Scholars Journal of Applied Medical Sciences (SJAMS) ISSN 2320-6691 (Online)

Sch. J. App. Med. Sci., 2015; 3(2A):602-607 ISSN 2347-954X (Print)

©Scholars Academic and Scientific Publisher
(An International Publisher for Academic and Scientific Resources)
www.saspublisher.com

Research Article
Factors Influencing Adherence to Diabetes Medication in Turkey
Taşkaya Serap1*, Şahin Bayram2
1
Aksaray University, School of Health Collage, Department of Health Care Management, 68100, Turkey
2
Hacettepe University, Economics and Administrative Sciences, Department of Health Administration, 06800, Turkey

*Corresponding author
Serap Taskaya
Email: [email protected]

Abstract: The aim of this study was to investigate factors affecting medication adherence of patients with diabetes. A
cross-sectional study was conducted on patients with diabetes being treated with insulin and oral hypoglycemic agents.
Outpatients of a public hospital, which is located in Aksaray province in Turkey, are the universe of the study.
Questionnaire forms were used as the instruments for data collection in the study. The questionnaire was completed by
360 patients with diabetes during face to face interviews. Socio demographic (age, gendder, marital status, education,
insurance and income) and clinic variables (duration of diabetes, medication type, disease type, complication and
comorbidity), health belief, health literacy level, patient-physician relationship and drug adherence were determined for
each patient. The multiple regression analyses showed that education, type of diabetes, diabetes duration, health belief
and health literacy level had a statistically significant association with levels of medication adherence (p < 0.05) but there
is no relationship between patient-physician relationship and medication compliance. Thus, adherence to medication had
related to higher education levels, Type 2 diabetes, longer duration of disease, higher health beliefs and higher health
literacy level. Educational and biopsychological aspects of patient’s health are very important for diabetes adherence.
The results of the study are expected to provide important evidence-based information on factors affecting medication
adherence of diabetes to improve the diabetes management and war against disease for clinicians, health managers and
health workers.
Keywords: Diabetes, Factors, Adherence, Questionnaire

INTRODUCTION diagnosed with diabetes and prevalance of disease was

In this century, one of the most fundamental estimated 14.58% in 2013. In this report, it was found
issue related to health system is chronic diseases in both that prevalance of world diabetes is 8.3% and 6.8 % in
developed and developing countries. Complications, European countries [3]. As seen, Turkish people are
deaths and costs from these diseases are rising among more likely to be diagnosed with diabetes compared
worldwide. However scientific knowledge and with world and European countries. Besides this,
experience show that, chronic diseases and problems diabetes is the eighth leading cause of death and tenth
resulting from it can be reduced by implementation of leading couse of disability in this country [4].
evidence-based interventions and so, the management
and evaluation of chronic disease issues are getting In diabetes management, providing glycemic
more important in every society [1]. control plays a main role in care and achieving it
depends on patients adherence to the medical treatment
Diabetes is a chronic disease which affects faithfully [5]. Adherence to medication is defined as the
millions of peoples in the world and prevalence of it has extent to which the patient’s medication-taking
an increasing trend year by year. Much of this disease behavior corresponded with the prescribed medication
occurs in developing countries IDF. However it affects regimen [6], while non-adherence is described as taking
the all the socio economic status, age and gender less than 80% of the prescribed treatment [7].
althought a few preventable risk factors are responsible
for this condition such as obesity, unhealthy diet, When the patients are well suited to treatment,
physical inactivity. Diabetes is also one of the most better clinical results are achieved. For example
important challenges to health services due to the higher patient’s quality of life perception is positively affected
costs, higher mortality and morbidity rates [1-2]. and morbidity, mortality and medical costs are
declining [2-8]. However non-adherence to medication
In the Turkey, according to the International has been a common problem in the worldwide among
Diabetes Federation, nearly 7 million people have been chronic disease like diabetes. In this regard, adherence
602
Taşkaya Serap et al., Sch. J. App. Med. Sci., 2015; 3(2A):602-607

to medical treatment regimen is an extremely sensitive was taken from all the patients. An institutional ethical
subject for health professionals. committee approval to conduct the study was obtained
on February 10th 2014.
Above mentioned, because of the critical
importance of compliance to medical treatment, Data collection
determination of the factors that affect the status of Data were collected with a self-administered
adherence is required in order to ensure a good disease questionnaire was conducted among patients with
management. According to WHO, the factors diabetes. The questionnaire consisted of five parts:
influencing adherence to treatment that are identified socio-demographic and clinic information, Health
from the studies may be classifed to five categories Belief Model, patient-physician relationship, health
such as patient-centered factors (sociodemographic literacy, medication adherence.
factors, health, belief, health literacy, forgetfullnes etc.),
therapy-related factors (complexity of the medical Factors potentially associated with adherence
regimen, duration of treatment side-effects etc.), to diabetic medication include socio-demographic
healthcare system factors (poorly developed health variables and clinical characteristics such as age,
services, non-existent reimbursement, patient-physician gender, marital status, education, income, insurance,
relationship etc.), social and economic factors(poor medical history (duration of diabetes and duration of
socioeconomic status, poverty, unemployment, lack of medication), type of diabetes, medication type,
effective social support networks etc.) and disease complication and co-morbidities.
factors(severity of symptoms, level of disability,
comorbidity, complication etc.) [9-10]. At the second section, to determine diabetes-
specific beliefs, Health Belief Model (HBM) Scale were
Although examining factors influencing is a used. Original scale consisting of 76 questions had been
core issue, a few studies on this subject has been revised by Hurley [11] and reduced to 11 items. Scale
performed in developing countries just like Turkey. was divided into 3 domains as perceived high benefits
Conducting more studies in these countries could be (3 questions), perceived high seriousness (4 questions),
helpful to fill in the knowledge gap and to formulate and perceived low barriers (4 questions). HBM
worldwide strategies [10]. So this study was conducted incorporated a five-point Likert scale ranging from 1
for the purpose of determining the factors affecting “strongly disagree” to 5 “strongly agree”.
adherence to medication of patients with diabetes in
Turkey. Factors potentially associated with adherence Physician-patient relationship was measured
to diabetic medication are determined as socio- by Patient-Physician Relationship Questionaire with
demographic variables (age, gender, marital status, nine items which was created by Van der Feltz-Cornelis
education, income, insurance), clinical characteristics et al. [12] Original scale vas designed to measure in
(duration of diabetes and duration of medication, type primary health care doctor-patient relationship. But then
of diabetes, medication type, complication and co- by various researchers, it was used in the hospital and
morbidities), health beliefs, health literacy and patient- outpatient clinics [13]. Scale was rated as of "1" score
physician relationship. The results of the study are "Strongly Disagree", "5" score "Strongly Agree".
expected to provide important evidence-based
information for clinicians, health managers and health In the fourth section of the survey, to measure
workers. the health literacy of patients, Short Health Literacy
Screening Tool (Brief Health Literacy Screening Tool-
MATERIALS AND METHODS BRIEF) were used. Scale has 4 questions and
Patient selection incorporated a five-point Likert scale ranging from 1
A cross-sectional research design was used to “never” to 5 “always” [14].
study the adherence of patients to medical treatment.
Outpatients of a public hospital, which is located in Finally, adherence to treatment was assessed
Aksaray province in Turkey, are the universe of the through the self-report Morisky Medication Adherence
study. The sample size of 360 was calculated with a Scale (MMAS) [15]. The total scale consists of 4 items,
confidence level of 95 percent. The questionnaire was "Yes-No" is answered. But then Erickson et al. [16]
completed by patients with diabetes during face to face used a five-point Likert scale ranging from "1 = never 5
interviews between February to April 2014. The sample = always". In this study, we also used 5 point Likert
included both Type 1 and Type 2 patients with scale.
diabetes.
Internal reliability was examined for those
Patients who met the following criteria were questionnaries and all Cronbach’s alpha coefficients
invited to participate in this study: patients above 15 were greater than 0.7 (ranging from 0.73 to 0.97). With
years; patients who had been diagnosed with diabetes; respect to construct validity, factor analysis were
and patient who were on prescribed antidiabetic performed. Using the Kaiser criterion, we calculated the
medications for at least six months. Informed consent 3-factor solutions for Health Belief Model and one

603
Taşkaya Serap et al., Sch. J. App. Med. Sci., 2015; 3(2A):602-607

factor solution for each other scales (Patient-Physician demographic variables and clinical characteristics,
Relationship Questionaire, Short Health Literacy health beliefs, patient-physician relationships and health
Screening Tool and Morisky Medication Adherence literacy levels that might influence adherence to
Scale). It was the same with orginal scale solutions. treatment of patient with diabetes.

Data analysis A total of 360 patients with diabetes

The collected data were stored in an Excel participated in this study. Table 1 describes socio-
Statistical Package for Social Science (SPSS), version demographic and clinic variables of these 360
11.5 for analysis. Frequencies was were computed for participants 64.4% were female and most patients were
categorical variables and means and standard deviation married (73.3%) and insured (91.4%). Most patients
calculated for continuous data. Multiple regression had primary education and (54.2%) and only few
analyses was conducted to evaluate the factors affecting (6.6%) had university education. The mean age was 51
medication adherence. P-value < 0.05 was considered years (SD ±14.45). Of the participants only 10% (n =
statistically significant. 36) had a diagnosis of type 1 diabetes and the duration
of the disease varied between 1 year and 35 years (m=
RESULTS AND DISCUSSION 8.94, SD = 7).
In this study, itwas aimed to investigate
whether there were variables such as socio-

Table 1: Socio-demographic and clinic factors of patient with diabetes

Variables n (%) Variables n (%)
Gender Types of Diabetes
Female 232 64.4 Type 1 36 10
Male 128 35.6 Type 2 324 90
Marital status Tedavi Tipi
Married 264 73.3 İnsülin 148 41.1
Not Married 96 26.7 Oral hypoglycemic agents 212 58.9
Education *Complication
No formal education 108 30 No 289 80.3
Primary school 131 36.4 Yes 71 19.7
Middle and high 91 25.3 **Comorbitity
University 30 8,3 No 224 62.2
Insurance Yes 136 37.8
No 31 8.6 Mean S.D.
Yes 329 91,4 Dur. of Diabetes 8.94 7
Mean S.D. Dur. of Medication 8.51 6.9
Age 53,19 14,45 *Complication: coroner disease, eye disease ve nephropaty
**Comorbidite: hypertension, collesterol, celiac disease, depression,
Income (TL) 1056,81 663,03
asthma

Multiple regression analyses technique was Firstly it was found that patients' medication
used to examine the proportion of variance in a adherence levels did not depend on gender, marital
adherence to medication as explained by a set of status, insurance and income (p>.05). Also, medication
independent variables such as socio-demographic and type, comorbidite and complication did not affect
clinic factors, health belief, patient physician adherence to diabetes medication. By contrast, the level
relationship and health literacy. As recommended by of education, duration of diabetes and type of diabetes
Tabachnick and Fidell [17], the independent variables are important factors for compliance to treatment.
should not be strongly related to each other. In this According to this, people who have more education,
study, age and duration of medication were excluded longer diabetes duration and Type 2 diabetes has more
because of the highly intercorrelated with duration of compliant to medical treatment.
diabetes.
Another findings that were found after using
Table 2 shows regression analysis results. At multiple regression, health literacy level and health
the end of the multiple linear regression analyses, belief of patients were statistically significantly
several factors were found to have significant effects on associated with medication adherence. However
the medication adherence (R = 0.383; adjusted R2 = patient-physician relationship did not affect compliance
0.147; F = 12,160; p < 0.001). to diabetes medication. Beta values which used to the
contribution of each variables to medication adherence

604
Taşkaya Serap et al., Sch. J. App. Med. Sci., 2015; 3(2A):602-607

indicated that health literacy (β=.275) is the larger beta inflation factor (VIF) for multicollinearity were
coefficient than the other independent variables. between 1.022 and 1.269 ). In model, Durbin Watson
value was 1.874 that means there was no otocorrelation
Table 2 also shows no multicollinearity was problem.
found among the independent variables (variance

Table 2: Multiple regression analysis for adherence to medical treatment

Adherence to Medication
UnStd. β Std. β
Vaiables t p VIF
Coefficiency Coefficiency
Constant 1.191 2.476 .014
Health Literacy .209 .275 4.977 .000 1.269
Health Belief Model .273 .171 3.204 .001 1.186
Primary School -.297 -.124 -2.475 .014 1.049
Type of Diabetes .452 .118 2.313 .021 1.085
Duration of Diabetes .017 .104 2.094 .037 1.022
Adj. R2= 0.147 F=12.160 p=0.000 Durbin- Watson=1.874

At the end of the multiple regression analyses, Finally a significant higher rate of adherence to
educational situtiation had an positive influence on medication was observed in patients who exhibited
patients’ compliance to treatment. More educatued higher health belief compared with the others who did
patients showed higher rate of medical adherence than not have. This situation was thought that patients with
patients who had no formal education and graduated higher perceived seriousness, higher perceived benefits
from primary school. Because more educated people of treatment and the low perception of barrier were
tend to understand the consequences of non-adherence. more complient with treatment. Daniel and Messer [27],
It was found by some researchers to have better Gutierrez and Long [28] and Pourghaznein et al. [29]
adherence [18], while someone did not find a found that health belief was associated with adherence
relationship between education and adherence to to treatment.
medication [19].
We conclude that participants with higher
Another factor affecting adherence to medical healh belief and literacy level are more likely to have
treatment is type of disease. Being a patient with type 1 higher adherence to diabetes medications. Thus, low
was associated with decreased medication adherence literacy and health beliefs may be very important barrier
diabetes. Although some authors report that patients to achieving high rates of adherence to treatment. So
with type 1 diabetes was more compliant to medication educational and biopsychological aspects of patients
[20], others demonstrate patients with type 2 diabetes health are very important to improve treatment
showed better addherednce [21]. adherence. Patient should be educated about the
advantages of use of medical treatment regulary and
In this research, it was found out that there was common unappropriate beliefs must be eliminated.
a relationship betmeen duration of diabetes and Further research is also recommended in order to
adherence rate. Thus, a person which had longer identify interventions to improve treatment adherence in
diabetes history had more adherent than patient with patients with diabetes on a larger sample population.
shorter diabetes duration. These interrelationship were
consistent with Ayman et al. [18], and were inconsistent A major limitation of our study is that we
with Bezie et al. [22] and Arslan [23]. By contrast, collected data only at a single hospital. So results of
some researhers found there was no relationship study can not be generalized to the universe.
between duration of diabetes and adherence to medical Nevertheless, it can say that the results achieved, can
treatment [19]. reveal important clues regarding diabetes. Another
limitation of this research, questionnaire was presented
CONCLUSION in the hospital environment to respondents. Especially
Our results confirm that the most important the objective assessment regarding the patient-physician
factor contributing to adherence to medication was the relationship can be affected in the hospital environment.
level of patients’health literacy. Patients who had higher However, to avoid to this problem, the patient
health literacy were care and knew about their disease completed the survey when they received the services.
and it’s consequences ans so they were more compliant. Because of this reason, it was considered that it could
Some studies had found that adherence to medication reflects their actual ideas
was better when patient had higher health literacy [24,
25], but some studies had found there was no
relationship between them [26].

605
Taşkaya Serap et al., Sch. J. App. Med. Sci., 2015; 3(2A):602-607

ACKNOWLEDGMENT 15. Morisky DE, Green LW, Levine DM;

The author thanks Aksaray Hospital managers Concurrent and predictive validity of a
to give us permission to conduct this study. Also we selfreported measure of medication adherence.
would like to thanks to all patients who participated in Med Care, 1986; 24(1): 67-74.
this study. 16. Erickson SR, Coombs JH, Kirking DM, Azimi
AR; Compliance From Self-Reported Versus
REFERENCES Pharmacy Claims Data With Metered-Dose
1. Turkish Ministry Of Health; Türkiye Kronik İnhalers. The Annals of Pharmacotherapy, 2001;
Hastalıklar ve Risk Faktörleri Sıklığı Çalışması. 35(9): 997-1003.
Anıl Matbaa Ltd., Şti, Ankara, 2013: 3-4. 17. Tabannick BG, Fidel LS; Using Multivariate
2. Singh D; How can chronic disease management Statistics. 5th edition. NY: Harpercollins College
programmes operate across care settings and Publishers, New York, 2008: 61-91.
providers? World Health Organization 18. Ayman SA, Abd-El-Aziz AE; Degree of
Publication, Copenhagen, 2008: 1-4. compliance towards therapeutic tasks among
3. International Diabetes Federation; IDF Diabetes diabetic patients attending a health ınsurance
Atlas. 6th edition, IDF Publication, Belgium, setting ın Cairo. The Egyptian Journal of
2013: 58-59. Hospital Medicine, 2007; 27: 234– 244.
4. Turkish Ministry of Health;Türkiye Hastalık 19. Sweileh MW, Aker O, Hamooz S; Rate of
Yükü Çalışması. Aydoğdu Ofset, Ankara, 2006: compliance among patients with diabetes
8-15. mellitus and hypertension. An-Najah Univ J Res
5. Farsaei S, Sabzghabaee AM, Zargarzadeh AH, (N. Sc.), 2005; 19: 1-12.
Amini M; Adherence to glyburide and 20. Peyrot M, Rubin RR, Lauritzen T, Snoek FJ,
metformin and associated factors in type 2 Matthews DR, Skovlund SE; Psychosocial
diabetes in Isfahan, Iran. Iranian Journal of problems and barriers to improved diabetes
Pharmaceutical Research, 2011; 10(4): 933-939. management: results of the Cross-National
6. Wu JR, Moser DK, Chung ML, Lennie TA; Diabetes Attitudes, Wishes and Needs (DAWN)
Objectively measured, but not self-reported, study. Diabet Med., 2005; 22(10): 1379–1385.
medication adherence independently predicts 21. Ruggiero L, Spirito A, Coustan D, McGarvey
event-free survival in patients with heart failure. ST, Low KG; Self-Reported Compliance With
J Card Fail., 2008; 14(3): 203–210. Diabetes Self-Management During Pregnancy.
7. Delamater AM; Improving patient adherence. Int J Psychiatry Med., 1993; 23(2): 195-207.
Clinical Diabetes, 2006; 24(2): 71-77. 22. Bezie Y, Molina M, Hernandez N, Batista R,
8. Hagerstrom GE; Personal Factors, Perceptions, Niang S, Huet D; Therapeutic compliance: A
Influences and Their Relationship with prospective analysis of various factors involved
Adherence Behaviors in Patients with Diabetes. in the adherence rate in type 2 diabetes. Diabetes
Georgia State University. USA, 2010: 13-24. & Metabolism, 2006; 32(6): 611-616.
9. World Health Organization; Adherence to long- 23. Arslan E; Tip 2 Diabetes Mellituslu Hastaların
term therapies: Evidence for action. WHO Press, Tedaviye Uyumlarını Etkileyen Faktörler. Dicle
Geneva, 2003: 27-38. Üniversitesi. Diyarbakır, 2011: 37-38.
10. Jin J, Sklar GE, Oh VMS, Li SC; Factors 24. Cho YI, Lee SYD, Arozullah AM, Crittenden
affecting therapeutic compliance: A review from KS; Effects of health literacy on health status
the patient’s perspective. Therapeutics and and health service utilization amongst the
Clinical Risk Management, 2008; 4(1): 269–286. elderly. Social Science & Medicine, 2008; 66(8):
11. Hurley A; The health belief model; Evaluation of 1809-1816.
diabetes scales. Diabetes Educator, 1990; 16(1): 25. Gazmararian JA, Williams MV, Peel J, Baker
44-48. DW; Health literacy and knowledge of chronic
12. Van der Feltz-Cornelis CM, Van Oppen P, Van disease. Patient Educ Couns., 2013; 51(3): 267–
Marwijk HW, De Beurs E, Van Dyck R; A 275.
patient doctor relationship questionnaire (Pdrq- 26. Kim S, Love F, Quistberg DA, Shea JA;
9) in primary care: Development and Association of health literacy with self-
psychometric evaluation. Gen Hosp Psychiatry, management behavior in patients with diabetes.
2004; 26(2): 115–120. Diabetes Care, 2004; 27(12): 2980–2982.
13. Wenger HC, Chen HC, Chen HJ, Lu K, Hung 27. Daniel M, Messer LC; Perception of disease
SY; Doctors’ emotional intelligence and the severity and barriers to self care predict glysemic
patient– doctor relationship. Medical Education, control in aboriginal persons with type 2 diabetes
2008; 42(7): 703–711. mellitus. Chronic Diseases in Canada, 2002;
14. Haun J, Noland-Dodd V, Varnes J, Graham- 23(4): 130-138.
Pole J, Rienzo B, Donaldson P; Testing the 28. Gutierrez J, Long JA; reliability and validity of
BRIEF health literacy screening tool. Federal diabetes specific health beliefs model scales in
Practitioner, 2009: 24-31. patients with diabetes and serious mental illness.

606
Taşkaya Serap et al., Sch. J. App. Med. Sci., 2015; 3(2A):602-607

Diabetes Research and Clinical Practice, 2011;

92(3): 342-347.
29. Pourghaznein T, Ghaffari F, Hasanzadeh F,
Chamanzari H; The relationship between health
beliefs and medication adherence in patients
with type 2 diabetes: A correlation-cross
sectional study. using multivariate statistics,
2013; 10(4): 1-7.

607