Emulating the GRADE Trial Using Real-World Data

ABSTRACT OBJECTIVESTo emulate the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) trial using real-world data prior to its publication. GRADE is the first comparative effectiveness study to directly compare second-line glucose-lowering medications with respect to their ability to lower hemoglobin A1c (HbA1c). DESIGN AND SETTINGIn this observational cohort study, we applied GRADE trial criteria to claims and laboratory data from OptumLabs? Data Warehouse (OLDW), a U.S. nationwide claims database, between 1/25/2010 and 6/30/2019. PARTICIAPNTSAdults with type 2 diabetes with hemoglobin A1c (HbA1c) 6.8-8.5% on metformin monotherapy, identified according to GRADE trial specifications. INTERVENTIONSGlimepiride, liraglutide, sitagliptin, and insulin glargine. MAIN OUTCOME MEASURESThe primary outcome was time to HbA1c ≥7.0% and secondary outcomes were other metabolic, microvascular, macrovascular, and safety outcomes specified by GRADE. Propensity scores were estimated using the gradient boosting machine method and inverse propensity score weighting was used to emulate randomization of the treatment groups, which were then compared using Cox proportional hazards regression. RESULTSWe identified 8252 patients (19.7% of adults starting the study drugs in OLDW) meeting GRADE eligibility criteria (glimepiride arm=4318, liraglutide arm=690, sitagliptin arm=2993, glargine arm=251). The glargine arm was excluded from analyses due to small sample size. Median times to HbA1c ≥7.0% were 442 (95% CI, 394-480) days for glimepiride, 764 (95% CI, 741-NA) days for liraglutide, and 427 (95% CI, 380-483) days for sitagliptin. Liraglutide was associated with lower risk of reaching HbA1c ≥7.0% compared to glimepiride (HR 0.57 [95% CI, 0.43-0.75]) and sitagliptin (HR 0.55 [95% CI, 0.41-0.73]). Results were consistent for the secondary outcome of time to HbA1c >7.5%. There were no significant differences among treatment groups for the remaining secondary outcomes. CONCLUSIONSIn this emulation of the GRADE trial, liraglutide was significantly more effective at maintaining glycemic control than glimepiride or sitagliptin when added to metformin monotherapy. There is value in generating timely evidence on medical treatments using real-world data as a complement to prospective trials. SUMMARY BOXESWhat is already known about this topic?Real-world data is an important source of information about clinical practice, comparative effectiveness and safety, and health outcomes and has the potential to generate timely, pragmatic evidence on medical treatments as a complement to prospective clinical trials.Multiple classes of second-line glucose-lowering medications have been approved for the management of type 2 diabetes, with limited evidence of their comparative effectiveness with respect to glycemic control.What this study adds?We emulated the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) randomized clinical trial using data from a U.S. nationwide administrative claims database to identify the strengths and limitations of using real-world data to emulate prospective comparative effectiveness trials, particularly when examining medications in contexts that may not be the standard of care.Liraglutide is more effective than glimepiride and sitagliptin at lowering HbA1c, supporting its preferential use when substantial glycemic reduction is needed.Advanced causal inference analytic methods applied to observational data can be used to efficiently and effectively emulate clinical trials.