ObjectiveThis study aimed to investigate the preferences regarding risks, benefits and other treatment attributes of patients with type 2 diabetes mellitus (T2DM) in China when selecting a second-line anti-hyperglycaemic medicine.MethodsA discrete choice experiment with hypothetical anti-hyperglycaemic medication profiles was performed using a face-to-face survey administered to patients with T2DM. The medication profile was described using seven attributes: treatment efficacy, hypoglycaemia risk, cardiovascular benefits, gastrointestinal (GI) adverse events, weight change, mode of administration and out-of-pocket cost. Participants chose between medication profiles by comparing attributes. Data were analysed using a mixed logit model with marginal willingness to pay (mWTP) and maximum acceptable risk (MAR) calculated. The preference heterogeneity within the sample was explored using a latent class model (LCM).ResultsA total of 3327 respondents from five major geographical regions completed the survey. Treatment efficacy, hypoglycaemia risk, cardiovascular benefits and GI adverse events were major concerns among the seven attributes measured. Weight change and mode of administration were of lesser concern. Regarding mWTP, respondents would pay ¥236.1 (US$36.6) for an anti-hyperglycaemic medication with an efficacy of 2.5% points reduction in HbA1c, while they were willing to accept a weight gain of 3 kg only if they received a compensation of ¥56.7 (US$8.8). Respondents were willing to accept a relatively large increase in hypoglycaemia risk (MAR=15.9%) to improve treatment efficacy from intermediate (1.0% points) to high (1.5% points). LCM identified the following four unobserved subgroups: trypanophobia, cardiovascular-benefits-focused, safety-focused and efficacy-focused and cost-sensitive.ConclusionPatients with T2DM prioritised free out-of-pocket costs, highest efficacy, no hypoglycaemia risk and cardiovascular benefits over weight change and mode of administration. There exists great preference heterogeneity among patients, which should be taken into account in healthcare decision-making processes.