Distribution Of Baseline Covariates Research Articles

When the Ends do not Justify the Means: Learning Who is Predicted to Have Harmful Indirect Effects.

There is a growing literature on finding rules by which to assign treatment based on an individual's characteristics such that a desired outcome under the intervention is maximized. A related goal entails identifying a subpopulation of individuals predicted to have a harmful indirect effect (the effect of treatment on an outcome through mediators), perhaps even in the presence of a predicted beneficial total treatment effect. In some cases, the implications of a likely harmful indirect effect may outweigh an anticipated beneficial total treatment effect, and would motivate further discussion of whether to treat identified individuals. We build on the mediation and optimal treatment rule literatures to propose a method of identifying a subgroup for which the treatment effect through the mediator is expected to be harmful. Our approach is nonparametric, incorporates post-treatment confounders of the mediator-outcome relationship, and does not make restrictions on the distribution of baseline covariates, mediating variables, or outcomes. We apply the proposed approach to identify a subgroup of boys in the MTO housing voucher experiment who are predicted to have a harmful indirect effect of housing voucher receipt on subsequent psychiatric disorder incidence through aspects of their school and neighborhood environments.

Association of Bariatric Surgery With Major Adverse Liver and Cardiovascular Outcomes in Patients With Biopsy-Proven Nonalcoholic Steatohepatitis

No therapy has been shown to reduce the risk of serious adverse outcomes in patients with nonalcoholic steatohepatitis (NASH). To investigate the long-term relationship between bariatric surgery and incident major adverse liver outcomes and major adverse cardiovascular events (MACE) in patients with obesity and biopsy-proven fibrotic NASH without cirrhosis. In the SPLENDOR (Surgical Procedures and Long-term Effectiveness in NASH Disease and Obesity Risk) study, of 25 828 liver biopsies performed at a US health system between 2004 and 2016, 1158 adult patients with obesity were identified who fulfilled enrollment criteria, including confirmed histological diagnosis of NASH and presence of liver fibrosis (histological stages 1-3). Baseline clinical characteristics, histological disease activity, and fibrosis stage of patients who underwent simultaneous liver biopsy at the time of bariatric surgery were balanced with a nonsurgical control group using overlap weighting methods. Follow-up ended in March 2021. Bariatric surgery (Roux-en-Y gastric bypass, sleeve gastrectomy) vs nonsurgical care. The primary outcomes were the incidence of major adverse liver outcomes (progression to clinical or histological cirrhosis, development of hepatocellular carcinoma, liver transplantation, or liver-related mortality) and MACE (a composite of coronary artery events, cerebrovascular events, heart failure, or cardiovascular death), estimated using the Firth penalized method in a multivariable-adjusted Cox regression analysis framework. A total of 1158 patients (740 [63.9%] women; median age, 49.8 years [IQR, 40.9-57.9 years], median body mass index, 44.1 [IQR, 39.4-51.4]), including 650 patients who underwent bariatric surgery and 508 patients in the nonsurgical control group, with a median follow-up of 7 years (IQR, 4-10 years) were analyzed. Distribution of baseline covariates, including histological severity of liver injury, was well-balanced after overlap weighting. At the end of the study period in the unweighted data set, 5 patients in the bariatric surgery group and 40 patients in the nonsurgical control group experienced major adverse liver outcomes, and 39 patients in the bariatric surgery group and 60 patients in the nonsurgical group experienced MACE. Among the patients analyzed with overlap weighting methods, the cumulative incidence of major adverse liver outcomes at 10 years was 2.3% (95% CI, 0%-4.6%) in the bariatric surgery group and 9.6% (95% CI, 6.1%-12.9%) in the nonsurgical group (adjusted absolute risk difference, 12.4% [95% CI, 5.7%-19.7%]; adjusted hazard ratio, 0.12 [95% CI, 0.02-0.63]; P = .01). The cumulative incidence of MACE at 10 years was 8.5% (95% CI, 5.5%-11.4%) in the bariatric surgery group and 15.7% (95% CI, 11.3%-19.8%) in the nonsurgical group (adjusted absolute risk difference, 13.9% [95% CI, 5.9%-21.9%]; adjusted hazard ratio, 0.30 [95% CI, 0.12-0.72]; P = .007). Within the first year after bariatric surgery, 4 patients (0.6%) died from surgical complications, including gastrointestinal leak (n = 2) and respiratory failure (n = 2). Among patients with NASH and obesity, bariatric surgery, compared with nonsurgical management, was associated with a significantly lower risk of incident major adverse liver outcomes and MACE.

Assessing Heterogeneity of Treatment Effects in Observational Studies.

Here we describe methods for assessing heterogeneity of treatment effects over prespecified subgroups in observational studies, using outcome-model-based (g-formula), inverse probability weighting, doubly robust, and matching estimators of subgroup-specific potential outcome means, conditional average treatment effects, and measures of heterogeneity of treatment effects. We compare the finite-sample performance of different estimators in simulation studies where we vary the total sample size, the relative frequency of each subgroup, the magnitude of treatment effect in each subgroup, and the distribution of baseline covariates, for both continuous and binary outcomes. We find that the estimators' bias and variance vary substantially in finite samples, even when there is no unobserved confounding and no model misspecification. As an illustration, we apply the methods to data from the Coronary Artery Surgery Study (August 1975-December 1996) to compare the effect of surgery plus medical therapy with that of medical therapy alone for chronic coronary artery disease in subgroups defined by previous myocardial infarction or left ventricular ejection fraction.

Target Population Statistical Inference With Data Integration Across Multiple Sources—An Approach to Mitigate Information Shortage in Rare Disease Clinical Trials

A major challenge for rare disease clinical trials is the limited amount of available information for making robust statistical inference. While external data present information integration opportunities to enhance statistical inference, conventional data combining methods, for example, meta-analysis, usually do not adequately address study population differences. Matching methods, on the other hand, directly account for population characteristics but often lead to inefficient use of data by underutilizing unmatched data points. Aiming at a better bias-variance tradeoff, we propose an intuitive integrated inference framework to borrow information from all relevant data sources and make inference on the response of interest over a target population precisely characterized by the joint distribution of baseline covariates. The method is easily implemented and can be complemented by modern statistical learning or machine learning tools. Statistical inference is facilitated by the bootstrap. We argue that the integrated inference framework not only provides an intuitive and coherent perspective for a variety of clinical trial inference problems but also has broad application areas in clinical trial settings and beyond, as a quantitative data integration tool for making robust inference in a target population precise manner for policy and decision makers.

Association of Metabolic Surgery With Major Adverse Cardiovascular Outcomes in Patients With Type 2 Diabetes and Obesity

Although metabolic surgery (defined as procedures that influence metabolism by inducing weight loss and altering gastrointestinal physiology) significantly improves cardiometabolic risk factors, the effect on cardiovascular outcomes has been less well characterized. To investigate the relationship between metabolic surgery and incident major adverse cardiovascular events (MACE) in patients with type 2 diabetes and obesity. Of 287 438 adult patients with diabetes in the Cleveland Clinic Health System in the United States between 1998 and 2017, 2287 patients underwent metabolic surgery. In this retrospective cohort study, these patients were matched 1:5 to nonsurgical patients with diabetes and obesity (body mass index [BMI] ≥30), resulting in 11 435 control patients, with follow-up through December 2018. Metabolic gastrointestinal surgical procedures vs usual care for type 2 diabetes and obesity. The primary outcome was the incidence of extended MACE (composite of 6 outcomes), defined as first occurrence of all-cause mortality, coronary artery events, cerebrovascular events, heart failure, nephropathy, and atrial fibrillation. Secondary end points included 3-component MACE (myocardial infarction, ischemic stroke, and mortality) and the 6 individual components of the primary end point. Among the 13 722 study participants, the distribution of baseline covariates was balanced between the surgical group and the nonsurgical group, including female sex (65.5% vs 64.2%), median age (52.5 vs 54.8 years), BMI (45.1 vs 42.6), and glycated hemoglobin level (7.1% vs 7.1%). The overall median follow-up duration was 3.9 years (interquartile range, 1.9-6.1 years). At the end of the study period, 385 patients in the surgical group and 3243 patients in the nonsurgical group experienced a primary end point (cumulative incidence at 8-years, 30.8% [95% CI, 27.6%-34.0%] in the surgical group and 47.7% [95% CI, 46.1%-49.2%] in the nonsurgical group [P < .001]; absolute 8-year risk difference [ARD], 16.9% [95% CI, 13.1%-20.4%]; adjusted hazard ratio [HR], 0.61 [95% CI, 0.55-0.69]). All 7 prespecified secondary outcomes showed statistically significant differences in favor of metabolic surgery, including mortality. All-cause mortality occurred in 112 patients in the metabolic surgery group and 1111 patients in the nonsurgical group (cumulative incidence at 8 years, 10.0% [95% CI, 7.8%-12.2%] and 17.8% [95% CI, 16.6%-19.0%]; ARD, 7.8% [95% CI, 5.1%-10.2%]; adjusted HR, 0.59 [95% CI, 0.48-0.72]). Among patients with type 2 diabetes and obesity, metabolic surgery, compared with nonsurgical management, was associated with a significantly lower risk of incident MACE. The findings from this observational study must be confirmed in randomized clinical trials. ClinicalTrials.gov Identifier: NCT03955952.

Índices de propensión: el deseo de parecerse al ensayo clínico

Observational studies lack randomization. Therefore, there is no assurance that the different study groups have the same distribution of baseline covariates,...