The Singapore National Healthcare Group Diabetes Registry--descriptive epidemiology of type 2 diabetes mellitus

<p><b>INTRODUCTION</b>The National Healthcare Group (NHG) launched an enterprise-wide diabetes registry in 2007. We describe the epidemiology of type 2 diabetes mellitus from 2005 to 2008.</p><p><b>MATERIALS AND METHODS</b>Patients with encounters in NHG from 2005 were identified for inclusion into the Diabetes Registry from existing stand-alone diabetes registries, ICD9CM diagnosis codes, anti-hyperglycaemic medication and laboratory confirmation. Variables extracted for analysis were demographics (age, gender, ethnicity), diabetes-related comorbidities and complications, most recent anti-hyperglycaemic agents dispensed, and the most recent glycated haemoglobin (HbA1C) measurement.</p><p><b>RESULTS</b>The diabetes registry grew 32% from 129,183 patients in 2005 to 170,513 patients in 2008, making up 12% to 15% of all patients in NHG. About half of the type 2 diabetes patients were aged 45 to 64 years. Females were generally older with a median age of 63 to 64 years vs 59 to 61 years in males. The Indian ethnic group accounted a disproportionately higher 13% of patients. Over 95% of type 2 patients had at least one diabetes-related comorbid condition, and diabetes-related complications were principally renal and cardiovascular complications. The majority (86.2% to 89.2%) of primary care patients were on oral anti-hyperglycaemic agents; however, the rate of insulin treatment increased from 10.8% to 13.8%. HbA1C levels in 2008 improved over that in 2005, with the percentage of patients with good glycaemic control improving with age.</p><p><b>CONCLUSION</b>The registry has enabled a baseline assessment of the burden and the care of type 2 diabetes patients in NHG, which will provide critical "evidence" for planning future programmes.</p>

Metabolic control in type 2 diabetes correlates weakly with patient adherence to oral hypoglycaemic treatment

<p><b>INTRODUCTION</b>Patient adherence to treatment is viewed as essential to good metabolic control in diabetes. Our primary objective was to determine if self-reported patient adherence correlated strongly with metabolic control. Our secondary objective was to determine the natural grouping of factors which influence adherence.</p><p><b>MATERIALS AND METHODS</b>Data were collected using a questionnaire set with 5-point Likert scales. Primary analysis was done using Spearman's correlation coefficient between self-reported composite adherence scores and HbA1c. Secondary analysis was done using exploratory factor analysis.</p><p><b>RESULTS</b>The primary analysis suggests that patient adherence to the treatment regime is weakly correlated to metabolic control. Calculated Spearman's rho was 0.197, with a two-tailed P value of 0.027. The secondary analysis demonstrates the natural clustering of factors that influence patient adherence to treatment. A 6-factor solution was found to account for most of the variance in the data. We also found that feelings of frustration, anxiety, and depression were associated with a lack of knowledge about diabetes treatment. In addition, belief in traditional medicine correlated strongly with ethnicity.</p><p><b>CONCLUSION</b>A good treatment regime for type 2 diabetes mellitus influences metabolic outcome far more than patient adherence.</p>

Measuring the quality of care of diabetic patients at the specialist outpatient clinics in public hospitals in Singapore

<p><b>INTRODUCTION</b>This study aims to measure the quality of care for patients with diabetes mellitus at selected Specialist Outpatient Clinics (SOCs) in the National Healthcare Group.</p><p><b>MATERIALS AND METHODS</b>The cross-sectional study reviewed case-records of patients from 6 medical specialties who were on continuous care for a minimum of 15 months from October 2003 to April 2005. Disproportionate sampling of 60 patients from each specialty, excluding those co-managed by Diabetes Centres or primary care clinics for diabetes, was carried out. Information on demographic characteristics, process indicators and intermediate outcomes were collected and the adherence rate for each process indicator compared across specialties. Data analysis was carried out using SPSS version 13.0.</p><p><b>RESULTS</b>A total of 575 cases were studied. The average rate for 9 process indicators by specialty ranged from 47.8% to 70.0%, with blood pressure measurement consistently high across all specialties (98.4%). There was significant variation (P <0.001) in rates across the specialties for 8 process indicators; HbA1c, serum creatinine and lipid profile tests were over 75%, while the rest were below 50%. The mean HbA1c was 7.3% +/- 1.5%. "Optimal" control of HbA1c was achieved in 51.2% of patients, while 50.6% of the patients achieved "optimal" low-density lipoprotein (LDL)-cholesterol control. However, 47.3% of patients had "poor" blood pressure control. Adherence to process indicators was not associated with good intermediate outcomes.</p><p><b>CONCLUSIONS</b>There was large variance in the adherence rate of process and clinical outcome indicators across specialties, which could be improved further.</p>