The Net Value of Diabetes Management in South Asia: Evidence from the CARRS Randomized Control Trial in India and Pakistan

Economic dimensions of chronic disease control in low- and middle-income countries are understudied, relative to the large morbidity and mortality burden in those countries and the associated direct and indirect medical expenditures. We analyze detailed patient-level data on clinical risk factors and outcomes, linked to healthcare utilization and spending, as collected for 1146 poorly controlled type 2 diabetes patients in urban South Asia, in the Centre for Cardiometabolic Risk Reduction in South Asia (CARRS) Trial. These patients at baseline had glycated haemoglobin (HbA1c)≥8.0% and either: Systolic Blood Pressure (BP)≥140mmHg or low-density lipoprotein cholesterol (LDLc)≥130mg/dl, with a median duration of diabetes of 7 years. Patients were randomized to usual care or a multicomponent quality improvement intervention featuring clinical decision-support software to improve physician’s responsiveness to treatment modification and non-physician care coordinators to enhance patients’ adherence to therapy.

We estimate the net value of the additional resource use of the intervention by comparing the value of reduced risk of major complications and death relative to the incremental resource use for the period 2011-17. Using a variety of risk prediction models (UKPDS, WHO, Asia-recalibrated RECODE, and India-specific models), we estimate the reduction in risks of fatal stroke, fatal acute myocardial infarction, and all-cause mortality, and then use a range of values of a life-year based on multiples of per capita income to estimate the net value of the additional medical spending associated with the intervention. We contrast the change in net value for the intervention and control groups, and perform sensitivity analyses (using inverse probability weighting and multiple imputations) to account for trial attrition and missing data. We find that the intervention substantially improved health outcomes as proxied by multiple risk factor control (HbA1c<7%, and BP<130/80mmHg, or LDLc<100mg/dl), at modest additional cost relative to the avoided treatment spending and survival gains. The estimates of net value reinforce the promising overall results from the CARRS Trial, demonstrating that a simple multifactorial intervention that includes attention to physician's incentives and decision support, as well as patient adherence, can improve health outcomes in resource-constrained settings. We also compare estimates of the avoidable admissions rate and net value with those developed for several other Asian economies (Japan, Hong Kong, Taiwan, and Singapore), demonstrating in comparative perspective the substantial economic gains that could be obtained from better diabetes management in South Asia by following the pragmatic intervention of the CARRS trial.