International Journal for Quality in Health Care 2009; Volume 21, Number 5: pp. 356– 362 10.

1093/intqhc/mzp030
Advance Access Publication: 14 August 2009

Evaluation of HIV/AIDS clinical care

quality: the case of a referral hospital
in North West Ethiopia
YIBELTAL KIFLIE ALEMAYEHU1, OLUMA YOSEPH BUSHEN2 AND AYALEW TEGEGN MULUNEH1
1
Jimma University, PO Box 378, Jimma, Ethiopia, and 2Tulane University Technical Assistance Program – Ethiopia, PO Box 844, code
1250, Addis Ababa, Ethiopia

Abstract
Objective. To assess the quality of clinical care provided to patients with HIV in Felege Hiwot Referral Hospital.
Approach and design. Normative evaluation based on Donabedian’s structure – process – outcome model of health care
quality. Cross-sectional study design was employed to gather data in September 2007.
Setting. Felege Hiwot Referral Hospital is a government hospital in North West Ethiopia. The hospital is providing clinical
care for patients infected with HIV free of patient charge since 2005.
Measures. The evaluation used 10 process and 5 outcome indicators of quality measured by reviewing 351 randomly selected
patient records and interview with 368 patients. Resource inventory was conducted to assess the availability of trained staff,
laboratory facilities and drugs required for provision of HIV care.
Results. All resources recommended by the national antiretroviral therapy (ART) Implementation Guideline including trained
staff, laboratory facilities and drugs were continuously available, except for a shortage of cotrimoxazole. Despite this, impor-
tant components of care and treatment recommended by national treatment guidelines were not delivered for significant
portion of patients. The study showed that only 45.9% of patients eligible for cotrimoxazole prophylactic therapy (CPT) and
76.8% of patients eligible for ART were actually taking CPT and ART, respectively. Compliance with national guidelines to
monitor patients was also found to be a major problem.
Conclusion. Availability of resources alone does not ensure the quality of HIV care and treatment. The study results indicate
a need for regular monitoring and improvement of processes and outcomes of care in the Ethiopian Health System.
Keywords: quality measurement, quality improvement, quality indicators

Introduction but also on appropriate execution of recommended com-

ponents of care including prevention and treatment of
Like in other Sub-Saharan African Countries, HIV/AIDS is opportunistic infections, use of Highly Active Antiretroviral
a major public health, social and economic problem in Therapy (HAART) and laboratory and clinical follow-up of
Ethiopia. As part of the multi-sectoral response to mitigate patients [4 – 8].
the impact of the pandemic and through the support of Felege Hiwot Referral Hospital is one of the three referral
PEPFAR and Global Fund, the government is providing hospitals in Amhara National Regional State of Ethiopia.
medical care services to People Living with HIV/AIDS free The hospital uses the national guidelines for the management
of charge [1 – 3]. of opportunistic infections and ART to manage
Time and again, researches have revealed that medical care HIV-infected patients [4, 5]. By the end of August 2007, a
prolongs and improves the quality of life of patients infected total of 6396 patients were registered for HIV care out of
with HIV/AIDS. However, the outcomes of clinical care for which 5031 (78.7%) started ART. Out of those patients who
patients depend not only on attending a clinical care provider started ART, 3225 (64.1%) were reported as taking

Address reprint requests to: Yibeltal Kiflie Alemayehu, Jimma University, PO Box 378, Jimma, Ethiopia. Fax: þ251 471
110969/þ251 471 117839; E-mail: [email protected]

International Journal for Quality in Health Care vol. 21 no. 5

# The Author 2009. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/
licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is
properly cited. 356
 Quality of HIV care in Ethiopia

antiretroviral drugs actively in the hospital while the remain- patients in the two categories of care. In addition, an exit
ing 1806 (35.9%) failed to follow-up, transferred out to other interview was conducted with 368 patients visiting the HIV
health facilities or died [9]. clinic during the data collection period to assess satisfaction
To date, in Ethiopia, there exists neither an established of patients. Interviews were conducted by trained data collec-
system of quality management nor a comprehensive study tors using a satisfaction questionnaire adapted from the
assessing the quality of HIV clinical care. But some studies HIVQUAL project of New York State Department of
conducted outside Ethiopia showed that quality of medical Health AIDS Institute [17], which included 29 satisfaction
care is indeed a major problem that needs to be addressed measuring variables against a five-level satisfaction score.
systematically [10 – 14]. A specific study conducted to assess SPSS version 12.0.1 software was used to analyze the data.
adherence of patients with HAART in Addis Ababa also
showed that adherence to antiretroviral drugs is a problem
even though it is better when compared with that of patients
Inclusion and exclusion criteria
in more urban settings [15].
Patients who have been in care and treatment for at least 6
months were eligible for record review. With the purpose of
assessing recent practices, patients who had no clinic visit
Objective during the 1-year period prior to data collection were
excluded. Patients who transferred in or out to other health
To assess the quality of clinical care provided to patients
facilities were also excluded from the study to keep focus on
infected with HIV in Felege Hiwot Referral Hospital.
quality practice in the study hospital. For the patient interview,
patients who visited the clinic for the first time were excluded
as they may not have adequate prior experience with the hos-
Indicator selection pital to provide meaningful answers to the questions.
Quality of care indicators were set based on national guide-
lines for the treatment of patients with HIV/AIDS in
Ethiopia [4, 5] after a thorough review on the quality of
Results
HIV/AIDS care literature. The selected indicators were again
Resource availability
reviewed by physicians working on HIV/AIDS control
program as well as HIV care-providing facilities for their HIV clinical care in Felege Hiwot Referral Hospital is pro-
local relevance. After all these procedures, a total of 15 indi- vided in a separate clinic. Daily, the ART unit is staffed with
cators were selected. Ten indicators were on the process of two general practitioners and seven nurses who took part in
care while the rest five indicators were on outcomes of care. in-service trainings on HIV/AIDS clinical care provision
based on the national guideline. The HIV clinic uses labora-
tory services from two sources: the hospital laboratory
Measurement department and the Regional Health Research Laboratory,
which is located in the hospital’s compound. The hospital’s
Donabedian’s structure – process – outcome model of health care laboratory provides routine laboratory tests, including HIV
quality was employed as a framework for the study [14, 16]. testing, microscopy for TB diagnosis and stool examination.
Data on the structure, processes and outcomes of care were For other HIV/AIDS-related laboratory tests including CD4
collected cross-sectionally in September 2007. The measures measurement, liver function test, renal function test
included: and hematology analysis, it uses the Regional Research
(1) Availability of resources required to provide HIV clini- Laboratory. All the required laboratory tests, based on the
cal care service National ART Implementation Guideline, are available in either
(2) Compliance of HIV clinical care practice with national the hospital’s laboratory unit or the regional laboratory.
guidelines During the 6-month period prior to data collection, the
(3) Outcomes of care including behavioral, immunologic hospital’s drug store lacked stock of the first line antiretro-
and clinical conditions viral drugs, Stavudine and Zidovudine. The cumulative
Resource inventory was conducted to assess the availability period without stock of these drugs was for 2 and 8 days,
of resources, including trained staff, laboratory facilities, anti- respectively. But the drug supply was not interrupted in the
retroviral drugs and opportunistic infection prophylactic dispensing unit. There was no stock of cotrimoxazole in
drugs. An inventory tool prepared based on the resource either the store or the dispensary for 2 months prior to and
requirements of the national program for health facilities pro- during the data collection period.
viding ART was used for this purpose.
The process and outcome indicators were assessed by
Background characteristics of study subjects for
reviewing 351 randomly selected patient records. Of the total
chart review
records reviewed, 233 (66.4%) were patients on ART and the
remaining 118 (33.6%) were on chronic care before ART Table 1 presents the background characteristics of study
initiation, which was based on the actual proportion of population. The review included 351 records of patients. The

357
Alemayehu et al.

Table 1 Background characteristics of study subjects for presented in Table 2 and findings revealed major problems
chart review, Felege Hiwot Referral Hospital, September in adherence with national guidelines for HIV clinical care
2007 provision.
According to the national guidelines, 267 (76.1%) of the
Variable Category Number Percent 351 sample patients were eligible for ART. Among these,
.................................................................................... only 205 (76.8%) were actually taking ART while 34 (12.7%)
Sex Male 159 45.3 of them never started ART and the rest 28 (10.5%) failed to
Female 192 54.7 continue treatment after initiation. Of the 118 patients who
Total 351 100.0 were in care before ART initiation, 34 (28.8%) were actually
Age (completed ,5 9 2.6 eligible for ART but had not yet started it.
years) 5– 14 14 4.0 Out of the 351 patients included in this study, 236
15 328 93.4 (67.2%) had a most recent CD4 count of less than 350
Total 351 100.0 cells/ml that makes them eligible for CPT according to the
Marital status Married 97 27.6 national guidelines. The result showed that only 179 (75.8%)
Never married 34 9.7 of these eligible patients were prescribed CPT during their
Widowed 28 8.0 last HIV clinic visit. Furthermore, the clinic was able to
Divorced 50 14.2 maintain continuity of care only for 146 (81.6%) of those
Not documented 142 40.5 patients who were prescribed with CPT during their last visit.
Total 351 100.0 Exit interviews with patients attending the HIV clinic
Educational No education 92 26.2 showed that only 74.3% of patients prescribed with CPT
status Primary education 29 8.3 during their previous clinic visit took their prophylactic
Secondary education 70 19.9 drugs. Continuity of care was assessed against this standard
Tertiary education 18 5.1 and was found to be maintained for only 249 (70.9%)
Not documented 142 40.5 patients. It was 88.0% for patients taking ART and 37.3%
Total 351 100.0 for patients in care before ART initiation.
Occupation Civil servant 26 7.4 According to the national guideline on management of
Farmer 12 3.4 patients with HIV infection [4], a patient with HIV is
Housewife 19 5.4 expected to visit HIV clinic at least once every 6 months
Merchant 19 5.4 prior to ART initiation and every 3 months after ART
Daily labor 39 11.1 initiation. Continuity of care was assessed against this stan-
Others 24 6.8 dard and was found to be maintained for only 249 (70.9%)
Not documented 212 60.4 patients. It was 88.0% for patients taking ART and 37.3%
Total 351 100.0 for patients in care before ART initiation.
Address (Zone) Bahir Dar (The city 200 57.0
where the hospital is Behavioral and health outcomes
located)
Outside of Bahir Dar 151 43.0 Achievement of desired behavioral and health outcomes was
Total 351 100.0 measured by the level of patients’ adherence to prescribed
On HIV Yes 233 66.4 drugs, immunologic status of patients, clinical improvement
treatment / No 118 33.6 (improvement in functional status of bedridden or ambulat-
ART Total 351 100.0 ory patients) and satisfaction of patients. Table 3 presents
the findings on specific outcome indicators of care and
treatment.
median duration in care and treatment was found to be 15.5 Adherence to antiretroviral drugs and to CPT was
months (IQR: 12.13 months). One hundred twenty-one measured by interviewing patients on the number of doses
(34.5%) were enrolled within 1 year prior to the data collec- missed during 1-week period prior to the date of interview.
tion period while the remaining 230 (65.5%) have been fol- If a patient was taking 95% or more of prescribed pills, this
lowing care for more than 1 year. was defined as good adherence. The result from the inter-
view showed that 338 (97.7%) of patients on ART and 182
(65.2%) of those on CPT had good adherence to their
Compliance of HIV clinical care with national
respective prescribed drugs. Among respondents of the inter-
guidelines
view who were prescribed with CPT, it was found that 72
Compliance with national guidelines [4, 5] for the provision (25.8%) of them were not taking the drugs at all.
of HIV clinical care was assessed based on 10 selected Out of the 331 patients who had their CD4 count
process indicators on 351 patient records, of which 233 measured at least once, first-time CD4 count was greater
(66.4%) were on ART and the remaining 118 (33.6%) were than 200 cells/ml for 150 (45.3%) of the patients. After
patients in care prior to ART initiation. The performance of varying period in care and/or treatment, most recent CD4
the hospital in regard to the selected process indicators is count was greater than 200 cells/ml for 200 (60.4%) of

358
 Quality of HIV care in Ethiopia

Table 2 Summary of performance on process indicators of quality, Felege Hiwot Referral Hospital, September 2007

Processes indicators Number of eligible Achievement,

patients P (95% CI)
.............................................................................................................................................................................

Care during enrollment to HIV clinic

Proportion of patients who received their first-time CD4 count within 2 weeks 121 90.9% (85.8 – 96.0)
after first HIV clinic visit
Opportunistic infection prevention
Proportion of patients whose CD4 count is ,350 cells/ml who currently are 236 45.9% (39.5 – 52.3)
on cotrimoxazole prophylactic therapy
ARV therapy and follow-up
Proportion of patients with continued care 351 70.9% (66.1 – 75.7)
Proportion of patients who had CD4 count measured at least once during the 351 20.8% (16.6 – 25.0)
last 6 months
Proportion of patients eligible for ARV who are currently on ARV 267 76.8% (71.7 – 81.9)
Proportion of patients on ART who are started on ART within 2 weeks after 233 86.3% (82.5 – 91.1)
clinical eligibility is confirmed
Proportion of patients on NVP who had LFT at least once within 1 month 179 0.0%
after initiation of NVP-based ART
Proportion of patients on ART for whom adherence is measured on last three 218 28.0% (22.1 – 33.9)
visits
Record completeness
Proportion of patients with previous ARV regimen change for whom reason 26 53.8% (35.3 – 72.3)
for change in regimen is documented
Proportion of patients on ARV who had registered own and contact address 351 41.9% (36.7 – 47.1)

Table 3 Summary of performance on outcome indicators of

quality, Felege Hiwot Referral Hsopital, September 2007

Outcome indicators n Achievement,

P (95% CI)
....................................................................................

Adherence to medications
Proportion of patients on 346 97.7% (96.1 – 99.3)
ARV with at least 95% (good)
reported adherence on last visit
Proportion of patients on 279 65.2% (59.6 – 70.8)
cotrimoxazole prophylaxis with
at least 95% reported
adherence on last visit
Immunologic and clinical response to HIV care Figure 1 CD4 count of patients on HAART across time
Proportion of patients with 331 60.4% (55.1 – 65.7) after ART initiation, Flege Hiwot Referral Hospital,
CD4 count greater than 200 September 2007.
cells/ml
Proportion of either 92 71.7% (62.5 – 80.9)
bedridden or ambulatory them. Among patients who were taking ART, CD4
patients who have measurements at baseline, 6, 12 and 18 months after ART
improvement in functional were found documented for 199, 69, 53 and 23 patients,
status respectively. The result of this study showed an increase in
median CD4 count with duration on ART. Fig. 1 presents
Patient satisfaction
the trend in median CD4 count with duration on ART. As
Average of patient 368 78.0% (73.8 – 82.2)
shown in the figure, the median baseline CD4 count
satisfaction score
was found to be 275 cells/ml (IQR: 157, 403 cells/ml) at

359
Alemayehu et al.

Table 4 Summary of satisfaction levels by dimensions of missing important components of care recommended by
satisfaction, Flege Hiwot Referral Hospital, September 2007 the national guidelines. Although there were important
improvements in many patients’ immunologic and clinical
Dimensions of Average Number of Mean conditions, achievement of these desired patient outcomes
patient number of questions satisfaction needed further improvement. The presence of all the
satisfaction respondents score required resources, except an observed shortage of cotrimox-
.................................................................................... azole, to provide HIV care reveals a vast opportunity for the
Availability of 345 7 3.4 hospital to improve process and outcome indicators of
resources quality. A relatively better quality of care was provided to
Perceived quality 317 7 4.0 patients on enrollment in care with 90.9% of the patients
of care received their first CD4 count within 2 weeks after first HIV
Health worker– 320 8 4.4 clinic visit. For the remaining components of care expected
patient interaction to be provided as follow up care, the hospital failed to
Accommodation 312 5 4.0 provide recommended care for significant portion of its
and time HIV-infected patients.
management A number of studies have shown that CPT prevents the
Affordability of 365 2 3.2 occurrence of fatal opportunistic infections among patients
care with HIV infection [18 – 21]. World Health Organization and
Average of all 327 29 3.9 UNAIDS also recommended the use of CPT for patients
indices with symptomatic HIV infection [22]. Considering these rec-
ommendations, the national CPT guideline has made a rec-
ommendation to prescribe CPT for HIV-infected patients
with CD4 count of less than 350 cells/ml [5]. In this respect,
18 months of treatment when compared with that of
the finding from this evaluation study showed that out of the
baseline CD4 count which was 127 cells/ml (IQR: 74, 191
sample patients who were eligible for CPT, only about 45.9%
cells/ml).
were actually on CPT. This considerably low achievement
Patients’ self-reported functional status was one of the
was because of failure to prescribe CPT for eligible patients
patient monitoring measures recommended by the national
during clinic visits, lower rate of continuity of care and poor
treatment guideline. Working, ambulatory and bed ridden
compliance of patients with CPT.
were the categories used for this purpose. Among the study
Timely initiation of appropriate ART is one of the basic
subjects, functional status was assessed and documented
components of chronic HIV care which determines the
during enrollment to care for 246 (70.1%) of them. Out of
outcome of HIV clinical care. Many studies have proved that
these, 154 (62.6%) were working; 77 (31.3%) were ambulat-
antiretroviral therapy reduces morbidity and mortality among
ory and 15 (6.1%) were bedridden. Of the total 92 patients
patients with HIV [6, 7, 23, 24]. The Ethiopian standard on
who were either bedridden or ambulatory, 66 (71.7%)
the time to start ART is based on the World Health
showed improvement from ambulatory to working, bedrid-
Organization’s recommendation for resource-limited settings
den to ambulatory or bedridden to working.
[25]. According to the national ART guideline, patients need
On the basis of the patient satisfaction survey, the average
to be started on ART if they have clinical stage IV disease,
level of satisfaction was 3.9 out of 5 points (78.0%). Average
clinical stage III disease with CD4 count less than 350 cells/
rate of satisfaction was relatively higher for variables related
ml or those with CD4 count less than 200 cells/ml irrespec-
with provider– patient interaction for which the average satis-
tive of their clinical stage [4].
faction score was 4.4 out of 5 points. On the other hand, the
According to this study, the study hospital was able to
satisfaction score was relatively lower for affordability of care
keep 76.8% of eligible patients on ART. Nearly 13% of the
which was mainly related with the non-medical cost of visit-
eligible patients were never started on ART and the remain-
ing the HIV clinic. The average satisfaction score for this
ing 10.5% failed to follow-up after starting ART. This
dimension was 3.2 out of 5 points. Table 4 presents the
showed that there are HIV patients who are eligible but have
number of satisfaction questions under each satisfaction
not yet been put on ART. In concrete terms, more than a
dimension and the level of satisfaction of patients on each of
quarter (28.8%) of HIV-infected patients who were in care
satisfaction dimensions.
before ART initiation did not start ART although they
should have been started according to the recommendation
in the national treatment guideline. This finding appears to
Discussion be an issue for discussion among HIV/AIDS program man-
agers who are determined to expand the service to health
This evaluation study used an indicator-driven approach for centre level, with the aim of universal access to ART, prob-
the measurement of HIV care quality and has addressed ably without due attention to the quality of care in the exist-
three components of quality—structure, process and ing delivery sites.
outcome. The results show that a significant portion of Continuity of care was maintained for 88.0% for patients
patients attending HIV clinic in the study hospital were in care after ART initiation and 37.3% before ART initiation.

360
 Quality of HIV care in Ethiopia

Those patients who lost to follow-up require putting in place patients, respectively [32]. If this holds true in our setting,
a strong system of patient-tracing mechanism. Unfortunately, the use of standardized patients would have increased the
documentation of patients’ own and contact address was values of the measured process indices by up to 10%, but
poor; only 41.9% of patients had a complete address docu- most of the indicators would still fall below the expected
mented in their records, making tracing of patients difficult. level of performance.
Time and again, studies have shown that outcome of
HAART is strongly associated with adherence to antiretro-
viral drugs [26 – 30]. This study showed that good adherence Conclusions
to antiretroviral drugs was achieved in 97.7% of the patients.
This result was higher compared with a similar measure in a Felege Hiwot Referral Hospital has adequate resources to
more urbanized setting of Addis Ababa where the pro- provide HIV/AIDS clinical care. However, important com-
portion of patients with good adherence was 81.2% which ponents of care recommended by the national guidelines
was considered even better when compared with the situ- were found to be missed for significant proportions of
ation in the developed world [15]. However, adherence to patients. Although immunologic and clinical improvements
CPT was low; only 65.2% of patients prescribed with CPT were observed among patients attending the hospital, there is
had good adherence to CPT during the week prior to the still a huge gap to maximize benefits to all eligible patients.
data collection day and 25.8% were not taking the drug at all The presence of these gaps despite the availability of ade-
even though it was prescribed by their providers. quate human and material resources shows that availability of
The overall process of care for patients with HIV is to resources alone will not ensure the quality of clinical services
keep patients protected from opportunistic infections, by to patients with HIV infection signifying the need to con-
improving their immunological status [4]. A patient’s CD4 tinuously monitor and improve process and outcome
count less than 200 cells/ml is a strong marker for the poss- measures of quality.
ible occurrence of opportunistic infections [31]. The last
CD4 count among the sample HIV-infected patients was
found to be greater than 200 cells/ml in 60.4% (n ¼ 351) of References
them when compared with that of first-time CD4 count
which showed that only 45.3% of the patients had CD4 1. FMOH/FHAPCO. AIDS in Ethiopia: sixth report. http://
count of greater than 200 cells/ml. CD4 measures at base- www.etharc.org/aidsineth/publications/AIDSinEth6th_en.pdf
line, 6, 12 and 18 months after ART initiation also showed (12 December 2006, date last accessed).
improvement in immunologic status of patients with time on 2. World Health Organization. Progress on Global Access to HIV
ART. Despite this improvement, there is a huge gap for Antiretroviral Therapy: A Report on ‘3 by 5’ and Beyond. Geneva:
improvement to keep all patients with CD4 count of at least World Health Organization, 2006.
200 cells/ml. This finding should alert relevant stakeholders
to give emphasis on prophylactic therapies and work towards 3. UNAIDS. 2006 Report on the global AIDS epidemic. 2006
early enrollment of patients to care and treatment. 4. Federal HIV/AIDS Prevention and Control Office/Federal
Quality of care provided to patients was found to vary Ministry of Health. Guidelines for management of opportunis-
across different components of care. The best care provided tic infections and antiretroviral treatment in adolescents and
was during first time enrollment of patients during which adults in Ethiopia. http://www.etharc.org/arvinfo/
90.9% of patients received their first-time CD4 measurement ethOIARTguideline_2007.pdf (11 November 2007, date last
within 2 weeks of enrollment. For most of the rest care com- accessed).
ponents expected to be provided during follow-up, the 5. Federal MOH/Federal HAPCO. Guideline for cotrimoxazole
system was able to provide recommended care only for less prophylaxis in HIV/AIDS care and treatment. http://www.
than half of the patients. This finding is comparable to a etharc.org/arvinfo/cotromoxizoleguide.pdf (12 May 2006, date
study in the USA, which showed that patients are more likely last accessed).
to receive the recommended HIV care better during their 6. Palella FJ, Jr, oria-Knoll M, Chmiel JS et al. Survival benefit of
first visit than in follow-up care [11]. initiating antiretroviral therapy in HIV-infected persons in
The patient record was the main source of process indi- different CD4þ cell strata. Ann Intern Med 2003;138:620 –6.
cators used in this evaluation study. The accuracy and com- 7. Mocroft A, Lundgren JD. Starting highly active antiretroviral
pleteness of the patient record may result in either therapy: why, when and response to HAART. J Antimicrob
underestimating or overestimating some of the indicators. To Chemother 2004;54:10 –3.
date, there are no studies done in Ethiopia to compare the
8. Liu C, Weber K, Robison E et al. Assessing the effect of
reliability of methods to measure health care quality, so the
HAART on change in quality of life among HIV-infected
direction of bias introduced by inherent limitations of chart women. AIDS Res Ther 2006;3:6.
abstraction is not well understood in our context. However,
some studies outside the country have shown that using 9. FHAPCO. Monthly ART update as of September 10, 2007. 2007
chart review may underestimate quality indicators by a 10. Wilson IB, Landon BE, Marsden PV et al. Correlations among
margin of 5.4 and 10.6% when compared with measures measures of quality in HIV care in the United States: cross sec-
through the use of clinical vignettes and standardized tional study. Br Med J 2007;335:1085.

361
Alemayehu et al.

11. McGlynn EA, Asch SM, Adams J et al. The quality of health 22. World Health Organization, UNAIDS. Antibiotic recommen-
care delivered to adults in the United States. N Engl J Med dation for African HIV patients. 2000. 2007
2003;348:2635 – 45.
23. Enrico Girardia. Impact of combination antiretroviral therapy
12. Jencks SF, Cuerdon T, Burwen DR et al. Quality of medical care on the risk of tuberculosis among persons with HIV infection.
delivered to Medicare beneficiaries: a profile at state and AIDS 2000;14:1985 –91.
national levels. J Am Med Assoc 2000;284:1670– 6.
24. Phillips AN, Staszewski S, Weber R et al. HIV viral load
13. Schuster MA, McGlynn EA, Brook RH. How good is the response to antiretroviral therapy according to the baseline CD4
quality of health care in the United States? 1998. Milbank Q cell count and viral load. J Am Med Assoc 2001;286:2560 –7.
2005;83:843 –95.
25. World Health Organization. Scaling up antiretroviral therapy in
14. Donabedian A. An Introducation to Quality Assurance in Health resource limitted settings: treatment guidelines for a public
Care. New York: Oxford University Press, 2003. health approach. http://www.who.int/hiv/pub/prev_care/en/
arvrevision2003en.pdf, 2007.
15. Tadios Y, Davey G. Antiretroviral treatment adherence and its
correlates in Addis Ababa, Ethiopia. Ethiop Med J 26. Mannheimer SB, Matts J, Telzak E et al. Quality of life in
2006;44:237 –44. HIV-infected individuals receiving antiretroviral therapy is
related to adherence. AIDS Care 2005;17:10 –22.
16. Donabedian A. Evaluating the quality of medical care. 1966.
Milbank Q 2005;83:691– 729. 27. Paterson DL, Swindells S, Mohr J et al. Adherence to protease
inhibitor therapy and outcomes in patients with HIV infection.
17. New York State Department of Health AIDS Institute, Johns
Ann Intern Med 2000;133:21 –30.
Hopkins University. HIV clinical resource: adult and adolescent
indicators. http://www.hivguidelines.org/Content.aspx?pageID= 28. Hogg RS, Heath K, Bangsberg D et al. Intermittent use of
36 (15 July 2006, date last accessed), 2006. triple-combination therapy is predictive of mortality at baseline
and after 1 year of follow-up. AIDS 2002;16:1051 –8.
18. McNaghten AD, Hanson DL, Jones JL et al. Effects of
antiretroviral therapy and opportunistic illness primary chemo- 29. Bangsberg DR, Perry S, Charlebois ED et al. Non-adherence to
prophylaxis on survival after AIDS diagnosis. Adult/ highly active antiretroviral therapy predicts progression to
Adolescent Spectrum of Disease Group. AIDS AIDS. AIDS 2001;15:1181 –3.
1999;13:1687 –95.
30. Garcia de Olalla P, Knobel H, Carmona A et al. Impact of adher-
19. Grimwade K, Sturm AW, Nunn AJ et al. Effectiveness of cotri- ence and highly active antiretroviral therapy on survival in HIV-
moxazole prophylaxis on mortality in adults with tuberculosis infected patients. J Acquir Immune Defic Syndr 2002;30:105–10.
in rural South Africa. AIDS 2005;19:163 –8.
31. Hogg RS, Yip B, Chan KJ et al. Rates of disease progression by
20. Walker AS, Mulenga V, Ford D et al. The impact of daily cotri- baseline CD4 cell count and viral load after initiating triple-drug
moxazole prophylaxis and antiretroviral therapy on mortality therapy. J Am Med Assoc 2001;286:2568 –77.
and hospital admissions in HIV-infected Zambian children. Clin
32. Peabody JW, Luck J, Glassman P et al. Comparison of vignettes,
Infect Dis 2007;44:1361– 7.
standardized patients, and chart abstraction: a prospective vali-
21. Anglaret X, Chene G, Attia A et al. Early chemoprophylaxis dation study of 3 methods for measuring quality. JAMA
with trimethoprim-sulphamethoxazole for HIV-1-infected 2000;283:1715 –22.
adults in Abidjan, Cote d’Ivoire: a randomised trial. Cotrimo-CI
Study Group. Lancet 1999;353:1463 – 8.
Accepted for publication 19 July 2009

362