ImportanceMetabolic surgery leads to weight loss and improved health, but these outcomes are highly variable. Poor weight loss is associated with lower circulating levels of glucagon-like peptide-1 (GLP-1).ObjectiveTo assess the efficacy and safety of the GLP-1 receptor agonist, liraglutide, 3.0 mg, on percentage body weight reduction in patients with poor weight loss and suboptimal GLP-1 response after metabolic surgery.Design, Setting, and ParticipantsThe Evaluation of Liraglutide 3.0 mg in Patients With Poor Weight Loss and a Suboptimal Glucagon-Like Peptide-1 Response (BARI-OPTIMISE) randomized placebo-controlled trial recruited adult patients at least 1 year after metabolic surgery who had experienced 20% or less body weight loss from the day of surgery and a suboptimal nutrient-stimulated GLP-1 response from 2 hospitals in London, United Kingdom, between October 2018 and November 2019. Key exclusion criteria were type 1 diabetes; severe concomitant psychiatric, gastrointestinal, cardiac, kidney or metabolic disease; and use of insulin, GLP-1 receptor analogues, and medication that can affect weight. The study period was 24 weeks followed by a 4-week follow-up period. Last participant follow-up was completed in June 2020. All participants and clinical study personnel were blinded to treatment allocation. Of 154 assessed for eligibility, 70 met trial criteria and were included in the study, and 57 completed follow-up.InterventionsLiraglutide, 3.0 mg, once daily or placebo as an adjunct to lifestyle intervention with a 500-kcal daily energy deficit for 24 weeks, on a 1:1 allocation by computer-generated randomization sequence, stratified by surgery type (Roux-en-Y gastric bypass [RYGB] or sleeve gastrectomy [SG]) and type 2 diabetes status.Main Outcome and MeasuresThe primary outcome was change in percentage body weight from baseline to the end of the 24-week study period based on an intention-to-treat analysis. Participant safety was assessed through monitoring of biochemical parameters, including kidney and liver function, physical examination, and assessment for adverse events.ResultsA total of 70 participants (mean [SD] age, 47.6 [10.7] years; 52 [74%] female) with a poor weight loss response following RYGB or SG were randomized to receive 3.0-mg liraglutide (n = 35) or placebo (n = 35). All participants received at least 1 dose of the trial drug. Eight participants discontinued treatment (4 per group), and 2 in the 3.0-mg liraglutide group and 1 in the placebo group were lost to follow-up. Due to COVID-19 restrictions, 3 participants in the 3.0-mg liraglutide group and 7 in the placebo group were unable to attend their final in-person assessment. Estimated change in mean (SD) percentage body weight from baseline to week 24 was −8.82 (4.94) with liraglutide, 3.0 mg (n = 31), vs −0.54 (3.32) with placebo (n = 26). The mean difference in percentage body weight change for liraglutide, 3.0 mg, vs placebo was −8.03 (95% CI, −10.39 to −5.66; P < .001). Adverse events, predominantly gastrointestinal, were more frequent with liraglutide, 3.0 mg (28 events [80%]), than placebo (20 events [57%]). There were no serious adverse events and no treatment-related deaths.Conclusion and RelevanceThese findings support the use of adjuvant liraglutide, 3.0 mg, for weight management in patients with poor weight loss and suboptimal GLP-1 response after metabolic surgery.Trial RegistrationClinicalTrials.gov Identifier: NCT03341429

Regression discontinuity designs (RDDs) have been developed for the estimation of treatment effects using observational data, where a treatment is administered using an externally defined decision rule, linked to a continuous assignment variable. Typically, RDDs have been applied to situations where the outcome of interest is continuous and non‐temporal. Conversely, RDDs for time‐to‐event outcomes have received less attention, despite such outcomes being common in many applications. We explore RDDs for a time‐to‐event outcome subject to right censoring. An accelerated failure time (AFT) approach is used to establish a treatment effect estimate for a fuzzy RDD (where treatment is not always strictly applied according to the decision rule). This estimation approach is robust to different levels of fuzziness and unobserved confounding, assessed using simulation studies and compares favourably to established structural AFT models. A motivating example is presented in which models are fitted to estimate the effect of metformin on mortality and cardiovascular disease rate using real observational data from UK Primary Care.