one-third (32) of facilities. Late or no submission of reports also
emerged as areas in need of
immediate remediation in 10 facilities Sub-optimal data utilization disproportionately
(six HC IIIs, two HC IVs, and two affected HC IIIs as was highlighted by 23 (41%) of HC
referral hospitals). In a few of the IIIs, six (21%) of HC IVs, and two (10%) hospitals.
facilities (7), the lack of backup copies
of relevant summary reports at the site was also noted as a key area for improvement.

2. The area with most glaring weaknesses identified by the teams was that of Data Management. Areas
mostly affected comprised:

a) Irregular updating of registers in

Irregular updating of data collection tools was
almost half 47 (44%) of the facilities.
highlighted in 26 (45%) HC IIIs, 10 (36%) HC IVs and
This, as noted earlier, affected mainly
11 (55%) hospitals.
HIV/AIDS data management systems
but also TB registers, and ranged from

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 60
 complete non-updating to delayed updating and was the cause of gross deviations from the
actual figures across various reports. The irregular updating of tools was compounded by lack of
data quality controls including report validation before submission as well as lack of feedback
from next level as noted in over one-third (37 or 35%) of facilities These facilities comprised 22
(38%) HC IIIs, 10 (36%) HC IVs, and five (25%) hospitals.

a) At almost one-third of the facilities, lack of primary and secondary data collection tools was also
highlighted by facility staff as an area that needed remediation. In some facilities, this was
aggravated by poor quality registers that were easily torn, and multiple versions, or even wrong
versions, of tools. In six facilities (five HC IIIs and one HC IV), the tool acquisition process was ill
defined.

a) Almost one in five facilities (19 or 18%)

identified inadequate data quality Lack of data collection tools disproportionately
assurance as a key area needing affected lower level facilities with 21 (36%) HC III,
improvement. In these facilities, lack of eight (29%), HC IVs and three (15%) hospitals.
SOPs/guidelines for handling, reporting,
and archiving data and reports at sites and outreaches, as well as lack of awareness of SOPs,
were among areas highlighted. This affected all levels including 12 (21%) HC IIIs, four (14%) HC
IVs, and two (15%) hospitals.

b) Despite the longitudinal nature of patient data that characterize TB/HIV information systems,
issues of computerization were noted in some high-volume facilities. At least 12 such facilities
(six HC IVs, five HC IIIs, and one hospital) were not computerized. Even where electronic medical
records system were available or had once been available, inadequate power supply,
dysfunctional computers, a backlog of data due to untimely data entry, and failure of such
systems to capture all TB data variables were noted as key challenges in 10 facilities including six
HC IVs, three HC IIIs, and one hospital.

c) Confidentiality of client information was lacking in a number of mainly lower level facilities (18
total, 13 HC IIIs and five HC IVs). Additionally, some facilities had inadequate storage space for
client information and site reports. Unlockable data rooms were also noted.

d) Client management approach was also highlighted as an area that required improvement in vie
facilities (three HC IIIs and two HC IVs). Areas of concern included poor linkages of TB patients to
relevant services, sub-optimal TB case detection, and inadequate patient privacy.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 61
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 62
4. DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS

This DQA provided an opportunity to assess M&E systems both at the facility and district level, and to
validate data reported for HIV/TB services. While the findings revealed largely adequate or acceptable
M&E systems, some gaps were identified in key areas which, if not addressed, could have a negative
impact on the quality of data reported. The main areas noted to have gaps at both levels included data
management processes and data use for program improvement. Several reasons were highlighted,
including, among others, lack of appropriate reporting tools, lack of knowledge of indicators, lack of
trained staff, and incomplete documentation. As we embark on the process of rolling out new tools,
there is need to have in place clear processes for ensuring a steady supply of tools at the facilities. There
is need to ensure that all staff involved in documenting results both at the facility and district levels are
trained and provided the necessary mentoring during this process. The MoH RPMT could be an avenue
for achieving this goal, working in collaboration with other stakeholders including WHO, UNAIDS, Global
Fund, and PEPFAR.

Overall, this assessment noted low utilization of TB/HIV program data at both the facility and district
level. Few facilities were actively using their data at source, and for those using data, the scope of data
utilization was unclear. There was low level of understanding of indicator definitions at the facility and
district level yet these indicators are designed to track specific programs with targets. Innovative
approaches to increasing utilization of the data, including such measures as performance driven
financing of work plans and regular tracking of performance against targets, could potentially promote
data use at the source. Increased data use at the source will ultimately create demand for more timely
and accurate data.

One of the potential causes of the continuing weaknesses in data systems appears to be a lack of regular
supervision of districts and facilities, although facility supervision appeared to be better. There was a
notable absence of supervision and feedback on data quality to districts from MoH, especially from ACP
and RC. Perhaps human resource and financial constraints could account for this. However, for
sustainable improvement in subnational efforts, the MoH needs to lead these efforts and should
therefore urgently address this void in support supervision.

Lack of documented procedures in the form of SOPs, guidelines, and job aides for various aspects of the
data management processes in facilities and districts was noted as a recurring challenge. A large
proportion of facilities lacked documented procedures for quality assurance of data entry and
compilation of reports, archiving of source documents, followup of data quality issues, making
corrections to already submitted reports, confidentiality of patients, archiving of source documents, and
others. In the absence of appropriate guidance, errors are created or remain uncorrected leading to
several data quality issues. This void should be urgently addressed, and the MoH RC should be
encouraged to take a lead in this effort. This should involve practical measures such as development,
dissemination, and promotion of SOPs, user job aides, guidelines.

Instances where the various IPs supporting TB/HIV services were running their own separate
information systems appear to have declined significantly in recent times. This DQA showed increased
use of national tools which has helped strengthen the national system and is a positive step toward the
goal of one national M&E system. That said, there were a few isolated cases noted where IPs continue
to lead data management processes often to the total exclusion of DHTs. There were instances in which
tools are provided to facilities by IPs, without any involvement of DHTs. In fact, some facility
management teams also appear not to be involved either. There were a few cases cited in which even

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 63
compilation and submission of facility reports were undertaken by non-facility staff supported by IPs
without involvement of facility staff. The MoH, MEEPP, and IPs should work out measures to ensure that
IPs support facility and district processes and follow national guidelines.

There were few computerized data management systems for TB/HIV data, which means the opportunity
to reduce paperwork burden and associated errors is not fully exploited. For the few facilities that had
such systems, a high proportion lacked appropriate procedures for backup and quality control of data
entry. Furthermore, where backup was being implemented, there were several instances in which it was
grossly out of date. Efforts to scale up computerized patient information systems, together with
appropriate guidelines and procedures, quality assurance of data entry, data quality checks, and regular
backup should be undertaken.

Although there have been significant efforts to streamline information systems in the sector in which
many vertical and other IP-proprietary information systems were replaced to align with the national
system, the NTLP appears to continue to run a separate program. This program involves the DTLS
keeping TB information at the district level. The interface between the two systems was not clear nor
was the reason for continuing to operate these two systems in parallel. It is likely that this will lead to
duplication of efforts. Moreover, the DTLS is the only one with skills to prepare the TB/HIV report and
the program has not empowered the health unit staff to compile the TB report because reports were
previously prepared by the DTLS after visiting facilities and updating the district TB register. There is
need to build the skills of the health workers to compile the TB/HIV report that is uploaded to the DHIS 2
database. The MoH, especially the RC and NTLP program, should urgently review the place of the two
parallel systems with a view of strengthening recording and reporting of all TB data through the national
system.

Furthermore, while data in the district TB register was more likely to match the joint manual count, use
of this data for national reporting through the DHIS 2 would require harmonization of reporting periods.
The DTLS visits facilities and abstracts data from the unit TB registers throughout the quarter and not
necessarily at the end of the quarter since waiting for the end of the quarter to visit the facilities would
delay reporting to the program. Since there is some form of centralized client information at the district
level, the TB program should consider establishing web-based data entry by facilities so that the district
has timely patient data, can reconcile information for patients that have been transferred, and can
generate reports for the district TB program.

The data validation exercise revealed gross inaccuracies in the data reported to the central level for
monitoring of TB/HIV services. For all seven variables assessed, depending on the variable, 23–44% of
facilities grossly over-reported figures in the DHIS 2 database by more than 10% compared to validated
figures at the facility, while 18–38% under-reported these variables by more than 10%. Depending on
the variable, only 23–47% of the facilities reported numbers in DHIS 2 database that were within the
adequate range (0 to ±5%) of the manually verified figures.

Whereas the overall totals reported tended to be balanced out by both over-reporting and under-
reporting by facilities, for three variables the totals were still beyond the 10% margin of the validated
figures. Data reported for only two variables were within the adequate margin of error of +5%. On the
other hand, slightly better findings were noted when comparing manual counts with data documented
in the copy of the HIMS 106a report at the sites. The balancing of overall totals should not be a
consolation because the wrong figures at facility level can potentially misinform decisions in facilities
that could have dire consequences to program implementation. For instance, under-reporting could
misinform quantification and ordering of commodities which could ultimately lead to stock outs. And

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 64
over-reporting could lead to ordering excessive supplies and risks waste including expiry of expensive
commodities for TB/HIV services.

The findings of the DQA revealed that information systems for HIV/AIDS data tended to be slightly worse
than the corresponding systems for TB. This was particularly the case for pre-ART service data that was
over-reported in the copy of the HMIS 106a report at facilities by 55%, and in the DHIS 2 database by
23%. The HIV/AIDS data collection tools in facilities were also more likely to be incomplete. There are
several potential explanations for this. HIV/AIDS programs tend to have larger numbers of clients and
the need for a longer follow-up period compared to TB, and the use of electronic medical records
systems that could alleviate the paperwork burden is not widespread as yet. Secondly, TB services tend
to have more dedicated staff such as DTLS and RTLFP who often focus on TB services and information
systems which is not the situation with HIV/AIDS services. Thirdly, the institutionalized quarterly
performance and data review meetings at the district and regional levels provide opportunity for review
and feedback regarding quality of TB data. However, there is no similar process for reviewing HIV data.
These findings underscore the need for increased attention to HIV/AIDS information systems across all
levels, including steady provision of data collection and reporting tools, provision of feedback and
supervision, and better use of the data. Innovative use of the dedicated TB staff such as the DTLSs and
RTLFPs, including having them pay attention to HIV/AIDS data as well, presents an opportunity to be
exploited if the integration of TB and HIV services were extended to this level.

These findings are consistent with previous DQAs and anecdotal observations from the reporting
systems and make a very compelling case for concerted investments and efforts to improve information
systems at all levels of the health system. The weak M&E system seen during this DQA may explain
some of the observed inaccuracies that occur, since when patient records are not completed well or
records are not available, reported data are not accurate. Other reasons that could explain these
findings include, among others, transcription errors. Generally, DHIS 2 data are entered in the system by
the district biostatistician or HMIS FP which provides an opportunity for transcriptional errors.
Additionally, changes may be made to the facility HMIS form but it is unclear if similar changes are
effected in DHIS 2. To address some of these issues, there is need to strengthen data management
processes at facilities and districts. Additionally, to limit potential errors in data entry, facilities with a
large clientele should be allowed to enter data directly in DHIS 2.

There were some limitations to this DQA exercise. The facilities were selected from those that are
supported by PEPFAR IPs, which could limit generalizability to all facilities nationwide. Nonetheless,
there were also strengths to the DQA. The sample used was randomly selected and representative of
the health facilities supported by PEPFAR IPs, thus findings can be generalized to other sites. The DQA
was also able to assess systems at the district level where data are ultimately collected and used. The
collaboration with multiple stakeholders gives the findings more credibility and acceptability.
Additionally, through this exercise the number of staff with the capacity to conduct DQAs has increased.
Finally, the use of standardized tool helped to ensure comparability of findings.

In summary, the findings have highlighted gaps in data management systems that ultimately affect data
quality reported. There is a need to improve the systems and to encourage use of data to identify issues
early and correct them.

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 65
5. REFERENCES

1. Cox JA, Lukande RL, Nelson AM, Mayanja-Kizza H, Colebunders R, Van Marck E, Manabe YC. An
autopsy study describing causes of death and comparing clinico-pathological findings among
hospitalized patients in Kampala, Uganda. PLoSOne. 2012;7(3):e33685. Epub 2012 Mar 14.
PMID: 22432042.

2. Ministry of Health Uganda. 2013. National Tuberculosis and Leprosy Program (NTLP): Annual
Report 2012/2013

3. World Health Organization: Global Tuberculosis Report 2013: WHO/HTM/TB/2013.11 Geneva,

Switzerland

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 66
6. ANNEXES

Annex 1: Districts and Facilities in the TB/HIV DQA, by IP

District 1 Sub County District IP Health Facility Level
BAYLOR/COMPREHENSIVE
Adjumani Ciforo EASTERN & WEST NILE Ciforo HC III
Patongo Town
Agago Council PLAN/NU HITES Patongo HC III
Apac Chegere PLAN/NU HITES Teboke (Gvt) HC III
Apac Ibuje PLAN/NU HITES Alenga HC III
MILDMAY/COMPREHENSIVE
Buikwe Ssi-Bukunja CENTRAL Ssi HC III
MILDMAY/COMPREHENSIVE
Buikwe Nyenga CENTRAL St. Francis Nyenga Hospital
BAYLOR/COMPREHENSIVE
Bukedea Bukedea EASTERN & WEST NILE Bukedea HC IV
BAYLOR/COMPREHENSIVE
Bukedea Kidongole EASTERN & WEST NILE Kidongole HC III
MILDMAY/COMPREHENSIVE
Bukomansimbi Butenga CENTRAL Butenga HC IV
Baylor College of
Bundibugyo Busaru Medicine/PIDC/SNAPS – WEST Busaru HC IV
Baylor College of
Bundibugyo Kasitu Medicine/PIDC/SNAPS – WEST Ntandi HC III
Busia Buhehe MSH/STAR-EASTERN Buhehe HC III
Busia Lumino MSH/STAR-EASTERN Lumino HC III
Butaleja Budumba MSH/STAR-EASTERN Budumba HC III
MILDMAY/COMPREHENSIVE
Butambala Ngando CENTRAL Ngando HC III
Buvuma Town
Buvuma Council WALTER REED/MUWRP Buvuma HC IV
JOHN SNOW INC./STAR East
Buyende Nkondo Central Nkondo HC III
Dokolo Bata PLAN/NU HITES Bata HC III
Dokolo Kangai PLAN/NU HITES Kanga HC III
Gulu Awach PLAN/NU HITES Awach HC IV
Gulu Koro PLAN/NU HITES Lapainat HC III
IDI/Expand HCT & CARE IN
Hoima Kyangwali CLINICS, HOSPITALS & HC IV Kyangwali HC IV
Ibanda Nyamarebe EGPAF/STAR-SW Nyamarebe HC III
Isingiro Town
Isingiro Council EGPAF/STAR-SW Kyabirukwa HC III
Kabuyanda
Isingiro Town Council EGPAF/STAR-SW Kabuyanda HC IV

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 67
 District 1 Sub County District IP Health Facility Level
Isingiro Kashumba EGPAF/STAR-SW Kashumba HC III
Jinja Budondo TASO/HIV/AIDS & TB/PCT Lukolo HC III
Jinja Busedde TASO/HIV/AIDS & TB/PCT Mpambwa HC III
Buwenge Town
Jinja Council TASO/HIV/AIDS & TB/PCT Buwenge HC IV
Kabale Bufundi EGPAF/STAR-SW Bufundi HC III
Kabale Northern
Kabale Division EGPAF/STAR-SW Rugarama Hospital
Kabale Kabale Rubaya EGPAF/STAR-SW Rubaya HC IV
Baylor College of
Kabarole East Division Medicine/PIDC/SNAPS – WEST Kataraka HC IV
Kalangala Kalangala Kalangala DHO/HCT & Care Kalangala HC IV
JOHN SNOW INC./STAR East
Kaliro Namwiwa Central Namwiwa HC III
MILDMAY/COMPREHENSIVE
Kalungu Lwabenge CENTRAL Kiragga HC III
IDI/COMPREHENSIVE
Kampala Rubaga Division KAMPALA CITY COUNCIL Kitebi HC III
Kampala Kampala Central IDI/COMPREHENSIVE
Central Division Division KAMPALA CITY COUNCIL Makerere University HC III
Makindye Makindye IDI/COMPREHENSIVE
Division Division KAMPALA CITY COUNCIL Mukwaya Hospital
Rubaga IDI/COMPREHENSIVE
Division Rubaga Division KAMPALA CITY COUNCIL Lubaga Hospital
Kamuli Town JOHN SNOW INC./STAR East
Kamuli Council Central Kamuli Hospital
JOHN SNOW INC./STAR East
Kamuli Kitayunjwa Central Kitayunjwa HC III
Kanungu Kirima EGPAF/STAR-SW Kirima HC III
Kanungu Rugyeyo EGPAF/STAR-SW Rugyeyo HC III
Nyamwamba Baylor College of Kasese Municipal
Kasese Division Medicine/PIDC/SNAPS – WEST Council HC III
Baylor College of
Kasese Kisinga Medicine/PIDC/SNAPS – WEST Kagando Hospital
Kayunga Town
Kayunga Council WALTER REED/MUWRP Ntenjeru HC III
Kayunga Bbaale WALTER REED/MUWRP Baale HCIV
Wabwoko-
Kayunga Kitimbwa WALTER REED/MUWRP Wabwoko HC III
Kibaale Town IDI/Expand HCT & CARE IN
Kibaale Council CLINICS, HOSPITALS & HC IV Kibaale HC IV HC IV
Family Health
Resource Centre
Kiruhura Kinoni EGPAF/STAR-SW Clinic HC III
Kiryandongo Kiryandongo IDI/Expand HCT & CARE IN Kiryandongo Hospital

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 68
 District 1 Sub County District IP Health Facility Level
Town Council CLINICS, HOSPITALS & HC IV
Kisoro Nyakabande EGPAF/STAR-SW Mutolere (St. Francis) Hospital
Kitgum Town
Kitgum council PLAN/NU HITES St. Joseph'S Kitgum Hospital
Kole Aboke PLAN/NU HITES Aboke HC IV
Wekomiire/
Kyegegwa Town Baylor College of Wekomiire St.
Kyegegwa Council Medicine/PIDC/SNAPS – WEST Thereza HC III
Baylor College of
Kyenjojo Kigarale Medicine/PIDC/SNAPS – WEST Kigarale HC III
Luwero Town JOHN SNOW INC./STAR East
Luuka Council Central Ikumbya HC III
Bombo Town MILDMAY/COMPREHENSIVE
Luwero Council CENTRAL Bishop Asili Ceaser HC IV
MILDMAY/COMPREHENSIVE
Luwero Kisekka CENTRAL St. Luke Namaliga HC III
MILDMAY/COMPREHENSIVE
Lwengo Kyazanga CENTRAL Lwengo Kinoni HC III
MILDMAY/COMPREHENSIVE
Lwengo Magale CENTRAL Kyazanga HC IV
Manafwa Oluffe TASO/HIV/AIDS & TB/PCT Magale HC IV
BAYLOR/COMPREHENSIVE
Maracha Mukungwe EASTERN & WEST NILE Ovujo HC III
Nyendo-
Senyange MILDMAY/COMPREHENSIVE
Masaka Division CENTRAL Kiyumbya HC IV
Kimanya-
Kyabakuza MILDMAY/COMPREHENSIVE
Masaka Division CENTRAL St. Joseph Kitovu Hospital
MILDMAY/COMPREHENSIVE
Masaka Kityerera CENTRAL Masaka RRH
JOHN SNOW INC./STAR East
Mayuge Bukonde Central Kityerera HC IV
Mbale Nyondo MSH/STAR-EASTERN Bufumbo HC IV
Mbale Northern
Mbale Division MSH/STAR-EASTERN Nyondo HC III
Clinic -
Mbale Bugamba MSH/STAR-EASTERN TASO Mbale Clinic Special
Kamukuzi
Mbarara Division MUFM/MJAP Bugamba HC IV
Mbarara Laropi MJAP/MUFM Ruharo Mission Hospital
BAYLOR/COMPREHENSIVE
Moyo Metu EASTERN & WEST NILE Laropi HC III
Mubende Town BAYLOR/COMPREHENSIVE
Moyo Council EASTERN & WEST NILE Metu HC III

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 69
 District 1 Sub County District IP Health Facility Level
MILDMAY/COMPREHENSIVE Mubende
Mubende Kitenga CENTRAL Rehabilitation Centre HC III
MILDMAY/COMPREHENSIVE
Mubende Ntenjeru CENTRAL Kalonga HC III
Mukono Nakisunga WALTER REED/MUWRP Kojja HC IV
Mukono Nakitoma WALTER REED/MUWRP Seeta-Nazigo HC III
Kalongo MILDMAY/COMPREHENSIVE
Nakasongola (Nakasongola) CENTRAL Nakitoma HC III
MILDMAY/COMPREHENSIVE
Nakasongola Sigulu Islands CENTRAL Bamugolodde HC III
JOHN SNOW INC./STAR East
Namayingo Mutumba Central Mutumba HC III
BAYLOR/COMPREHENSIVE
Nebbi Nyaravur EASTERN & WEST NILE Kucwiny HC III
BAYLOR/COMPREHENSIVE
Nebbi Alero EASTERN & WEST NILE Angal St. Luke Hospital
Nwoya Pajule PLAN/NU HITES Alero HC III
Pader Kabwangasi PLAN/NU HITES Pajule HC IV
Pader Pader PLAN/NU HITES Pader HCIV
Pallisa Apopong MSH/STAR-EASTERN Kabwangasi HC III
Pallisa Kalisizo MSH/STAR-EASTERN Apopong HC III
Rakai Rakai MUSPH/FELLOWS/Rakai Kalisizo Hospital
Rakai Buhunga MUSPH/FELLOWS/Rakai Rakai Hospital
Rukungiri
Southern
Rukungiri Division EGPAF/STAR-SW Buhunga HC IV
St. Karolii Lwanga
Rukungiri Kyere EGPAF/STAR-SW Nyakibale Hospital
BAYLOR/COMPREHENSIVE
Serere Shuuku EASTERN & WEST NILE Kyere HC III
Sheema Buteza EGPAF/STAR-SW Shuuku HC IV
Soroti Northern
Sironko Division MSH/STAR-EASTERN Buteza HC III
BAYLOR/COMPREHENSIVE
Soroti Gweri EASTERN & WEST NILE Soroti RRH
BAYLOR/COMPREHENSIVE
Soroti Tubur EASTERN & WEST NILE Gweri HC III
Tororo Eastern BAYLOR/COMPREHENSIVE
Soroti Division EASTERN & WEST NILE Tubur HC III
Tororo Merikit TASO/HIV/AIDS & TB/PCT Tororo Hospital
Kira Town
Tororo Council TASO/HIV/AIDS & TB/PCT Merikit HC III
MILDMAY/COMPREHENSIVE
Wakiso Nangabo CENTRAL Bweyogerere HC III

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 70
 District 1 Sub County District IP Health Facility Level
Paidha Town MILDMAY/COMPREHENSIVE
Wakiso Council CENTRAL Kasangati HC IV
BAYLOR/COMPREHENSIVE
Zombo Zeu EASTERN & WEST NILE Paidha HC III
BAYLOR/COMPREHENSIVE
Zombo Abanga EASTERN & WEST NILE Zeu HC III
BAYLOR/COMPREHENSIVE
Zombo Abanga EASTERN & WEST NILE Pakadha HC III

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 71
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 72
Annex 2: District TB/HIV Data Verification and System Assessment Tool
District TB HIV Data System Assessment Tool

Health District:

A) TB_SCREEN: Percentage of PLHIV in HIV clinical care who were screened for
TB symptoms at the last clinical visit
Variables:
1. The number of PLHIV who were screened for TB symptoms at the last
clinical visit to an HIV care facility during the reporting period disaggregated
by age & sex
2. Number of HIV-positive adults and children who received at least one of
the following during the reporting period: clinical assessment (WHO staging)
OR CD4 count OR viral load - number Pre-ART
3. Number of HIV-positive adults and children who received at least one of
the following during the reporting period: clinical assessment (WHO staging)
OR CD4 count OR viral load - Number current on ART

B) TB_STAT: Percentage of registered new and relapsed TB cases with

Indicator Reviewed: documented HIV status
Variables:
4. Number of registered new and relapsed TB cases with documented HIV
status, during the reporting period disaggregated by number HIV-positive and
number HIV negative.
5. Total number of registered new and relapsed TB cases, during the reporting
period.

C) TB_ART: Percentage of HIV-positive new and relapsed registered TB cases

on ART during TB treatment
Variables:
6. The number of registered TB cases with documented HIV-positive status
who start or continue ART during the reporting period
7. Number of registered new and relapsed TB cases with documented HIV
status, during the reporting period - number HIV Positive.

Date of Review:

Reporting Period
October 2014 through March 2015
Verified:

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 73
 COMMENTS
(Please provide
Check the Answer that applies
detail for each
- Yes Completely
Component of the response
- Partly
M&E System Detailed responses
- Not at All
will help guide
- Not Applicable
strengthening
measures.)

Part 2. Systems Assessment

I - M&E Structure, Functions and Capabilities to handle HIV/TB information

Are there staff designated to review the quality of
TB data (i.e., accuracy, completeness and timeliness)
1 received from health facilities. (Briefly
describe who & how the District TB Register is filled
at this site)

Does the district have staff designated the

responsibiity of reviewing and approving the district
TB quarterly report prior to submission to the
2
national level.
(Note that the reviewer can be from either the
district or national levels)

Is there a procedure in place to ensure the TB

quarterly district report is completed and submitted
3
in the event the responsible staff is unavailable (e.g.
shared duties, a team approach etc.)

Is feedback systematically provided to all health

facilities on the quality of their TB reporting (i.e.,
4 accuracy, completeness and timeliness). (Specify
who provides feedback as well as the frequency and
form of providing the feedback)

Does the district conduct regular supervisory visits

to the health facilities in the district according to the
5 guidelines.(…If yes, when was the last last round of
support supervision (month and year (mm/yyyy)
and which facilities were visiited?)

Does the district receive regular feedback from the

national program on the quality of their submitted
6
reports. (…If yes, specify when the district last
recieved feedback (mm/yyyy) and from whom?)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 74
 Does the district receive regular supervisory visits
from the national program according to the
7 guidelines on supervision. (…If yes, specify the team
that provided the support supervision and date for
the last visit (month and year (mm/yyyy)))

II- Understanding of Indicator Definitions and Reporting Guidelines

Has the district been provided with the National written M&E guidelines for
its sub-reporting level on …
………. what should be recorded in the source
documents/registers?
8 (probe and comment on whether the team at the
district understands the questions/variables to be
filled in the registers)

………. what should be included on the HMIS

quarterly report (HMIS 106a TB/LP section) ?
(probe and comment on whether the data
9
aggregation team at the district understands how
the HMIS 106a quarterly summary reports are
compiled)

10 ……...… how the reports should be submitted?

11 ……...… to whom the reports should be submitted?

12 ……… when the reports are due?

Are there written district specific instructions to

13 ensure standardized recording and reporting of
TB/HIV data?

III - Availability of Data-collection Tools and Reporting Forms for TB/HIV

sevices
Does the District have the national TB Register to be
14
used at their reporting level?

…If yes, is refgister consistently used at the district?

15
(probe and list reasons incase not used consistently)

If there are multiple organizations that are

implementing activities under the TB/HIV services at
16
the District, do they use the national reporting forms
and follow the same reporting timelines?
(Name the organization/ Implementing Partners in

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 75
 the comment section and tools used if different)

Are there sufficient stocks of blank registers at the

District?
17 (Enquire and comment on when the existing stock
of TB registers are expected to run out -estimate
based on list of tools provided in #18 above)

Does the district monitor stocks of data collection

18 tools/reporting forms at facilities to ensure their
continuous availability (i.e. forecasting.)?

IV - Data Management Processes

Does the district have computerized TB/HIV data
and/ reports
19
(If yes, specify what data is computerized and the
computer packages used)

Are there quality controls in place for when data

from paper-based forms are entered into a
computer to ensure the accuracy of data entry (e.g.
edit and/or logic checks, post-data entry verification,
20
etc.)?
(In the comment section, briefly explain the data
quality controls and also enquire about their
computerized data backup program)

Does the district routinely create back-up files for

21
the TB/HIV data?

If yes, is the latest date of back-up appropriate given

22 the frequency of update of the computerized system
(e.g., back-ups are weekly or monthly).

Does the district have a written procedure to

address late, incomplete, inaccurate and missing
23
reports; including following-up with service points
on data quality issues?

……… Has the district ever uncovered errors in the

reports from health facilities?
24
(If yes describe how these errors have been resolved
in the comments section)

25 Is there a written policy that states for how long

source documents and reporting forms should to be

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 76
 retained?

Does the district have a writen policy that describes

how program documents (e.g. source documents
26
and reporting forms) should be archived (e.g. filing
cabinets, storage rooms etc.)?

V - Use of data for decision making

Is there evidence of TB/HIV data use at the District?
(check for evidence of data analysis & reporting
27
other than the official reports e.g. charts, graphs,
maps, etc. (ask to see if not displayed)?

...If yes, is there a staff assigned staff to develop

28
them regularly?

Is there a staff at the district who takes lead in

29
analysis and interpretation of TB/HIV data?

Is the analyzed data/results presented/disseminated

to other information system stakeholders in a timely
manner so that the information can be used to
30
inform decisions?
(Probe about examples of target audiences for the
data e.g. DHT, IPs, &Other stakeholders)

Are there any programmatic decisions taken by the

31
district based on analyzed TB/HIV data/results?

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 77
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 78
Annex 3a: Health Facility TB/HIV Data Verification and Validation Tool
Health Facility Data Verification and Validation Tool

Service Delivery Site:

Health
Health District: Sub-
District:
A) TB_SCREEN: Percentage of PLHIV in HIV clinical care who were screened for TB symptoms at the last clinical
visit
Variables 1. The number of PLHIV who were screened for TB symptoms at the last clinical visit to an HIV care
facility during the reporting period disaggregated by age & sex
2. Number of HIV-positive adults and children who received at least one of the following during the
reporting period: clinical assessment (WHO staging) OR CD4 count OR viral load - number Pre-ART
3. Number of HIV-positive adults and children who received at least one of the following during the
reporting period: clinical assessment (WHO staging) OR CD4 count OR viral load - Number current on ART

Indicator Reviewed: B) TB_STAT: Percentage of registered new and relapsed TB cases with documented HIV status
Variables: 4. Number of registered new and relapsed TB cases with documented HIV status, during the reporting
period disaggregated by number HIV-positive and number HIV negative.
5. Total number of registered new and relapsed TB cases, during the reporting period.

C) TB_ART: Percentage of HIV-positive new and relapsed registered TB cases on ART during TB treatment
Variables: 6. The number of registered TB cases with documented HIV-positive status who start or continue ART
during the reporting period
7. Number of registered new and relapsed TB cases with documented HIV status, during the reporting
period - number HIV Positive.

Date of Review:
Reporting Period
October 2014 through March 2015
Verified:
Part 1: Data Verifications
TB_STAT TB_ART
TB _SCREEN
(Percentage of (Percentage of HIV-
(Percentage of PLHIV in HIV
A - Documentation registered new and positive new and
clinical care who were screened
Review: relapsed TB cases with relapsed registered
for TB symptoms at the last
documented HIV TB cases on ART
clinical visit)
status) during TB treatment)
Variable 1: Variable 2: Variable 3: Variable 4: Variable 5: Variable 6: Variable 7:
# PLHIV # HIV+ve # HIV+ve # registered Total The number #
who were adults and adults and new and number of of registered registered
screened children children relapsed TB registered TB cases new and
for TB who who cases with new and with relapsed
symptoms received at received at documented relapsed TB documented TB cases COMMENTS
at the last least one of least one of HIV status, cases, HIV+ve with
Review availability clinical visit the the during the during the status who documente
and completeness of all to an HIV following following reporting reporting start or d HIV
care facility during the during the period period continue status,
indicator data sources during the reporting reporting disaggregated ART during during the
for the selected reporting period: period: by # HIV+ve & the reporting
period clinical clinical # HIV negative reporting period - #
reporting period. assessment assessment period HIV+ve
(WHO (WHO
staging) or staging) or
CD4 count CD4 count
or viral load OR viral load
- # Pre-ART - # current
on ART
Variable 1
Review available data
sources for the reporting Variable 2
period being verified. Are Variable 3
1 all necessary data sources
Variable 4
available for review? (Y/N)
Variable 5
Briefly Comment as Variable 6

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 79
 appropriate Variable 7
If no, determine how this
might have affected reported
numbers.
Variable 1

Are all available data Variable 2

sources complete (essentil Variable 3
data fields)?
Variable 4

2 Briefly Comment as Variable 5

appropriate Variable 6

Variable 7
If no, determine how this
might have affected reported
numbers.
Review the data sources: Variable 1
Is information available Variable 2
covering the period under Variable 3
review
Variable 4

3 Briefly Comment as Variable 5

appropriate Variable 6

Variable 7

If no, determine how this

might have affected reported
numbers.

Part 1: Data Verifications

TB_STAT TB_ART
TB _SCREEN
(Percentage of (Percentage of HIV-
(Percentage of PLHIV in HIV
B - Recounting registered new and positive new and
clinical care who were screened
reported Results: relapsed TB cases with relapsed registered TB
for TB symptoms at the last
documented HIV cases on ART during TB
clinical visit)
status) treatment)
Recount results from Variable 1: Variable 2: Variable 3: Variable 4: Variable 5: Variable 6: Variable 7:
# PLHIV # HIV+ve # HIV+ve # registered Total The number # registered
data source, compare who were adults and adults and new and number of of registered new and
the verified numbers to screened children children relapsed TB registered TB cases relapsed TB
for TB who who cases with new and with cases with
the service delivery site symptoms received at received at documented relapsed TB documented documented COMMENTS
reported numbers and at the last least one of least one of HIV status, cases, HIV+ve HIV status,
clinical visit the the during the during the status who during the
explain discrepancies (if to an HIV following following reporting reporting start or reporting
any). care facility during the during the period period continue period - #
during the reporting reporting disaggregated ART during HIV+ve
reporting period: period: by # HIV+ve the
period clinical clinical and # HIV reporting
assessment assessment negative period
(WHO (WHO
staging) or staging) or
CD4 count CD4 count
or viral load OR viral load
- # Pre-ART - # current
on ART
Enter the number of
clients reported by the site
during the reporting
4
period from the site
summary report (DHIS2).
[A]
Oct - Dec 2014

Jan - Mar 2015

Recount the number of
clients during the
5 reporting period by
reviewing the data source
(Register(s)). [B]

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 80
 Oct - Dec 2014
Jan - Mar 2015
Calculate the ratio of
6 reported to recounted
numbers.[A/B]
#DIV/ #DIV/ #DIV/0
Oct - Dec 2014
0! 0! !
#DIV/ #DIV/ #DIV/ #DIV/0 #DIV/ #DIV/0 #DIV/0
Jan - Mar 2015
0! 0! 0! ! 0! ! !
Variable 1
For each indicator; Variable 2
What are the reasons for
Variable 3
the discrepancy (if any)
7 observed (i.e., data entry Variable 4
errors, arithmetic errors, Variable 5
missing data source,
Variable 6
other)?
Variable 7

C - Cross-check reported results with other data sources:

Cross-checks can be performed by examining separate inventory records during the reporting
period to see if these numbers corroborate the reported results.
CROSS-CHECK 1: Cross- TB_STAT TB_ART
TB _SCREEN
check primary data source (Percentage of (Percentage of HIV-
(Percentage of PLHIV in HIV
(Register) with secondary registered new and positive new and
clinical care who were screened
relapsed TB cases with relapsed registered TB
data source (onsite HMIS for TB symptoms at the last
documented HIV cases on ART during TB
106a). (If cross-checks are clinical visit)
status) treatment)
different from the planned Variable 1: Variable 2: Variable 3: Variable 4: Variable 5: Variable 6: Variable 7:
cross-check, i.e. the cross- # PLHIV # HIV+ve # HIV+ve # registered Total The number # registered
who were adults and adults and new and number of of registered new and
checks entered on the screened children children relapsed TB registered TB cases relapsed TB
Information page, specify the for TB who who cases with new and with cases with
symptoms received at received at documented relapsed TB documented documented
cross-checks performed in at the last least one of least one of HIV status, cases, HIV+ve HIV status, COMMENTS
the comment cells to the clinical visit the the during the during the status who during the
to an HIV following following reporting reporting start or reporting
right.) care facility during the during the period period continue period - #
during the reporting reporting disaggregated ART during HIV+ve
reporting period: period: by # HIV+ve the
period clinical clinical and # HIV reporting
assessment assessment negative period
(WHO (WHO
staging) or staging) or
CD4 count CD4 count
or viral load OR viral load
- # Pre-ART - # current
on ART
Enter the number of
clients reported by the
site during the reporting
1.1
period from the site
summary report (onsite
HMIS 106a)
Oct - Dec 2014
Jan - Mar 2015
Recount the number of
clients during the
reporting period by
1.2
reviewing the data source
(Use joint count from the
Register(s))
Oct - Dec 2014
Jan - Mar 2015
Calculate % difference for Variable 1
1.3 cross check 1: Variable 2

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 81
 What are the reasons for Variable 3
the discrepancy (if any) Variable 4
observed (i.e., data entry
Variable 5
errors, arithmetic errors,
missing data source, Variable 6
other)? Variable 7

#DIV/ #DIV/ #DIV/ #DIV/0 #DIV/ #DIV/0 #DIV/0

Oct - Dec 2014
0! 0! 0! ! 0! ! !
#DIV/ #DIV/ #DIV/ #DIV/0 #DIV/ #DIV/0 #DIV/0
Jan - Mar 2015
0! 0! 0! ! 0! ! !
C - Cross-check reported results with other data sources:
Cross-checks can be performed by examining separate inventory records during the reporting
period to see if these numbers corroborate the reported results.
CROSS-CHECK 2: Cross- TB_STAT TB_ART
in HIV clinical care
check the Joint Count from (Percentage of (Percentage of HIV-
Facility TB Register with the who were screened registered new and positive new and
District TB Register count. for TB symptoms at relapsed TB cases with relapsed registered TB
(If cross-checks are different documented HIV cases on ART during TB
the last clinical visit) status) treatment)
from the planned cross-
Variable 4: Variable 5: Variable 6: Variable 7:
check, i.e. the cross-checks load - Varia # registered Total The number # registered
entered on the Information ble 3: new and number of of registered new and COMMENTS
relapsed TB registered TB cases relapsed TB
page, specify the cross- cases with new and with cases with
checks performed in the dc documented relapsed TB documented documented
HIV status, cases, HIV+ve HIV status,
comment cells to the right.) during the during the status who during the
reporting reporting start or reporting
period period continue period - #
disaggregated ART during HIV+ve
by # HIV+ve the
and # HIV reporting
negative period
Recount the number of
2. clients during the
1 reporting period from the
district TB register. [A]
Oct - Dec 2014
Jan - Mar 2015
Enter the number of
clients during the
2. reporting period
2 recounted from the facility
TB register in 5 above - .
[B]
Oct - Dec 2014
Jan - Mar 2015
Calculate % difference for Variable 1
cross check 1: Variable 2
What are the reasons for
Variable 3
the discrepancy (if any)
2.
observed (i.e., data entry Variable 4
3
errors, arithmetic errors, Variable 5
missing data source,
Variable 6
other)?
Variable 7

#DIV/ #DIV/ #DIV/0 #DIV/ #DIV/0 #DIV/0

Oct - Dec 2014 0!
0! 0! ! 0! ! !
#DIV/ #DIV/ #DIV/ #DIV/0 #DIV/ #DIV/0 #DIV/0
Jan - Mar 2015
0! 0! 0! ! 0! ! !

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 82
Annex 3b: Health Facility TB/HIV System Assessment Tool
Health Facility System Assessment Tool
Service Delivery Site:

Health District: Health Sub_District

A) TB_SCREEN: Percentage of PLHIV in HIV clinical care who were

screened for TB symptoms at the last clinical visit
Variables:
1. The number of PLHIV who were screened for TB symptoms at the
last clinical visit to an HIV care facility during the reporting period
disaggregated by age & sex
2. Number of HIV-positive adults and children who received at least
one of the following during the reporting period: clinical assessment
(WHO staging) OR CD4 count OR viral load - number Pre-ART
3. Number of HIV-positive adults and children who received at least
one of the following during the reporting period: clinical assessment
(WHO staging) OR CD4 count OR viral load - Number current on ART

B) TB_STAT: Percentage of registered new and relapsed TB cases

Indicator Reviewed: with documented HIV status
Variables:
4. Number of registered new and relapsed TB cases with documented
HIV status, during the reporting period disaggregated by number HIV-
positive and number HIV negative.
5. Total number of registered new and relapsed TB cases, during the
reporting period.

C) TB_ART: Percentage of HIV-positive new and relapsed registered

TB cases on ART during TB treatment
Variables:
6. The number of registered TB cases with documented HIV-positive
status who start or continue ART during the reporting period
7. Number of registered new and relapsed TB cases with documented
HIV status, during the reporting period - number HIV Positive.

Date of Review:

Reporting Period Verified: October 2014 through March 2015

Check the
Answer COMMENTS
Component of the M&E that applies (Please provide detail for each response
System - Yes Detailed responses will help guide strengthening
Completely measures. )
- Partly
- Not at All

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 83
 - Not
Applicable

Part 2. Systems Assessment

I - M&E Structure, Functions and Capabilities to handle HIV/TB information

Is the responsibility for recording TB/HIV service delivery

information in the Registers clearly assigned to the
1 relevant staff?
(Briefly describe who & how the Registers [TB register, Pre
ART register, ART register] are filled at this site)

Are there processes in place to ensure that data

compilation and reporting is completed in the event that
the responsible staff is not available to do the job (e.g.
2
shared duties, a team approach etc.)?
(Briefly describe for all Registers used for TB/HIV Services
[TB register, Pre ART register, ART register])

Are there designated staff responsible for reviewing

periodic reports prior to submission to the next level?
3
(Note that the reviewer can be from any level - Health
Facility, sub-district, district or national levels)

Does the Health facility receive regular feedback on the

quality of their submitted reports?
4
(Specify persons, frequency and form of providing the
feedback)

Does the health facility receives regular TB/HIV supportive

supervisory visits from district and/or national level staff
and/or other Organisation according to the guidelines?
5
(…If yes, specify the team that provided the support
supervision and date for the last visit (month and year
(mm/yyyy))

Are the person(s) responsible for collecting and reporting

TB/HIV date trained/ oriented in data collection &
reporting?
(In the comments section specify:
6
1) if the site has an HMIS/Records person, 2) if the records
person assists with TB/HIV reporting,
3) if the site has a training/orientation plan for new staff in
data collection)

II- Understanding of Indicator Definitions and Reporting Guidelines

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 84
 Has the site been provided with the National written M&E
guidelines for its sub-reporting level on …

……….what should be recorded in the source

documents/registers.
7 (probe and comment on whether the data collection
team at the site understands the questions/variables to be
filled in the various source documents/registers)

………. what should be included on the HMIS quarterly

report.
8 (probe and comment on whether the data collection
team at the site understands how the HMIS 106a
quarterly summary reports are compiled)

9 ……...… to whom the reports should be submitted.

10 ……… when the reports are due.

……… in case of errors, does the site know how to

communicate and effect changes in a report that was
11 previously submitted to the district.
(in the comment section briefly describe the process used
at the site)

Are the written instructions provided adequate to ensure

12
standardized recording and reporting of TB/HIV data.

Do the facility staff understand how to derive the

variables for the indicator "Percentage of PLHIV in HIV
13
clinical care who were screened for TB symptoms at the
last clinical visit" in the register and HMIS 106a report

Do the facility staff understand how to derive the

14 variables for the indicator "Percentage of registered new
and relapsed TB cases with documented HIV status"

Do the facility staff understand how to derive the

variables for the indicator "Percentage of HIV-positive
15
new and relapsed registered TB cases on ART during TB
treatment"

III - Availability of Data-collection Tools and Reporting Forms for TB/HIV services

Does the Health facility have the national HMIS

forms/tools to be used at their reporting level?
16
(In the comment section explain if the following MOH
tools are currently available at the site (specify version
where possible), 1)HMIS 096: TB Register; 2)HMIS 080:

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 85
 Pre-ART Register; 3)HMIS 081: ART Register; 4)HMIS 106a:
Health Unit Quarterly Report,??4)TB Case management
HMIS 089D Intensified TB case finding guide, HMIS 089E
TB client card and HIV care/ART cards

…If yes, are the standard forms/tools consistently used at

the Health facility?
17
(probe and list reasons incase standard forms/tools are
not used consistently)

If there are multiple organizations that are implementing

activities under the TB/HIV services at the site, do they
use the national reporting forms and follow the same
18
reporting timelines?.
(Name the organization/ Implementing Partners in the
comment section and tools used if different)

Are there sufficient stocks of blank primary data collection

tools/ /registers and summary HMIS forms at the facility.
(Enquire and comment on when the existing stock of
19
TB/HIV registers & HMIS summary forms are expected to
run out -estimate based on list of tools provided in #16
above)

Does the site have a regular refill program of the TB/HIV

20 Tools when they are used up?
(Enquire where and how the site gets these refills)

Do the Primary data collection tools/registers have all the

relevant questions/variables needed to compile the HMIS
21 106a reports?
(Enquire about how the staff obtain the various answers
needed in the compilation of the said reports)

IV- Data Management Processes

Does the established patient recording and reporting

system allow tracking of unique individuals within and
across Service Delivery Points to avoid double counting at
22
this site?
(e.g. can a patient with a clinical diognoses of supected TB
in the HIV Clinic be tracked through to the special TB unit)?

Are there data quality controls in place for ensuring

compilation of accurate quarterly (HMIS 106a) reports
23 (e.g. controls for detection of data inconsistencies;
incomplete /incorrect TB/HIV Reporting, missing data &
transcription errors)?

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 86
 Does the site have computerized TB/HIV data and/ reports
24 (If yes, specify what data is computerized and the
computer packages used)

Ask only where applicable, For computerized sites are

there quality controls in place for when data from
Registers/paper-based forms are entered into a computer
to ensure the accuracy of data entry (e.g. edit and/or logic
25
checks, post-data entry verification, etc.).
(In the comment section, briefly explain the data quality
controls and also enquire about their computerized data
backup program)

Confidentiality of patients information: Is the patient data

26 maintained according to national confidentiality
guidelines?

Are there SOPs/ Job Aids/Guidelines that describe how

27 site documents should be handled and archived clearly
written and accessible to all staff?

If yes, are the staff aware and using the above SOPs/ Job
Aids/Guidelines?
28
(Briefly explain, and also enquire if the site has the latest
TB/HIV Guidelines)

Is the data archiving/storage system at the site adequate?

(Assess data storage, security and ease of accessibility &
29
retrieval of TB/HIV site records (e.g. filing cabinets,
storage rooms etc.)

Is the HMIS 106a report submitted in a timely manner?

(Enquire and observe if the 'April - June 2015' quarterly
30
reports was compiled and ask to see copy, cross check on
the date of compilation)

V - Use of data for decision making

Is there evidence of TB/HIV data use at the site?

(check for evidence of data analysis & reporting other
31
than the official reports e.g. charts, graphs, maps, etc.
(ask to see if not displayed)?

Have the TB/HIV services staff been trained in data

32 analysis and interpretation?
(Probe for trainings done and duration)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 87
 Is there a staff at the site who takes lead in analysis and
33
interpretation of TB/HIV data?

Is the analyzed data/results presented/disseminated to

other information system stakeholders in a timely manner
so that the information can be used to inform decisions?
34
(Probe about examples of target audiences for the data
e.g. Health Unit administration, sub county council, HSD,
DHT &IPs)

Are there any programmatic decisions taken by the service

35 delivery site based on analyzed TB/HIV data/results.
(Ask to see example)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 88
Annex 3c: General Observations and Notable Good M&E Practices at Site

Data Verification and System Assessment Sheet - Service Delivery Site

Service Delivery Site:

Health District: Health Sub-District:

Part 3: General Observations & Notable Good M&E Practices at Site

i) General Observations about Site

ii) Notable Good M&E Practices at Site:

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 89
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 90
Annex 3d: Site Feedback Form
A Indicator:- Percentage of PLHIV in HIV clinical care who were screened for TB symptoms at the last
clinical visit

A1- Documentation review at site

Variable Availability of all Completeness of Availability of Good Areas for

required datasources all availabledata data for period practices Improvement
(registers etc) sources under review

# Screened for TB

Active Pre ART

Current on ART

A2 - Understanding of variable & Validation of Reported Data against this Site

Variable Understanding of how Validated Count at Validated Good Areas for

to derive variable for Site vs Report in Count at Site vs practices Improvement
reporting DHIS2 Reported in
HMIS 106a

# Screened for TB

Active Pre ART

Current on ART

B Indicator: - Percentage of registered new & relapsed TB cases with documented HIV
Status

B1- Documentation review at site

Variable Availabailty of all Completeness of Availability of Good Areas for

required datasouces all availabledata data for period practices Improvement
(registers etc) sources under review

Registered new &

Relapsed TB cases
with documented
HIV status

Total Number of
new & relapsed
registered TB cases

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 91
 B2 - Understanding of variable & Validation of Reported Data against this Site

Variable Understanding of how Validated Count at Validated Good Areas for

to derive variable for Site vs Report in Count at Site vs practices Improvement
reporting DHIS2 Reported in
HMIS 106a

Registered new &

Relapsed TB cases
with documented
HIV status

Total Number of
new & relapsed
registered TB cases

C Indicator - %age of HIV-positive registered new & relapsed TB cases on ART

C1- Documentation review at site

Variable Availability of all Completeness of all Availability of Good Areas for

required data sources available data data for period practices Improvement
(registers etc) sources under review

HIV-positive
registered new &
relapsed TB cases
on ART

HIV-positive
registered new &
relapsed TB cases

C2 - Understanding of variable & Validation of Reported Data against this Site

Variable Understanding of how Validated Count at Validated Good Areas for

to derive variable for Site vs Report in Count at Site vs practices Improvement
reporting DHIS2 Reported in
HMIS 106a

HIV-positive
registered new &
relapsed TB cases
on ART

HIV-positive
registered new &
relapsed TB cases

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 92
 D M&E System Assessment

Observations

Component Good practices Areas for Improvement

I - M&E Structure,
Functions and
Capabilities to
handle HIV/TB
information

II - Understanding
of Indicator
Definitions and
Reporting
Guidelines

III - Availability of
Data-collection
Tools & Reporting
Forms for TB/HIV
services

IV- Data
Management
Processes

V - Use of data for

decision making

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 93
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 94
Annex 3e: Site Recommendations and Action Plans

Data Verification and System Assessment Sheet - Service Delivery Site

Service Delivery Site:

Health District: Health Sub-District:

Recommendations for the TB/HIV Services

Based on the findings of the systems’ review and data verification at the service site,
please describe four key challenges to the data quality identified and recommended
strengthening measures, with an estimate of the length of time the improvement
measure could take. These should be discussed with the site staff.

Resources
Action Point (What is
Timeline Responsible Person(s)
(Specific required in
(When do (Person/organisation
Identified activites on order to
you hope that will be responsible
Weaknesses improving achieve carry
to achieve for accomplishing this
TB/HIV out agreed
this) task)
services) upon,
activities)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 95
Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 96
Annex 4: DQAI Team Members and Central Coordinating Team
DQA
Districts
Team # Name Organization Email address
Team 1 1 Dr Lydia Kiryabwire Knowledge hub [email protected]
2 Ms Margaret Basia NA [email protected]
3 Ms. Nampijja Proscovia Masaka RPMT [email protected] Butambala,
4 Immaculate Baseka MEEPP [email protected] Kalungu,
Lwengo,
5 IP Representative - - Raka
6 District Representative - - (10 sites)
7 USG Representative - -
MEEPP norah
8 Dr Norah Namuwenge Consultant [email protected]

Team 2 1 Leornard Rotich Pallisa District [email protected]

[email protected]
2 Mr. Kaggwa Charles Central RPMT m Luwero,
3 Beatrice Nandutu MEEPP [email protected] Mubende
Nakasongola
4 IP Representative - - ,
5 District Representative - - Wakiso
(9 sites)
6 USG Representative - -
MEEPP norah
7 Dr Norah Namuwenge Consultant [email protected]

Team 3 1 Dr Thomas Ochar Tororo Hospital [email protected]

2 Dr Ebony Quito Global Fund [email protected]
Dr. Nassozi Christine
3 Ogwang Central RPMT [email protected]; Kampala,
Dr Barbara Amuron/ [email protected]/"Barry, Buikwe,
4 Dr S. Barry MEEPP [email protected] Buvuma,
5 IP Representative - - Mukono
(10 sites)
6 District Representative - -
7 USG Representative - -
MEEPP norah
8 Dr Norah Namuwenge Consultant [email protected]

Team 4 1 Dr Ann Nakinsige Jinja RPMT [email protected] Buyende,

Jinja,
2 Tabula Denis Buvuma District [email protected]
Kaliro,
[email protected]
Luuka,
3 Joseph Mukasa Jinja RPMT m;
Mayuge,
4 Dr Margaret Mugerwa MEEPP [email protected] Namayingo
5 IP Representative - - (10 sites)

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 97
 DQA
Districts
Team # Name Organization Email address
6 District Representative - -
7 USG Representative - -
[email protected]
8 Dr George Upenytho NTLP m

[email protected]
Team 5 1 Etwom Alfred Track TB Project [email protected]
2 Acaku Moses Mbale RPMT [email protected] Busia,
3 Fatuma Matovu MEEPP [email protected] Butaleja,
Manafwa,
5 IP Representative - - Mbale,
6 District Representative - - Tororo
(9 sites)
7 USG Representative - -
[email protected]
8 Dr George Upenytho NTLP m

Team 6 1 Ms Juliet Arica Gulu RRH [email protected]

2 Lazarus Onzima Moroto RPMT [email protected]
Pallisa,
3 Napyo Agnes Kasede Mbale RPMT [email protected] Sironko,
Bukede,
4 IP Representative - -
Serere,
5 District Representative - - Soroti
6 USG Representative - - (9 sites)
[email protected]
7 Dr George Upenytho NTLP m

Team 7 1 Dr. Julius Mugweri Gulu RPMT [email protected]

Agago,
2 Mr Obizu Moses Jinja RRH [email protected] Gulu,
3 Dr Hudson Balidawa Global Fund Kitgum,
Lamwo,
4 Ms Jenifer Balaba Bogere MEEPP [email protected] Nwoya,
5 IP Representative Pader,
Apac,
6 District Representative
Dokolo,
7 USG Representative Kole
[email protected] (12 sites)
8 Dr George Upenytho NTLP m

Team 8 1 Ivan Lukaabwe Baylor seconded [email protected] Bundibugyo,

MEEPP Kabarole,
2 Dr Solome Bakera Consultant [email protected] Kasese
[email protected] District,
3 Richard Ongom MEEPP m Kyegegwa,
Kyenjojo,
4 IP Representative - -

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 98
 DQA
Districts
Team # Name Organization Email address
Hoima,
5 District Representative - -
Kibaale,
6 USG Representative - - Kiryandongo
MEEPP norah (10 sites)
7 Dr Norah Namuwenge Consultant [email protected]

[email protected]
Team 9 1 Dr Stavia Turyahabwe Global Fund om
2 Tumwesigye Livingstone Mbarara RPMT [email protected]; Ibanda,
3 Joseph Kimera MEEPP Isingiro,
Kiruhura,
4 IP Representative - - Mbarara,
5 District Representative - - Sheema
(10 sites
6 USG Representative - -
MEEPP norah
7 Dr Norah Namuwenge Consultant [email protected]

Team 10 1 Antonita Nakya Kiwangala HC IV [email protected]

2 Dr. Asiimwe Justus Mbarara RPMT [email protected]
Kabale,
3 Dr. Micheal Kakinda Hoima RPMT [email protected]
Kanungu,
4 IP Representative - - Kisoro,
5 District Representative - - Rukungiri
(8 sites)
6 USG Representative -
MEEPP norah
7 Dr Norah Namuwenge Consultant [email protected]

Dr Edgar Kansime (Team [email protected]

Team 11 1 Leader) MEEPP m
2 Edson Niwagaba NA [email protected]
Adjumani,
3 Dr Alex Odama Arua RPMT [email protected]
Maracha,
4 Maria Nandago MEEPP [email protected] Moyo,
Nebbi,
5 IP Representative -
Zombo
6 District Representative - (9 sites)
7 USG Representative -
[email protected]
8 Dr George Upenytho NTLP m
Central Coordinating Team

1 Dr Mugerwa M Juliana MEEPP [email protected]

2 Dr Sarah Nabukera MEEPP [email protected]

[email protected]
3 Dr George Upenytho NTLP m

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 99
 DQA
Districts
Team # Name Organization Email address

4 Ms Jennifer Balaba MEEPP [email protected]

norah
5 Dr Norah Namuwenge Lead Consultant [email protected]

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 100
Annex 5: District Rating by M&E Component Assessed for TB/HIV System
Understanding of Availability of
M&E Structure, Indicator Data-collection
Functions and Definitions and and Reporting Data Use of data for Overall
Capabilities to handle Reporting Forms for TB/HIV Management decision District
District Facility HIV/TB information Guidelines services Processes making Score
Adjumani 2.7 2.7 2.6 2.6 2.6 2.6
Ciforo HC III 3.00 2.6 2.5 2.7 1.0 2.3
Agago 2.9 2.8 3 1.8 3 2.7
Patongo HC III 3.00 3.0 3.0 2.4 3.0 2.9
Apac 3 2.8 3 1.4 2.4 2.5
Alenga HC III 2.83 3.0 2.3 2.1 1.0 2.3
Teboke HC III 3.00 2.9 2.8 1.8 1.0 2.3
Buikwe 2.7 3 3 2.6 2.8 2.8
Nyenga Hospital 3.00 3.0 3.0 3.0 3.0 3.0
SsiBukunja 2.83 2.4 3.0 2.4 2.0 2.5
Bukedea 2.6 3 2.6 3 2.0 2.6
Bukedea HCIV 2.83 3.0 2.8 2.6 2.6 2.8
Kidongole HCIII 2.33 2.8 2.7 1.6 1.6 2.2
Bukomansimbi 2.9 2.7 3 1.6 2.4 2.5
Butenga HC IV 2.33 2.6 3.0 2.3 1.3 2.3
Bundibugy 2.7 3 2.8 1.4 1 2.2
Busaru HC IV 2.83 2.7 3.0 2.8 1.0 2.5
Ntandi HC III 2.83 2.7 3.0 2.4 1.0 2.4
Busia 2.6 2.7 2.2 2 3 2.5
Buhehe HC III 2.67 3.0 2.5 2.4 2.0 2.5
Lumino HC III 2.50 2.6 2.8 1.7 1.0 2.1
Butaleja 2.7 2.7 2.2 1.6 1 2
Budumba HC III 2.00 2.6 2.3 1.3 1.2 1.9
Butambala DTLS not available
Ngando HC III 2.83 2.8 2.8 1.0 1.3 2.2
Buvuma 2.3 2.7 3 2.6 3 2.7
Buvuma HCIV 3.00 2.4 2.7 2.4 3.0 2.7
Buyende 2.6 2.7 2.6 2.3 2.6 2.6
Nkondo HCCIII 2.67 2.8 2.7 2.4 2.3 2.6
Dokolo 2.7 2 2.6 1.4 2.6 2.3
Bata HC III 3.00 3.0 2.8 2.0 3.0 2.8
Kangai HC III 2.50 2.4 2.3 1.9 1.6 2.1
Gulu 2.7 3 3 2.3 3 2.8
Awach HC IV 3.00 3.0 2.7 2.6 3.0 2.9
Lapainat HC III 2.83 3.0 2.3 2.1 2.0 2.5
Homa 2.7 2.7 2.6 1.8 2 2.4
Kyangwali HC IV 2.17 2.4 2.3 1.9 1.8 2.1
Ibanda 3 3 2.6 2.3 3 2.8
Nyamarebe HCIII 2.83 2.8 3.0 1.9 1.8 2.5
Isingiro 2.7 2.8 3 1.7 3 2.7
Kabuyanda HCIII 3.00 3.0 3.0 2.7 2.6 2.9
Kashumba HC III 1.83 2.7 2.7 2.0 2.4 2.3
Kyabirukwa HCIII 2.83 2.6 3.0 2.3 1.8 2.5
Jinja 2.7 2.7 2 2.3 3 2.5
Buwenge HCIV 3.00 2.8 2.7 2.3 1.8 2.5
Lukolo HCIII 2.67 2.7 2.5 2.4 1.3 2.3
Mpambwa HCIII 2.67 1.4 2.0 1.9 1.0 1.8
KabaleE 2.9 2.7 2.8 1.4 2.2 2.4
Bufundi HC III 2.67 2.8 2.7 2.4 1.6 2.4
Rubaya Hospital 3.00 2.9 2.8 2.5 2.4 2.7
Rugarama Hospital 3.00 3.0 3.0 2.8 2.8 2.9
Kabarole 3 2.8 2.6 2.4 3 2.8

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 101
 Understanding of Availability of
M&E Structure, Indicator Data-collection
Functions and Definitions and and Reporting Data Use of data for Overall
Capabilities to handle Reporting Forms for TB/HIV Management decision District
District Facility HIV/TB information Guidelines services Processes making Score
Kataraka HC IV 2.67 2.7 3.0 2.2 2.8 2.7
Kalangala 2.9 2.7 3 1.4 3 2.6
Kalangala HC IV 2.83 2.9 2.7 2.4 2.2 2.6
Kaliro 2.9 2.7 2.6 1.7 3 2.6
Namwiwa HC III 3.00 2.8 3.0 2.0 3.0 2.8
Kalungu 2.9 2.7 3 1.4 1 2.2
Kiragga HC IV 2.83 2.2 3.0 2.1 1.0 2.2
Kampala
Central Div 2.7 3 2.4 3 3 2.8
MakerereHosp 2.33 1.2 1.8 1.1 1.0 1.5
Makindye Dizv DTLS not available
Mukwaya General
hospital 2.67 2.8 3.0 2.0 1.4 2.4
Kamuli 2.7 2.7 3 1.7 3 2.6
Kamuli Hospital 2.67 2.6 3.0 2.4 2.4 2.6
Kitayunjwa HC III 2.17 2.2 2.8 2.3 2.0 2.3
Kanungu 2.6 2.7 2.6 1.7 2 2.3
KirimaA HC III 2.33 2.3 2.7 1.9 1.0 2.0
Rugyeyo HC III 2.67 2.8 2.8 2.1 1.2 2.3
KASESE 2.6 3 2.6 2.2 1.2 2.3
Kagando Hospital 3.00 2.7 3.0 2.7 3.0 2.9
Kasese Town Council HC
III 2.83 2.7 2.7 2.6 2.6 2.7
Kayunga 3 3 3 3 2.8 3
Baale HC 4 2.67 2.8 3.0 2.6 2.2 2.7
Ntenjeru HC III 2.50 3.0 3.0 2.3 1.2 2.4
Wabwoko 3.00 2.7 3.0 2.4 3.0 2.8
Kibaale 2.4 2.5 2.6 2.2 2.8 2.5
Kibaale HC IV 3.00 2.9 2.7 2.8 2.6 2.8
Kiruhura 3 3 3 2.7 3 2.9
Family Health Resource
Centre 3.00 3.0 3.0 2.6 3.0 2.9
Kiryandongo DTLS not available
Kiryandongo Hospital 2.83 2.9 3.0 2.3 2.8 2.8
Kisoro 2.3 2.8 3 1.6 2.6 2.5
Mutolere Hospital 3.00 2.9 2.8 2.8 3.0 2.9
Kitgum 3 2.8 3 1.8 3 2.7
St. Josephs' Hospital
Kitgum 3.00 3.0 3.0 2.6 3.0 2.9
Kole 2.7 3 3 2.2 2.2 2.6
Aboke HC IV 2.83 2.8 2.7 2.1 3.0 2.7
Kyegegwa 2.9 2.7 3 1.4 1.8 2.4
Wekomiire HC III. 3.00 2.7 3.0 2.4 1.8 2.6
Kyenjojo 2.9 3 2.8 2 2.8 2.7
Kigaraale HC III 2.83 2.9 2.8 2.1 2.8 2.7
Luuka 2.9 2.7 2.6 1.7 2.5 2.5
Ikumbya HC III 2.83 3.0 3.0 2.4 2.0 2.7
Luweero 2.9 2.7 2.8 2.5 1.8 2.5
Bishop Asilli HCIV 3.00 2.9 2.8 2.8 3.0 2.9
St. Luke Namaliga HCIII 3.00 2.7 2.5 2.1 1.4 2.3
Lwengo 2.9 2.7 3 1.4 3 2.6
Kinoni HC III 2.00 2.9 2.5 2.3 1.0 2.1
Kyazanga HC IV 3.00 2.3 3.0 2.3 1.0 2.3
Manafwa 2.9 3 3 2.4 3 2.8

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 102
 Understanding of Availability of
M&E Structure, Indicator Data-collection
Functions and Definitions and and Reporting Data Use of data for Overall
Capabilities to handle Reporting Forms for TB/HIV Management decision District
District Facility HIV/TB information Guidelines services Processes making Score
Magale HC IV 2.33 2.6 2.7 2.1 1.0 2.1
Maracha 3 2.5 3 2.3 2.6 2.7
Ovujo HC III 2.83 2.6 2.7 2.8 1.4 2.4
Masaka 2.6 2.7 3 1.4 1 2.1
Kiyumba HC IV 3.00 2.8 2.8 2.0 1.0 2.3
Masaka Regional Referral
Hospital 2.83 2.4 2.8 2.8 3.0 2.8
St. Joseph's Hospital
Kitovu Mobile 2.50 2.4 2.8 2.1 1.0 2.2
Mayuge 2.9 2.8 2.8 2.1 2.4 2.6
Kityerera HC IV 2.67 3.0 3.0 2.3 3.0 2.8
Mbale 3 3 2 1.7 2.8 2.5
Bufumbo HC IV 2.83 2.7 2.8 2.0 2.6 2.6
Nyondo HC III 2.67 2.9 2.5 1.7 1.4 2.2
Taso Clinic Mbale 2.50 2.9 2.8 2.9 3.0 2.8
Mbarara 2.7 3 2.8 3 3 2.9
Buganba HCIV 2.67 3.0 3.0 2.3 2.6 2.7
Ruharo Mission Hospital 3.00 3.0 3.0 3.0 3.0 3.0
Moyo 3 3 3 1.8 3 2.8
Laropi HC III 3.00 3.0 3.0 3.0 3.0 3.0
Metu HCIII 3.00 3.0 2.7 2.5 2.6 2.8
Mubende 2.3 2.7 3 2.3 1.4 2.3
Kalonga HC III 2.83 2.8 2.7 1.6 3.0 2.6
MubendeRehabiitation
HCIII 2.83 1.9 2.7 2.5 1.0 2.2
Mukono
Kojja HC IV 2.50 2.8 3.0 2.5 1.6 2.5
SeetaNazigo HCIII 2.50 2.3 3.0 2.8 3.0 2.7
Nakasongola 2.7 3 2.4 2.4 1.5 2.4
Bamugolode 2.67 2.8 2.8 1.9 1.2 2.3
Nakitoma 2.83 3.0 3.0 2.0 1.2 2.4
Namayingo 3 2.7 3 1.6 2.5 2.5
Mutumba HHCIII 2.33 2.4 3.0 2.3 3.0 2.6
Nebbi 2.1 3 3 2.1 1 2.3
Angal Hospital 3.00 3.0 3.0 2.9 3.0 3.0
Kucwiny HCIII 3.00 3.0 3.0 2.6 3.0 2.9
Nwoya 3 2.8 3 2.2 3 2.8
Alero HC III 3.00 2.9 3.0 2.3 3.0 2.8
Pader 2.6 2.8 2.8 1.8 3 2.6
Pader HC IV 3.00 3.0 3.0 2.3 3.0 2.9
Pajule HC IV 3.00 3.0 2.7 2.4 3.0 2.8
Pallisa 2.7 3 2.4 2.6 2.3 2.6
Apopong HC III 3.00 2.7 2.2 2.0 1.0 2.2
KabwangasiI HCIII 2.33 2.9 3.0 1.6 1.5 2.3
Rakai 2.9 3 2.8 1.4 1 2.2
Kalisozo Hospital 2.67 2.3 2.5 2.7 1.0 2.2
Rakai Hosptal 2.33 2.8 2.7 2.6 1.2 2.3
Rubaga Division 2.3 2.7 1.8 1.7 3 2.3
Kitebi HCIV 3.00 3.0 2.8 2.9 3.0 2.9
LubagaHosp 3.00 3.0 3.0 3.0 3.0 3.0
Rukungirir 2.6 2.7 2.8 1.4 2.8 2.4
Buhunga HC IV 2.67 3.0 3.0 1.9 2.0 2.5
Nyakibale Hospital 2.33 3.0 3.0 2.4 3.0 2.8
Serere 2.7 3 3 2.8 2.8 2.9

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 103
 Understanding of Availability of
M&E Structure, Indicator Data-collection
Functions and Definitions and and Reporting Data Use of data for Overall
Capabilities to handle Reporting Forms for TB/HIV Management decision District
District Facility HIV/TB information Guidelines services Processes making Score
Kyere HCIII 2.67 2.7 2.7 2.4 1.3 2.3
Shema 2.9 2.5 3 2.1 3 2.7
Shuuku HCIV 3.00 2.8 3.0 2.5 2.6 2.8
Sironko 2.4 3 3 2.3 2 2.5
Buteza HCIII 2.33 2.4 2.2 2.1 1.6 2.1
Soroti 2.9 2.5 3 1.9 1.5 2.3
Gweri HCIII 2.50 2.8 3.0 2.3 1.8 2.5
Soroti Regional Referal
Hospital 3.00 2.7 3.0 2.8 3.0 2.9
Tubur HC III 2.83 2.4 3.0 2.8 2.3 2.7
Tororo 3 2.8 3 2.1 3 2.8
Merikit HC III 2.33 2.1 2.5 1.6 1.4 2.0
Tororo General Hospital 2.67 2.8 3.0 2.4 2.2 2.6
Wakiso 2.4 2.7 2.4 1.7 3 2.4
Bweyogerere HCIII 2.67 2.3 2.2 2.3 2.8 2.5
Kasangati HCIV 2.33 3.0 2.8 2.4 3.0 2.7
Zombo 2.1 3 3 2.1 1 2.2
Paidha HC III 3.00 2.4 3.0 2.7 2.6 2.7
Pakadha HCIII 3.00 3.0 2.3 2.8 1.6 2.5
Zeu HC III 2.67 2.3 2.7 2.5 2.4 2.5

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 104
 Annex 6: Deviation in DHIS 2 vs. Manual Recounts for Seven Variables
No. of HIV+ No. of new and Total no. of No. registered
adults and No. HIV+ relapsed TB registered No. of HIV+ TB cases with
No. PLHIV children in adults and cases with new and TB cases that documented
screened Pre-ART children in documented relapsed TB started or HIV-positive
District Site Level for TB care ART care HIV status cases continue ART status
Adjumani Ciforo HC III 18% 13% 16% 15% 7% 0% 0%
Agago Patongo HC III -4% 29% 72% -45% 9% -50% -56%
Apac Teboke HC III -5% 0% 29% 13% 38% 0% 25%
Apac Alenga HC III 5% 0% 6% 0% 0% -100% -100%
Buikwe Ssi Bukunja HC III 7% 37% 13% -20% -20% 0% 0%
Buikwe Nyenga Hospital 0% 0% 0% -7% -7% -55% -13%
Bukedea Kidongole HC III -4% 0% 0% 50% 50% 0% 0%
Bukedea Bukedea HCIV 1% 0% 1% 67% 67% 200% 50%
Bukomansimbi Butenga HCIV -4% 26% -6% -30% 261% -20% 20%
Bundibugyo Ntandi HC III -2% 0% -2% OR 100% 0% 0%
Bundibugyo Busaru HCIV -4% 9% -7% -50% -54% 0% 0%
Busia Lumino HC III -9% -27% 36% 0% 83% -50% 100%
Busia Buhehe HC III 65% 485% 30% -100% -100% -100% 100%
Butaleja Budumba HC III 0% 93% 5% -40% -100% 0% 0%
Butambala Ngando HC III -8% -24% -4% -11% -33% -17% -17%
Buvuma Buvuma HCIV 10% OR 7% 57% 286% -75% 0%
Buyende Nkondo HC III 71% 371% 2% 0% 0% 0% 0%
Dokolo Kangai HC III -88% -92% 12% 167% 217% 100% 33%
Dokolo Bata HC III -97% 9% -45% 16% 79% -11% -20%
Gulu Lapainat HC III -2% -7% 0% 0% 0% 0% OR
Gulu Awach HCIV -5% -4% -5% -45% 0% -100% -83%
Hoima Kyangwali HCIV -1% -15% 4% 17% 9% 0% -25%
Ibanda Nyamarebe HC III -20% 0% -20% 0% OR 0% OR
Isingiro Kashumba HC III -100% -100% -100% 13% 13% -100% 0%
Isingiro Kabuyanda HC III -82% 239% 43% -50% -50% -67% -60%
Isingiro Kyabirukwa HC III 13% 14% 21% 450% 1400% 200% -100%
Jinja Mpambwa HC III 8% -20% 20% OR OR OR OR
Jinja Lukolo HC III -8% -8% -50% -75% -50% 0% 0%
Jinja Buwenge HCIV 23% -42% 39% 5% 19% 0% 25%
Kabale Bufundi HC III -100% -100% -100% 0% 0% 0% 0%
Kabale Rubaya Hospital -31% 74% -39% 17% 0% 0% 0%
Kabale Rugarama Hospital -3% -6% -2% -17% -17% -40% -40%
Kabarole Kataraka HCIV 8% 0% -1% 0% 0% 0% OR
Kalangala Kalangala HCIV OR OR 8% 47% 47% 38% 38%
Kaliro Namwiwa HC III -12% 27% -21% 0% 0% 0% 0%
Kalungu Kiragga HCIV -98% 854% 4% 0% 0% 0% 0%
Kampala Kitebi HCIV 0% 0% 44% 43% 3% 79% 69%
Kampala Mukwaya Gen Hospital 33% -45% 38% 0% 27% 0% 25%
Kampala Lubaga Hospital 33% 84% 4% 6% 7% 10% 0%
Kampala Makerere Hospital 6% 13% -1% -33% -33% 0% 0%
Kamuli Kitayunjwa HC III -8% -51% 8% 50% 150% 0% 0%
Kamuli Kamuli Hospital 56% 42% 61% 3% 3% 0% 0%
Kanungu Rugyeyo HC III 40% 18% 49% 0% 100% 0% 100%
Kanungu Kirima HC III 52% 0% 52% OR OR 0% 0%
Kasese Kasese TC HC III -7% -11% -5% 15% 23% 0% 20%
Kasese Kagando Hospital -6% -16% -5% -32% 63% -27% -31%

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 105
 No. of HIV+ No. of new and Total no. of No. registered
adults and No. HIV+ relapsed TB registered No. of HIV+ TB cases with
No. PLHIV children in adults and cases with new and TB cases that documented
screened Pre-ART children in documented relapsed TB started or HIV-positive
District Site Level for TB care ART care HIV status cases continue ART status
Kayunga Ntenjeru HC III 3% -15% 17% 500% 1000% OR 100%
Kayunga Wabwoko HC III -97% -13% -100% -50% 0% -100% -50%
Kayunga Baale HCIV -5% -100% -3% 19% 44% 20% 14%
Kibaale Kibaale HCIV -18% -5% -1% 21% 29% -50% -9%
Kiruhura Family Health HC III 110% 332% 63% -38% -14% -38% -38%
Res Centre
Kiryandongo Kiryandongo Hospital -13% -15% -15% 17% 33% 0% 18%
Kisoro Mutolere Hospital 0% 0% 0% 14% -5% 23% 7%
Kitgum St. Josephs' Hospital 39% 3% 13% -15% 8% -29% -31%
Kitgum
Kole Aboke HCIV -4% -84% -11% 33% 55% 57% 129%
Kyegegwa Wekomiire HC III -5% -19% -2% 0% -20% -33% 0%
Kyenjojo Kigaraale HC III 0% 9% -2% 33% 33% 50% 50%
Luuka Ikumbya HC III -11% 10% -17% 0% 36% 0% 0%
Luwero St. Luke HC III 82% 53% 85% 40% 120% 100% 100%
Namaliga
Luwero Bishop Asilli HCIV 17% 132% 2% -6% 17% -25% -19%
Lwengo Kinoni HC III 28% OR 18% -5% -5% 0% 18%
Lwengo Kyazanga HCIV -3% 3% -4% 0% 0% 9% 0%
Manafwa Magale HCIV 4% -49% -3% -7% 75% 33% 25%
Maracha Ovujo HC III 233% 0% 40% OR OR 0% 0%
Masaka Kiyumba HCIV 26% 356% 6% 8% -33% 20% 20%
Masaka St. Joseph's Hospital -49% 26% 25% 4% 12% -50% -50%
Kitovu Mobile
Masaka Masaka RRH -38% -14% 58% -12% -12% -41% -18%
Mayuge Kityerera HCIV 21% 31% 17% 0% 9% 0% 0%
Mbale Taso Clinic Clinic- -7% -1% 2% -32% 16% -24% -47%
Mbale Special
Mbale Nyondo HC III 0% -100% 0% OR OR 0% 0%
Mbale Bufumbo HCIV -83% 526% -100% -29% -71% -50% -67%
Mbarara Bugamba HCIV 46% -6% 56% -56% -30% -100% -100%
Mbarara Ruharo Mission Hospital 57% -16% 74% -60% 20% -50% 25%
Moyo Laropi HC III -1% 0% -1% 0% 0% 0% 0%
Moyo Metu HC III 0% -57% 10% 0% 29% -50% -50%
Mubende Kalonga HC III 23% -20% -12% -50% 0% OR OR
Mubende Mubende HC III 380% 232% 523% 44% 56% 0% 75%
Rehabilitation
Mukono Seeta Nazigo HC III -89% -49% -100% 125% 100% 150% 250%
Mukono Kojja HCIV 11% 291% -9% 0% -6% 100% 11%
Nakasongola Nakitoma HC III -6% 0% -6% 233% 500% 200% 800%
Nakasongola Bamugolode HC III -44% OR -53% 164% 100% 50% 125%
Namayingo Mutumba HC III -15% -72% 1% 0% 33% -50% 0%
Nebbi Kucwiny HC III 426% 19% -5% OR OR 0% OR
Nebbi Angal Hospital 2% 50% 2% 8% 6% -27% -12%
Nwoya Alero HC III 4% -19% -15% 0% 1200% -100% -100%
Pader Pader HCIV -42% -59% -40% 130% 130% 700% 200%
Pader Pajule HCIV 7% -22% 11% 0% 0% -20% 0%
Pallisa Apopong HC III -17% -43% -100% 0% -100% 0% 0%
Pallisa Kabwangasi HC III 41% -23% 113% 0% 0% -50% -50%

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 106
 No. of HIV+ No. of new and Total no. of No. registered
adults and No. HIV+ relapsed TB registered No. of HIV+ TB cases with
No. PLHIV children in adults and cases with new and TB cases that documented
screened Pre-ART children in documented relapsed TB started or HIV-positive
District Site Level for TB care ART care HIV status cases continue ART status
Rakai Kalisizo Hospital 1% 1% 0% 5% 6% -10% -35%
Rakai Rakai Hospital 1% 27% 1% -22% -9% -29% -71%
Rukungiri Buhunga HCIV 11% 78% -1% 40% 40% 33% 33%
Rukungiri Nyakibale Hospital 16% -24% 18% -7% 2% 11% 7%
Serere Kyere HC III -7% -3% -8% 750% 800% 0% -100%
Sheema Shuuku HCIV 74% -41% 50% 256% -11% 200% 171%
Sironko Buteza HC III 33% 0% 200% 50% 650% 0% 0%
Soroti Gweri HC III 9% -3% 14% 33% 33% 50% 50%
Soroti Tubur HC III 12% 37% 5% 0% 0% 0% 0%
Soroti Soroti RRH -9% 31% 6% 6% 5% 9% 6%
Tororo Merikit HC III 48% 3171% 23% -20% -20% -50% -50%
Tororo Tororo Hospital 0% 0% 0% 14% 13% -31% -13%
Wakiso Kasangati HCIV -3% -19% -6% -6% -6% 0% 0%
Wakiso Bweyogerere HC III -41% -23% -4% 65% 47% 300% 33%
Zombo Zeu HC III 24% 13% 28% 75% -11% OR OR
Zombo Pakadha HC III 9% -55% -80% 100% 233% 100% 200%
Zombo Paidha HC III 249% 35% 1% -36% 14% 25% 38%

Quality of TB/HIV Data Reported by Health Facilities in Uganda: Findings from DQA Assessment, August 2015 107