Abstract
There is a lack of longitudinal data on type 2 diabetes (T2D) in low- and middle-income countries. We leveraged the electronic health records (EHR) system of a publicly funded academic institution to establish a retrospective cohort with longitudinal data to facilitate benchmarking, surveillance, and resource planning of a multi-ethnic T2D population in Malaysia. This cohort included 15,702 adults aged ≥ 18 years with T2D who received outpatient care (January 2002-December 2020) from Universiti Malaya Medical Centre (UMMC), Kuala Lumpur, Malaysia. The mean age of participants was 54.3 ± 12.6 years, with a T2D duration of 12.7 ± 4.8 years, HbA1c of 8.9 ± 2.6%, body mass index of 28.2 ± 6.2 kg/m2, and 47.4% were men. The top three comorbidities were dyslipidaemia (87.1%), overweight/obesity (69.4%), and hypertension (62.6%). The proportion of participants achieving HbA1c < 7%, blood pressure < 130/80 mmHg, and low-density lipoprotein cholesterol < 2.6 mmol/L was 27.8%, 24.8%, and 24.5%, respectively. The most common treatments were metformin (62.4%), sulfonylurea (32.8%), and insulin (32.7%). Given the lack of implementation of urinary albumin:creatinine ratio for early detection, chronic kidney disease (defined as estimated glomerular filtration rate < 60 mL/min/1.73m2) was underestimated at 7.5%. These findings highlight opportunities for improved data collection in a middle-income country in Southeast Asia. Apart from trend analysis, this cohort will be prospectively followed for ongoing benchmarking, surveillance, and ascertainment of clinical events, including death.