The analysis of subgroups is often used as a way to glean additional information from data sets. The strengths and weaknesses of this approach and new Journal policies concerning the reporting of subgroup analyses are discussed in this article.

/pdf/statistics-in-medicine-reporting-of-subgroup-analyses-in-4momuzf1gd.pdf

Statistics in Medicine — Reporting of Subgroup Analyses in Clinical Trials

Objective: To determine whether extended transthoracic esophagectomy for adenocarcinoma of the mid/distal esophagus improves long-term survival. Background: A randomized trial was performed to compare surgical techniques. Complete 5-year survival data are now available. Methods: A total of 220 patients with adenocarcinoma of the distal esophagus (type I) or gastric cardia involving the distal esophagus (type II) were randomly assigned to limited transhiatal esophagectomy or to extended transthoracic esophagectomy with en bloc lymphadenectomy. Patients with peroperatively irresectable/incurable cancer were excluded from this analysis (n = 15). A total of 95 patients underwent transhiatal esophagectomy and 110 patients underwent transthoracic esophagectomy. Results: After transhiatal and transthoracic resection, 5-year survival was 34% and 36%, respectively (P = 0.71, per protocol analysis). In a subgroup analysis, based on the location of the primary tumor according to the resection specimen, no overall survival benefit for either surgical approach was seen in 115 patients with a type II tumor (P = 0.81). In 90 patients with a type I tumor, a survival benefit of 14% was seen with the transthoracic approach (51% vs. 37%, P = 0.33). There was evidence that the treatment effect differed depending on the number of positive lymph nodes in the resection specimen (test for interaction P = 0.06). In patients (n = 55) without positive nodes locoregional disease-free survival after transhiatal esophagectomy was comparable to that after transthoracic esophagectomy (86% and 89%, respectively). The same was true for patients (n = 46) with more than 8 positive nodes (0% in both groups). Patients (n = 104) with 1 to 8 positive lymph nodes in the resection specimen showed a 5-year locoregional disease-free survival advantage if operated via the transthoracic route (23% vs. 64%, P = 0.02). Conclusion: There is no significant overall survival benefit for either approach. However, compared with limited transhiatal resection extended transthoracic esophagectomy for type I esophageal adenocarcinoma shows an ongoing trend towards better 5-year survival. Moreover, patients with a limited number of positive lymph nodes in the resection specimen seem to benefit from an extended transthoracic esophagectomy.

Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the mid/distal esophagus: five-year survival of a randomized clinical trial.

Sexual dimorphism in anatomical, physiological and behavioural traits are characteristics of many vertebrate species. In humans, sexual dimorphism is also observed in the prevalence, course and severity of many common diseases, including cardiovascular diseases, autoimmune diseases and asthma. Although sex differences in the endocrine and immune systems probably contribute to these observations, recent studies suggest that sex-specific genetic architecture also influences human phenotypes, including reproductive, physiological and disease traits. It is likely that an underlying mechanism is differential gene regulation in males and females, particularly in sex steroid-responsive genes. Genetic studies that ignore sex-specific effects in their design and interpretation could fail to identify a significant proportion of the genes that contribute to risk for complex diseases.

/pdf/sex-specific-genetic-architecture-of-human-disease-4keoo88s5g.pdf

Sex-specific genetic architecture of human disease

We consider the problem of identifying a subgroup of patients who may have an enhanced treatment effect in a randomized clinical trial, and it is desirable that the subgroup be defined by a limited number of covariates. For this problem, the development of a standard, pre-determined strategy may help to avoid the well-known dangers of subgroup analysis. We present a method developed to find subgroups of enhanced treatment effect. This method, referred to as 'Virtual Twins', involves predicting response probabilities for treatment and control 'twins' for each subject. The difference in these probabilities is then used as the outcome in a classification or regression tree, which can potentially include any set of the covariates. We define a measure Q(Â) to be the difference between the treatment effect in estimated subgroup Â and the marginal treatment effect. We present several methods developed to obtain an estimate of Q(Â), including estimation of Q(Â) using estimated probabilities in the original data, using estimated probabilities in newly simulated data, two cross-validation-based approaches, and a bootstrap-based bias-corrected approach. Results of a simulation study indicate that the Virtual Twins method noticeably outperforms logistic regression with forward selection when a true subgroup of enhanced treatment effect exists. Generally, large sample sizes or strong enhanced treatment effects are needed for subgroup estimation. As an illustration, we apply the proposed methods to data from a randomized clinical trial.

Subgroup identification from randomized clinical trial data

Impacts of adverse childhood experiences on health, mental health, and substance use in early adulthood: A cohort study of an urban, minority sample in the U.S.

Elise Whitely

Papers

Subgroup analyses in randomized trials: risks of subgroup-specific analyses; power and sample size for the interaction test.