Clinical Audit on the Provision of Diabetes Care in the Primary Care Setting by United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA) ()
1. Introduction
UNRWA has worked in the Near East for 64 years, providing health, education and social services for over 5 million Palestine refugees in Jordan, Lebanon, Syria, the West Bank and Gaza Strip. Due to the prevailing socioeconomic conditions, psychological stress and lack of access to healthy food and physical activities, non- communicable diseases (NCDs) such as, diabetes (DM), hypertension (HT), coronary artery disease (CAD) and cerebrovascular disease (CVD) are major problems for Palestine refugees The number of patients registered with DM has increased progressively and doubled in the last decade reaching 108,000 patients in 2012, without the data from Syria [1] . The exact prevalence of diabetes among Palestine refugees is not known, but may be assumed to be similar to the general population of the countries where they live, [2] i.e., 10.1% in Jordan, 7.8% in Lebanon, 8.6% in Palestine and 10.8% in Syria [3] ; hence, likely to be between 8% and 11% among adults 20 - 79 years of age. Around 11.0% of people ≥ 40 years attending UNRWA health facilities have DM [1] .
UNRWA has provided care for DM and HT at their primary health care centres since 1992 and the NCD strategy has been revised four times with the latest revision in 2009 [4] . The current strategy is structured around three main elements―healthy life style promotion emphasizes the importance of weight control and regular exercise; early detection of diabetes by active screening of at risk individuals; and treatment protocols and effective case management with dietary management, physical exercise and risk assessment and screening for cardiovascular, cerebrovascular and peripheral vascular disease to prevent secondary complications. Another important element of the strategy is to improve record keeping, program effectiveness and quality of services. UNRWA has started to roll out an E-health initiative which allows periodic cohort monitoring and evaluation [5] - [8] . To improve care delivery, periodic assessment of knowledge, attitudes and practices of care providers using validated questionnaires and data collection tools supplemented with clinical audit is useful [9] - [14] , the aim of this cross-sectional observational study was to assess the quality of diabetes care in the UNRWA primary care centres. In addition the study also assessed the UNRWA medical officers’ ability to deliver DM care by evaluating their knowledge, attitude and practice regarding patients and services, which will be presented in another paper.
2. Methods
2.1. Study Design
Cross-sectional prospective survey of patients with DM.
2.2. Setting
The study was conducted in the 32 largest PHCs, with 8 each in the following locations Gaza, Jordan, Lebanon and the West Bank. Syria was not included due to the on-going armed conflict. In 2012, UNRWA had 116 PHC clinics in these four Fields serving a population of 3,134,732 refugees. Each clinic is staffed by 2 to 6 doctors and a variable number of nurses. All screening, diagnosis and treatment services at the clinic are provided free- of-charge.
Palestine refugees who attend the clinics are screened for DM if they are ≥40 years old, at risk of non-com- municable diseases, or pregnant or planning to get pregnant. Blood glucose tests are done by laboratory technicians and diagnosis is confirmed by medical officers if the fasting blood glucose (FBG) is ≥126 mg/dl on 2 separate occasions [4] [15] . If the readings are between 100 - 125 mg/dl, a 75 g oral glucose tolerance test (OGTT) is performed to confirm or exclude diabetes. If FBG results are ≤100 mg/dl the patient is checked again in the following year. In the absence of facilities for testing for autoimmune markers for type 1 diabetes, children and young adults presenting with classical symptoms of diabetes―rapid loss of body weight with polyphagia, polyuria and polyhydria with or without ketonuria in the presence of random blood glucose ≥ 200 mg/dl or fasting glucose ≥ 126 mg/dl and requiring insulin treatment to control hyperglycemia are classified as type 1 DM. Persons diagnosed with DM are clinically assessed for co-morbidities and complications and these data along with demographic and clinical information are recorded in either patient registration files (hard copy) or in the E-Health system in health centres implementing electronic medical records. At registration, information on risk factors such as smoking, alcohol intake, physical activity and obesity (defined as body mass index ≥ 30 kg/m2) are recorded.
Patients are managed according to a standard algorithm defined in the UNRWA TI, [4] with diet and lifestyle advice and different classes of anti-diabetic drugs that include OADs namely Glibenclamide, Gliclazide and Metformin and Insulin injections. Patients with uncontrolled DM are seen weekly, fortnightly or monthly until their 2 h PPG value is ≤180 mg/dl and subsequently followed every 3 months. Once a year all DM patients undergo blood tests for total serum cholesterol and creatinine, urine examination for macro-proteinuria and dilated fundoscopic eye examination During quarterly visits, weight, blood pressure (BP), PPG value, as well as presence of complications (defined as blindness, end-stage renal failure, myocardial infarction (MI), congestive cardiac failure, stroke, and above-ankle amputation) are recorded.
2.3. Patient Population, Sampling Method and Size
There were 114,911 DM patients receiving care in UNRWA health centres at the end of 2011 [16] . A sample size of 1600 patients (50 from each selected health centre namely 400 from each field) i.e., 14% of the total diabetes population in the UNRWA health system was considered sufficient to provide a representative sample. The first fifty consecutive patients visiting the selected clinic on assessment days with confirmed DM and receiving care at the same UNRWA NCD clinic for at least one year, were included in the study after they gave a written informed consent. Recruitment at the clinic was stopped on reaching the target of 50 patients.
2.4. Source of Data, Variables, Reporting Formats and Analysis
Patient data were obtained from paper and/or E-Health records and through direct interviews and recorded on the pre-tested DCF [9] [10] modified slightly to accommodate UNRWA TI and procedures. Variables included in the DCF were DM type, risk factors, weight/height/waist measurements, blood pressure, and prior year's medical records including documented lab tests, complications, information on self-monitoring, health education and current medication.
The DCF was completed in the patient’s presence. To remove bias and ensure consistency across the study sites all patients were interviewed and clinically examined by staff responsible for NCD care at UNRWA Headquarters and field offices: Chief of Disease Prevention and Control and the Field Disease Control Officer, respectively. A local staff nurse or a trained NCD nurse assisted with the examination, and the local laboratory technician withdrew blood samples and facilitated the transportation for HbA1c testing at a non UNRWA central lab at the Augusta Victoria Hospital in East Jerusalem. HbA1c test was done using the High-performance liquid chromatography (HPLC) method (HLC®-723G8 Tosoh Corporation, Japan). Data were transferred into a computer and analysed with Epi-info 2000.
2.5. Ethics Approval
Approval for the study was obtained from UNRWA Headquarters and the four Fields. Patients gave written consent for participation in the study.
3. Results
The characteristics of the 1600 patients included in the audit are shown in Table 1. The mean age (SD) of the patients was 56.6 (12.6) years, ranging between 2 to 90 years. 37.5% patients with type 2 diabetes were <55
![]()
Table 1. Study population characteristics and key findings.
years. Two thirds of the entire patients were female (63.8%), which is in line with the gender distribution of patients attending UNRWA health centres (63.7%). Duration of DM treatment was used as a surrogate marker for DM duration and it varied considerably, ranging from 1 - 30 years, with a mean (SD) duration of 7.2 (4.9) years. People with type 1 DM (8.1 (4.9) years and type 2 DM with hypertension 7.8 (5.0) years) had longer treatment duration compared to type 2 DM without hypertension (5.9 (4.3) years).
Type 1 DM was seen in 68 (4.2%) cases, with a slight excess of females―37 to 31 males. Due to the selection criteria no case of gestational DM was included. Co-morbid hypertension was seen in 1102 (68.9%), all of them with type 2 DM. This data is similar to that seen in the entire UNRWA health system DM cohort in 2011.
Being overweight or obese was common in the study population, the mean (SD) BMI was 32.1 (6.4). Overall 90.3% were either overweight (26.3%) or obese (64.0%). Obesity was more common among women―73.1% compared to 47.9% among men. Smoking was more prevalent among men (34.1%) than women (11.3%). Few patients gave history of alcohol consumption.
In terms of record keeping and monitoring, data on age, gender, BMI, type of DM, duration and type of DM treatment, frequency and topic of health education sessions, and information on self-monitoring were available in all 1600 cases. Records of 2 h PPG, fasting serum cholesterol, serum creatinine and urine protein values in the previous one year were available in 94.7%, 96.4%, 91.4% and 87.5% cases respectively.
The most common treatment was OAD; 68.2% (1091) and 14.4% (231) were receiving OADs alone or in combination with insulin respectively. Insulin alone was used in 12.9% (207) cases and non-pharmacological lifestyle treatment (diet and exercise) alone in 3.9% (63) cases. Eight cases while attending the UNRWA health system were receiving treatment from outside. Of the 1322 cases on OADs, 89.3% (1181), 46.5% (615) and 18% (238) were receiving metformin, glibenclamide, and gliclazide alone or in combination respectively. Among subjects on insulin 72.5% were receiving < 60 IU/day and 77.8% were self-injecting. SMBG was reported by 32.6% of all and 66.2% with type 1 DM.
The mean (SD) 2 h PPG value for 1515 cases where information was available was 200.2 (74.5) mg/dl, with the range from 65 to 632 mg/dl. Table 2(a) and Table 2(b) show glycaemic control in relation to type of DM and type of treatment. In 44.4% cases the last recorded 2 h PPG was ≤180 mg/dl. The proportion of cases with acceptable control was similar for different types of DM and was highest among those on non-pharmacological treatment (81%), followed by those on OAD alone (47.9%); OAD in combination with insulin (31.6%) and insulin alone (30.9%)―a reflection of disease duration and severity.
(a) ![]()
(b) ![]()
Table 2. (a) Glycaemic control in relation to type of DM; (b) Glycaemic control in relation to treatment.
*Type of treatment not known in 8 cases.
The mean (SD) HbA1c was 8.3% (1.9). Using an HbA1c with cut off value of 7.0% as adequate control, only 27.9% cases were found to be under control, much lower than the 44.4% indicated by the last measured 2 h PPG value. When comparing results of HbA1c test with the last measured 2 h PPG values it was noted that while 86% of subjects with 2 h PPG ≥ 180 mg/dl had HbA1c values ≥ 7%, 56% of subjects with 2 h PPG ≤ 180 mg/dl also had HbA1c values ≥ 7%.
Results for serum cholesterol estimations were available in 96.4% (1543) cases; the mean fasting value (SD) was 191.8 (49.1) mg%. Hypercholesterolemia (≥200 mg%) was seen in 39.8% (614) cases. Results of serum creatinine test were available for 91.4% (1534) cases. The mean (SD) value was 0.865 (0.44) mg%. Elevated creatinine (≥1.2 mg%) was seen in 6.4% of all cases, 4.9% of type 1 DM cases, and 3.7% of Type 2 DM cases without hypertension but was twice as frequent in cases with hypertension - 7.5%. Urinary protein (macro albuminuria) estimation was done in 87.5% cases and was found raised in 10.3% cases.
Fundoscopy results were available for 757 cases (47.3%). UNRWA ophthalmologists had examined 408 (53.9%) and ophthalmologists from outside UNRWA had examined the remaining 349 (46.1%). There were 83 (11%) cases of retinopathy amongst those who had undergone fundoscopy.
The results of annual foot examination are shown in Table 3. Peripheral neuropathy was documented in 52.6% cases and was equally distributed among those with and without control based on the 2 h PPG results. Skin infections, both mycotic and bacterial, were documented in 17% cases.
There were 204 (12.8%) patients in the study population with late complications, 186 had one and 18 had two complications. No case had more than two late complications. Details of late complications are shown in Table 4.
4. Discussion
The clinical audit of the UNRWA Non Communicable Disease (NCD) Care Programme, with a focus on diabetes care validated the general UNRWA approach to health service delivery and confirmed its capacity to manage diabetes care in a primary health care setting. At the same time, it conducted an in depth assessment of protocols, procedures and performance in NCD care, documenting in particular the strengths of UNRWA diabetes care and more importantly, highlighting some critical shortcomings that will help define priorities for further improvement.
The clinical audit confirmed that UNRWA medical providers working in diabetes care generally follow the TI
![]()
Table 3. Results of foot examination.
![]()
Table 4. Late diabetes complications.
rigorously. In terms of record keeping and monitoring, data on most parameters were available in over 90% cases including records of lab tests and clinical examinations performed in the last one year which is better than in similar studies done in other developing countries in tertiary care centres, albeit some years ago [10] [17] .
Several shortcomings were identified some of which relate to financial constraints and the consequent effects on UNRWA TI and policy related to diabetes care. Only 45% of people with diabetes had 2 h PPG values within acceptable control (≤180 mg/dl) as defined by UNRWA TI. The control based on HbA1c testing (≤7.0%) was even lower at 28%, with the lowest control found in patients receiving insulin treatment (7.4%) and this is a cause of concern. Relying on 2 h PPG done only during clinic visit could be misleading. Non availability of HbA1c tests to monitor control due to cost constraints means that UNRWA has systematically overestimated its quality of DM control.
Comorbid hypertension was present in 69% (1102) cases and all of them were receiving antihypertensive medications. UNRWA TI defines blood pressure ≤ 140/90 mm Hg as the control target even for people with diabetes with co morbid hypertension. This level is higher than currently recommended International Diabetes Federation (IDF) and American Diabetic Association (ADA) target of ≤130/80 mm Hg for people with diabetes. Only 55.5% of cases with co-existing hypertension had their last recorded BP ≤ 140/90 mm Hg and only 28.2% had their BP ≤ 130/80 mm Hg. Hypertension control rates in this study are lower compared to the cumulative cohort data from six clinics in Jordan that use the E health record system where 87% of all patients had BP ≤ 140/90 mm Hg [7] . Uncontrolled hypertension in the setting of relatively poor glucose control is fertile ground for macro and microvascular complications particularly myocardial infarction, stroke, nephropathy and retinopathy. The fact that raised serum creatinine was seen twice as often in type 2 patients with co morbid hypertension and that 9.3% of cases had documented MI despite the relatively younger age, low prevalence of smoking and higher female gender mix in the study population points towards the need for more aggressive control of BP.
UNRWA TI defines the control target for total serum cholesterol at <200 mg/dl or <6.5 mmol/L. Cholesterol levels were elevated in 39.8% (614) cases. Because of financial constraints UNRWA cannot provide free treatment for hyperlipidaemia and as a consequence only half the patients (53.4%) with raised cholesterol were on lipid lowering drugs, most of them paying out of pocket. Including free supply of lipid lowering drugs, in particular, a statin and paying greater attention to lifestyle counselling will be necessary to address the risk of cardiovascular disease.
Insufficient focus on lifestyle counselling was another shortcoming identified. More than 90% of people with DM within the UNRWA system are either obese (64.0%) or overweight (26.3%). Around 70% have co existing hypertension and almost 40% have hyperlipidaemia as noted above and all these conditions are amenable to prevention through lifestyle measures; requiring greater attention to lifestyle counselling. According to UNRWA TI, patients should receive at least four health education sessions during assessment visits each year. The audit revealed that 17.6% patients received no self-care education and only 361 (22.6%) received four or more health education sessions, reflecting poor adherence to guidelines. Less than half the patients (40.6%) recalled receiving relevant lifestyle health education sessions (exercise and diet). Although foot complication rates are significant (e.g. 53% peripheral neuropathy and 13% foot infections), only 16% of patients recalled receiving foot care advice. The same applies for counselling on hypoglycaemia, even though 20% - 30% of patients were receiving insulin treatment. Better training, redefining roles and skilful deployment of non-medical health professionals may help improve self-care education and patient counselling.
The most prevalent early complication among patients was peripheral neuropathy seen in 52.6%. History of MI or undergoing angioplasty was noted in 9.3% cases and is similar to the 9.7% prevalence of MI among diabetic patients from the region reported earlier [18] . The high rate of MI reflects the need to address underlying risk factors and to strictly follow UNRWA’s secondary prevention strategy for people with DM and hypertension with stricter monitoring of BP and glycaemia control using HbA1c and introducing the use of statins.
The findings of this study are almost identical to those reported from tertiary care centres in other developing countries using a similar study protocol [10] [17] [19] - [22] . This indicates that in general people with type 2 DM have multiple comorbid conditions―overweight and obesity, hypertension, and dyslipidaemia. Control of hyperglycaemia, hypertension and hyperlipidaemia is less than satisfactory and complications are high. In many of these studies, the proportion of patients undergoing annual laboratory tests and examinations were much lower indicating poor adherence to protocols; level of control for glucose, blood pressure and lipids was lower and complication rates higher.
The high prevalence of overweight or obesity and other risk factors in the study population points towards a need for a more comprehensive and strategic response that goes beyond the activities of the NCD care programme alone to address such fundamental issues and the recently applied Family Health Team reform [23] offers an ideal reference framework with focus on persons and families to address early prevention of disease by addressing family and community risk factors, early detection through systematic screening and provision of comprehensive care.
While more efforts are required to raise awareness and improve lifestyles through health promotion; the prevailing socioeconomic and psychological stress from the security situation in refugee camps poses a big challenge. Limited funding and need to prioritise scarce resources also places constraints on policy recommendations.
5. Conclusion
In conclusion, UNRWA’s extensive experience in DM care in primary health care settings and the capacity, experience and rigour of their medical providers are a solid foundation on which to improve DM care. This study provides a basis to guide further actions aimed at modernizing and broadening DM care and address priorities for improvement by the UNRWA Health Department.
Acknowledgements
Special thanks to World Diabetes Foundation for financial and technical support to conduct the audit. We thank all UNRWA staff at the audited health centres and Field Control Officers for their cooperation in managing, follow up and monitoring patients.
Author Contributions
Anil Kapur, Yousef Shahin and Akihiro Seita designed the study and wrote the first draft which was further developed by other authors. All authors contributed to the writing of the subsequent drafts and revisions of the paper, and all authors have read and approved the final paper for submission.
NOTES
*Corresponding author.