1. Introduction

The global prevalence of Diabetes has already reached the epidemic level. According to predictions from the World Health Organization (WHO) developing countries bear the highest share of the diabetes epidemic. Currently, more than 70% of people with diabetes live in low- and middle income countries.¹ In 2011, the global prevalence of diabetes was estimated at 8.3%. Diabetes does not spare Africa. Although the current estimated prevalence in Africa is relatively lower (3.8%), the region is expected to experience the highest increase in its diabetes prevalence in the next two decades.²

African people with diabetes also suffer from poor access to quality diagnosis and treatment services contributing to high rates of complications and premature death.³ Delayed diagnosis increases the risk of complications and morbidity from diabetes exponentially. This situation is typical of the African region where only less than half of people estimated to have diabetes are actually diagnosed.⁴

In Ethiopia, the number of deaths attributed to diabetes reached over 21,000 in 2007. This estimate has increased to about 25,000 in 2011. For the wide majority of people with diabetes, medical care is provided at general outpatient clinics of hospitals. Due to the increase in the number of patients and better recognition of the problem, specialized diabetes clinics are being established only recently even in the capital. Most hospitals and health centers in and out of the capital provide diabetes care as part of their general outpatient services.⁵

However, how diabetes is being managed in Ethiopia is not well known and documented. A recent study conducted to assess the quality of care provided to patients with chronic non-communicable diseases (CNCDs) in South west Ethiopia has shown that there exists a huge gap in the quality of diabetes care. In that study, data was collected on multiple diabetes care quality indicators; however, the analysis aggregated multiple indicators and presented scores rather than performance on specific indicators.⁶ We further analysed these data on Diabetes management to better understand how diabetes is being managed in such resource constrained settings. More specifically, we looked at access to basic diabetes management services, diagnosis of diabetes, care and treatment of diabetes, and level of achievement of desired of glycemic control in the management of diabetes. The findings of the study are expected to be useful in designing and implementing further service improvement interventions with regard to the management of diabetes in Ethiopia and other similar settings.

2. Methods and Materials

2.1. Study Area and Period

The study on quality of care for patients with CNCDs in Southwest Ethiopia was conducted among people with Diabetes receiving follow-up clinical care in Jimma University Specialized Hospital and surrounding Health Centers in Jimma Zone.⁶ The data in this study, which is re-analysed in our current study, was collected between February and March 2010. Jimma zone is one of the 17 zones of Oromia National Regional State located in Southwest Ethiopia. The zone has a total population of 2.5 million.

2.2. Study Design

The design of the study was a health facility-based cross-sectional study involving retrospective review of records of people with diabetes in the study health facilities with Diabetes patients: Jimma University Specialized Hospital, Agaro Health centre and Asendabo Health Center.

2.3. Study Population

Our study population consisted of a representative sample of people with diabetes attending follow-up care in Jimma University Specialized Hospital and two surrounding Health Centers. At a larger scale, this population fairly represents people with diabetes attending clinical care in health facilities in resource constrained settings in general and Ethiopia in particular.

2.4. Sample and Sampling

We included data collected on 174 people with diabetes. The calculated sample was proportionally allocated to the study health facilities based on the number of diabetic cases attending those health facilities. Systematic random sampling technique was used to select medical records of patients. Patients enrolled to follow-up care at least six months before data collection period and who have at least one clinic visit during the one year review period were considered eligible for the study.

2.5. Data Collection

The procedure followed to develop the data collection tools and train data collectors is published in the methods section of the article on the quality of care for patients with CNCDs in Southwest Ethiopia.⁶ Data was collected by reviewing medical records of selected people with diabetes attending care in the study health facilities.

2.6. Data Analysis and Interpretation

We selected data collected on people with diabetes for further analysis. The data was analysed using SPSS for Windows Version 16.0. The data was analysed to assess the performance of study health facilities against specific process and outcome indicators of diabetes care. Numerical and graphic summaries are used to describe the findings. Chi-square tests, independent sample t-tests, Wilcoxon sign rank test, and bivariate correlation tests were used to measure the statistical significance of associations. P values less than 0.05 were considered to be statistically significant in all cases.

3. Results

3.1. Background Information

A total of 174 People with diabetes were included in this study. The majority of the cases, 114(65.5%), were attending diabetes follow-up care in Jimma University Specialized Hospital while the rest, 60 (34.5%), were attending Agaro Health Center and Asendabo Health Center. The majority, 121 (72.5%), of the study subjects were males. The median age of the study subjects was 50 years (IQR: 38, 60). One third of the cases were below the age of 40 years and 45% were between 40 and 60 years of age. Male people with diabetes had higher median age, 51 years (IQR: 38, 63), than their female counterparts with median age of 44 years (IQR: 32, 52).

3.2. Access to Health Facilities

Access to health facility was estimated using three proxy indicators: Distance from home to the health facility (in kilometres), time it takes to walk from home to facility (in hours) and public transport fee (in Ethiopian birr). The median distance from home of people with diabetes to the nearest health facility that provides diabetes care was found to be 10 kilometres (20 kilometres for the round trip) and ranges from 0.5 to 350kms. The median distance for those patients outside the town was 18 kilometres (36 kilometres round trip). About one fifth of the cases travelled more than 25 kilometres to reach to the nearest health facility and one tenth of the cases travelled more than 50 kilometres to reach to the nearest health facility.

The median time to walk from home to the nearest health facility was 2 hours (i.e. 4 hours for round trip) and ranged from 15 minutes to two and half days. For a quarter of the cases it took more than 10 hours to reach the nearest health facility. The median public transport fee for the distance from home to the nearest health facility was 5 Ethiopian birr (i.e. 10 birr for the round trip) and ranges from 75 cents to 150 birr. About 20% of the cases spend more than 10 Ethiopian birr to reach the health facility where they are receiving diabetes care.

To check for the consistency of the estimation, the distance from home to the health facility was correlated with the public transport fee. The correlation between these variables was very high (Spearman rho=0.96, p=0.000). The estimation of access to health facility was consistent between these measures. The findings also indicated that cases attended in the two Health Centers travelled less to reach to the facility as compared to those attended in Jimma University Specialized Hospital.

3.3. Diagnosis of Diabetes

The majority of the study subjects, 119 (68.4%) were diagnosed for Type 2 diabetes. The rest 55 (31.6%) had type I diabetes based on documented ketonuria and/or documented insulin use from the time of onset. In both types of diabetes, males outnumbered. Most of the cases with type I diabetes were being attended in Jimma University Specialized Hospital. The mean (SD) number of years since cases were diagnosed with diabetes was 5.4(4.3) years and it ranged from 6 months to 18.25 years.

Analysis of the age at the time of diagnosis showed that the average age at diagnosis for males and females was 45.2 years (95% CI: 42.1-48.3) and 36.8 years (95% CI: 32.1-41.5), respectively. The average age at diagnosis among people with Type 2 diabetes was 48.6(95% CI: 46.1-51.2) years.

Assessment of baseline laboratory investigation at the time of diagnosis showed that diabetes diagnosis in the study area was mainly based on measurement of fasting plasma glucose. This measurement was conducted in the diagnosis of 149 (85.6%) of the cases. For the diagnosis of the rest 24 (13.8%) people with diabetes, there was no evidence of any laboratory investigation. Diagnoses made without evidence of supporting laboratory investigation were mostly for Type I people with diabetes and for those diagnosed in Health Centers. For three-fourth of the cases, urinalysis to detect proteinuria was conducted at the time of diagnosis of diabetes. None of the cases were offered with Low Density Lipoprotein (LDL) cholesterol and Glycated Haemoglobin (HbA1c) tests at the time of diagnosis.

3.4. Diabetes care and Treatment

Observation of the type of Diabetes management that people with diabetes were attending at the time of the survey revealed that the majority (51.1%) of the cases were receiving hypoglycaemic agents and 46.6% were on insulin while 5(2.9%) were on non-pharmacologic management of diabetes.

The median number of annual scheduled clinic visits attended was 6 and ranges from 0 to 14. People with type 2 diabetes had more annual number of visits compared to those with type I diabetes. Only 15 (8.7%) of the cases had at least three visits during the previous six months period prior to data collection. Cases being attended in health Centers had also more visits as compared to those cases being attended in the hospital. There was no difference between male and female cases with regard to the annual number of scheduled visits.

The majority, 160 (92%), of the study subjects had at least one Fasting Blood Sugar (FBS) test during the one year period preceding the survey. However, only 31% had at least six FBS tests during the same period and none of them had HbA1c test. There was no statistically significant difference in number of FBS measurements between Health centres and the Hospital.

The majority, 154 (89.1%), of the study subjects had at least one Blood Pressure (BP) measurement during the last one year.

Other important tests/procedures like Urinalysis (for protein and glucose), dilated eye examination, and feet examination were rarely conducted during follow up. During the 12 months period prior to the study, only 8.6% of the cases had feet examination while only 4% had dilated eye examination at least once. The proportion of cases who received urinalysis for glucose and urinalysis for protein were 5.1% and 5.7% respectively. Only 1.1% of the cases had LDL cholesterol tests during the time period considered in this study.

During the one year period preceding the survey, 23 (13.2%) of the cases had been hospitalized at least once. Among those who were being attended in the health centers, 4(6.7%), were referred to the hospital. About one-third of the cases (32.2%) had at least one dosage change in their treatment regimen during the 12 months prior to the survey due to inadequate control of blood sugar levels.

3.5. Achievement of Desired Health Outcomes

Review of the follow up status of cases at the time of survey indicated that 127(73%) of them were regularly attending follow up clinic as per their appointment schedule while 44(25.3%) hadn’t attended the clinic for more than one month after their last date of appointment. Four cases were referred for better investigation and management.

The number of people with diabetes with evidence of at least one incident of uncontrolled blood sugar during last six months, based on FBS, was 117 (67.2%). Of these, only 51 (29.3%) had documented evidence of reception of one or more of the following interventions: adherence counselling, diabetes education and dose adjustment.

Analysis of blood pressure values for the most recent visit prior to the survey indicated that 128 (73.5%) had systolic blood pressure greater than 140mmhg. Only 5 (3.5%) cases had both of systolic and diastolic blood pressure values within the recommended range. Fifty-seven cases (32.8%) had BP results greater than 140/90mmhg on two occasions at least two weeks apart during the study period. Of these 39 (68.4%) had received additional services for their raised blood pressure.

4. Discussion

This study described how diabetes care is being provided in resource constrained settings. It examined access to and quality of diagnosis and treatment services to people with diabetes in hospital and health center settings.

The average age of type 2 people with diabetes at time of diagnosis in this study was 48.6 years and quite a remarkable number of Type 2 people with diabetes were below the age of 45 years. This average age at time of diagnosis of Type 2 diabetes is lower than traditionally expected as it was also true in the analysis of the changes in the mean age of diabetes diagnosis in other settings.^7,8

The findings in this study showed that people with diabetes had better physical access to Health facilities as compared to the findings in previous studies conducted in Ethiopia.⁹ The delivery of diabetes care services at Health centres also improved physical access to diagnosis and treatment services for people with diabetes. However, better access to physical health facilities did not ensure the proper management of diabetes; there were more gaps in quality of care in the health centers compared to the hospital. Only small portions of patients were found to receive essential tests to monitor the level of glycemic control and timely diagnose possible complications on a regular basis. Some of the tests were not conducted simply because the services were not available in health facilities (e.g. HbA1c). However, simple and feasible examinations like foot examination were also conducted for a very limited number of cases. The proportion of cases who received feet examination was even lower than lowest figures in other studies.¹⁰ Hence, further quality improvement measures can have significant role in better attainment of health outcomes even with the prevailing resource constraints.

Even if most of the study subjects had acceptable number of visits to health facilities, higher level of markers of poor adherence like hospitalization, referral, higher doses of medications, and frequent dosage changes. These need attention of stakeholders of diabetes management. Unless frequent scheduled visits improve adherence to treatment and prevent complications, the quality of follow up care will be in question. The impact will be severe if appropriate actions are not taken early.¹¹

Finally, the low level of disease control among people with diabetes in this study warrants urgent intervention. Improvements in diabetes education and self- management procedures can have a significant role in improving the management of diabetes.¹² Community-based care through the use of community health workers, including health extension workers, can improve knowledge and self-management of diabetes care in rural settings.¹³

This study was conducted in a teaching hospital where there are relatively better human and material resources and health centres supported by a teaching hospital. The poor quality of services despite these factors indicates that the quality of care for people with diabetes in other types of hospitals and health centres may be even worse calling for urgent quality improvement efforts in the country.

There were a few limitations associated with this study. The first was the data quality challenges faced during the data collection process. This study used data collected through review of medical records. Lack of appropriate and timely documentation of provided components of care, a common problem in developing countries, may result in underestimation of process indicators assessed in the study. The findings should, therefore, be interpreted in the context of such data quality issues. The second limitation was that the study was limited to analysis of patient data in three health facilities within one year period preceding the data collection. The study was conducted in Jimma University Specialized Hospital and health centres supported by the university. The findings, therefore, reflect diabetes management within the scope facilities supported by Jimma University. The findings could possibly better represent the management of diabetes in southwest Ethiopia.

5. Conclusions

Despite a relatively better access of patients to health facilities in Ethiopia, a huge gap exists in the provision of recommended components of diabetes care and the level of achievement of glycaemic control among people with diabetes in both hospital and health centre settings. There is an urgent need to improve the quality of diabetes services in resource constrained settings to ensure achievement of maximum health benefits to patients.

ACKNOWLEDGEMENTS

We would like to extend our acknowledgement to the participating health facilities and Jimma University for their support. We also thank professionals involved in the data collection process. Our special thanks go to Challi Jira and Dereje Nigussie who have played key role in the conduct of the previous study.

Conflict of interest: None declared

Authors’ contribution: Haregu TN participated in the design of the analysis, analysis of data and write up of the manuscript. Alemayehu YK participated in initiating the concept, interpretation of the findings and report write up. Both authors have read and approved the manuscript.

Disclaimer: The views expressed in this study are solely the views of the authors and do not necessarily reflect the views of any other organization.