Bayesian Hierarchical Models for Subgroup Analysis.

In conventional subgroup analyses, subgroup treatment effects are estimated using data from each subgroup separately without considering data from other subgroups in the same study. The subgroup treatment effects estimated this way may be heterogenous with high variability due to small sample sizes in some subgroups and much different from the treatment effect in the overall population. A Bayesian hierarchical model (BHM) can be used to derive more precise, and less heterogenous estimates of subgroup treatment effects that are closer to the treatment effect in the overall population. BHM assumes exchangeability in treatment effect across subgroups after adjusting for effect modifiers and other relevant covariates. In this article, we will discuss the technical details for applying one-way and multi-way BHM using summary-level statistics, and patient-level data for subgroup analysis. Four case studies based on four new drug applications are used to illustrate the application of these models in subgroup analyses for continuous, dichotomous, time-to-event, and count endpoints.

Statistical methods for handling missing data to align with treatment policy strategy.

The International Council for Harmonization (ICH) E9(R1) addendum recommends choosing an appropriate estimand based on the study objectives in advance of trial design. One defining attribute of an estimand is the intercurrent event, specifically what is considered an intercurrent event and how it should be handled. The primary objective of a clinical study is usually to assess a product's effectiveness and safety based on the planned treatment regimen instead of the actual treatment received. The estimand using the treatment policy strategy, which collects and analyzes data regardless of the occurrence of intercurrent events, is usually utilized. In this article, we explain how missing data can be handled using the treatment policy strategy from the authors' viewpoint in connection with antihyperglycemic product development programs. The article discusses five statistical methods to impute missing data occurring after intercurrent events. All five methods are applied within the framework of the treatment policy strategy. The article compares the five methods via Markov Chain Monte Carlo simulations and showcases how three of these five methods have been applied to estimate the treatment effects published in the labels for three antihyperglycemic agents currently on the market.

Acute kidney injury and renal replacement therapy independently predict mortality in neonatal and pediatric noncardiac patients on extracorporeal membrane oxygenation.

OBJECTIVE: To determine the independent impact of acute kidney injury (AKI) and renal replacement therapy (RRT) in infants and children who receive extracorporeal membrane oxygenation. Despite continued expertise/technological advancement, patients who receive extracorporeal membrane oxygenation have high mortality. AKI and RRT portend poor outcomes independent of comorbidities and illness severity in several critically ill populations. DESIGN: Retrospective cohort study. The primary variables explored are AKI (categorical complication code for serum creatinine > 1.5 mg/dL or International Statistical Classification of Diseases and Related Health Problems, Revision 9 for acute renal failure), and RRT (complication/Current Procedural Terminology code for dialysis or hemofiltration). Multiple variables previously associated with mortality in this population were controlled, using logistic stepwise regression. Decision tree modeling was performed to determine optimal variables and cut points to predict mortality. PATIENTS: Critically ill neonates (0-30 days old) and children (> 30 days but < 18 yrs old) in the Extracorporeal Life Support Organization registry. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Neonatal mortality was 2175 (27.4%) of 7941. Nonsurvivors experienced more AKI (413 [19%] of 2175 vs. 225 [3.9%] of 5766, p < .0001), and more received RRT (863 [39.7%] of 2175 vs. 923 [16.0%] of 5766, p < .0001) than survivors. Pediatric mortality was 816 (41.6%) of 1962. Pediatric nonsurvivors similarly experienced more AKI (264 [32.3%] of 816 vs. 138 [12.0%] of 1146, p < .0001) and RRT (487 [58.9%] of 816 vs. 353 [30.8%] of 1146, p < .0001) than survivors. After adjusting for confounding variables, the adjusted odds ratio for neonatal group was 3.2 (p < .0001) post AKI and 1.9 (p < .0001) given RRT. Similarly, the pediatric adjusted odds ratio for mortality was 1.7 (p < .001) post AKI and 2.5 (p < .0001) given RRT. AKI and RRT were essential in the neonatal and pediatric mortality decision trees. CONCLUSIONS: After adjusting for known predictors of mortality, AKI and RRT independently predict mortality in neonates and children, who receive extracorporeal membrane oxygenation. Ascertainment of AKI risk factors, testing novel therapies, and optimizing the timing/delivery of RRT may positively impact survival.