JMIR Publications

ABSTRACT

Background:

Type 2 diabetes mellitus (T2DM) and its related complications represent a growing economic burden for many countries and health systems. Diabetes complications can be prevented through better disease control but there is a big gap between the recommended treatment and the treatment that patients actually receive. The treatment of T2DM can be challenging because of comprehensive therapeutic targets and individual variability of the patients, leading to the need for precise, personalized treatment.

Objective:

This study aims to develop treatment recommendation models for T2DM based on deep reinforcement learning and then conduct a retrospective study to evaluate the reliability and effectiveness of the models.

Methods:

The data used in our study was the SingHealth Diabetes Registry, encompassing 189,520 T2DM patients with 6,407,958 outpatient visits from 2013 to 2018. The treatment recommendation model was built based on the 80% of data and its effectiveness was evaluated with the remaining 20% of data. Three treatment recommendation models were developed for anti-glycemic, anti-hypertensive and lipid-lowering treatments, by combining knowledge-driven model and data-driven model. The knowledge-driven model, based on the clinical guidelines and expert experiences, was first applied to select the candidate medications. And the data-driven model based on deep reinforcement learning was used to rank the candidates by the expected clinical outcomes. To evaluate the models, short-term outcomes were compared between the model-concordant treatment and the model-non-concordant treatment with confounder adjustment by stratification, propensity score methods and multivariate regression. For long-term outcome, model-concordant rates were included as independent variables to evaluate if the combined three treatments had positive impact on reduction of long-term complication occurrence or death at patient level via multivariate logistic regression.

Results:

The test data consisted of 36,993 patients for evaluating the effectiveness of the three treatment recommendation models. In 43.3% of patient visits, the anti-glycemic medications recommended by the model were concordant with the actual prescriptions of the physicians. The concordant rates for anti-hypertensive medications and lipid-lowering medications were 51.3% and 58.9% respectively. The evaluation results also showed that model-concordant treatments were associated with better glycemic control (odds ratio [OR], 1.73; 95% confidence interval [CI], 1.69-1.76), blood pressure control (OR, 1.26; 95% CI, 1.23-1.29), and blood lipids control (OR, 1.28; 95% CI, 1.22-1.35). We also found that patients with more model-concordant treatments were associated with lower risk of diabetes complications (including 3 macrovascular and 2 microvascular complications) and death, suggesting that the models have the potential of achieving better outcomes in long term.

Conclusions:

The comprehensive management by combining knowledge-driven model and data-driven model has good potential to help physicians improve the clinical outcomes of T2DM patients, achieving good control on blood glucose, blood pressure and blood lipids, and reducing the risk of diabetes complications in the long term.