Subgroup analysis and interaction tests: why they are important and how to avoid common mistakes

Ferreira, Juliana Carvalho; Patino, Cecilia María

doi:10.1590/s1806-37562017000000170

Cited by 18 publications

(19 citation statements)

References 2 publications

(3 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…25 , 27 Further, interaction tests are recommended to assess heterogeneous treatment effects. 4 , 25 , 27 Potential susceptibilities to false-positive findings because of multiple testing should be considered. 25 , 27 Even though results were often shown to be questionable, not investigating subgroup analyses might also cause misleading therapeutic recommendations.…”

Section: Methodsmentioning

confidence: 99%

“… 1 – 3 These are used to assess heterogeneity of treatment effects, concerning primary, secondary or adverse trial outcomes. 4 – 6 Corresponding investigations are generally based on the assumption that certain subgroups of patients may benefit more or less from a studied intervention. 5 , 7 Subgroup analyses are particularly useful when patient characteristics are associated with treatment effects and to define patients with increased risk profiles.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A systematic review of subgroup analyses in randomised clinical trials in cardiovascular disease

2021

View full text Add to dashboard Cite

Background: Subgroup analyses are frequently used to assess heterogeneity of treatment effects in randomised clinical trials. Inconsistent, improper and incomplete implementation, reporting and interpretation have been identified as ongoing challenges. Further, subgroup analyses were frequently criticised because of unreliable or potentially misleading results. More recently, recommendations and guidelines have been provided to improve the reporting of data in this regard. Methods: This systematic review was based on a literature search within the digital archives of three selected medical journals, The New England Journal of Medicine, The Lancet and Circulation. We reviewed articles of randomised clinical trials in the domain of cardiovascular disease which were published in 2015 and 2016. We screened and evaluated the selected articles for the mode of implementation and reporting of subgroup analyses. Results: We were able to identify a total of 130 eligible publications of randomised clinical trials. In 89/130 (68%) articles, results of at least one subgroup analysis were presented. This was dependent on the considered journal (p < 0.001), the number of included patients (p < 0.001) and the lack of statistical significance of a trial’s primary analysis (p < 0.001). The number of reported subgroup analyses ranged from 1 to 101 (median = 13). We were able to comprehend the specification time of reported subgroup analyses for 71/89 (80%) articles, with 55/89 (62%) articles presenting exclusively pre-specified analyses. This information was not always traceable on the basis of provided trial protocols and often did not include the pre-definition of cut-off values for the categorization of subgroups. The use of interaction tests was reported in 84/89 (94%) articles, with 36/89 (40%) articles reporting heterogeneity of the treatment effect for at least one primary or secondary trial outcome. Subgroup analyses were reported more frequently for larger randomised clinical trials, and if primary analyses did not reach statistical significance. Information about the implementation of subgroup analyses was reported most consistently for articles from The New England Journal of Medicine, since it was also traceable on the basis of provided trial protocols. We were able to comprehend whether subgroup analyses were pre-specified in a majority of the reviewed publications. Even though results of multiple subgroup analyses were reported for most published trials, a corresponding adjustment for multiple testing was rarely considered. Conclusion: Compared to previous reviews in this context, we observed improvements in the reporting of subgroup analyses of cardiovascular randomised clinical trials. Nonetheless, critical shortcomings, such as inconsistent reporting of the implementation and insufficient pre-specification, persist.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A systematic review of subgroup analyses in randomised clinical trials in cardiovascular disease

2021

View full text Add to dashboard Cite

show abstract

“…19 In practice, understanding treatment effects across patient subgroups are important to identify patient groups that respond better or worse to the intervention according to different baseline variables, 27 however, performing multiple testings should be avoided. 19,27,28 The number of subgroup analyses varied from 1 to 101 with a median of 13 from 130 eligible publications of randomized clinical trials in a recent systematic review, 4 compared with the range of 1 to 24 with a median of 4 in the study by Assmann and colleagues reviewing 50 clinical trials that were published in major journals in 1997. 7 When multiple subgroup testings are conducted, the potential susceptibility to a false-positive finding can be greatly inflated.…”

Section: Challenges Of Subgroup Analysesmentioning

confidence: 99%

“…25,29 To avoid overestimating subgroup effects because of inflated false-positive findings in 10 independent subgroup testings, the interaction tests can be assessed under a new criterion of 0.05Ä10 (0.005), using P < .005 as the significance level. 3,27,29 Selective reporting of post hoc subgroup observations or selective reporting of significant analyses regardless of the overall nonsignificant results is a data-generating method, instead of hypothesis testing. 19 This harmful research method has been described as a possible "fishing trip" or "fishing for significance".…”

Section: Challenges Of Subgroup Analysesmentioning

confidence: 99%

Research and scholarly methods: Subgroup analysis

Wan¹,

Ding

2022

J Am Coll Clin Pharm

View full text Add to dashboard Cite

Subgroup analysis is designed to evaluate whether an intervention has differing effects according to baseline characteristics of participants in clinical trials. This approach can help identify subgroups of patient populations, within a single trial or across multiple trials (meta-analysis), that may benefit from the intervention or can be hypothesis-generating for future trials. Properly prespecified subgroup analyses can increase the effect size of subgroups and the legitimacy of P-values in determining significance and hypothesis testing. However, many pitfalls and limitations exist and must be recognized, including post hoc data-generating, multiple testing, selective reporting, and insufficient sample size. Failure to account for these limitations could lead to erroneous interpretations that adversely impact clinical decision-making. This review will provide a general overview, best practices, and pitfalls to avoid for investigators involved in the design, conduct, or interpretation of subgroup analyses.

show abstract

“…Stratified analyses. We first performed a test for interaction with ICU admission using multivariable models to identify stratified groups with statistically significant differences [22,23]. We then performed analyses of variables in stratified groups with a significant interaction.…”

Section: Plos Onementioning

confidence: 99%

Association between early intensive care or coronary care unit admission and post-discharge performance of activities of daily living in patients with acute decompensated heart failure

et al. 2021

View full text Add to dashboard Cite

The management of acute decompensated heart failure often requires intensive care. However, the effects of early intensive care unit/coronary care unit admission on activities of daily living (ADL) in acute decompensated heart failure patients have not been precisely evaluated. Thus, we retrospectively assessed the association between early intensive care unit admission and post-discharge ADL performance in these patients. Acute decompensated heart failure patients (New York Heart Association I–III) admitted on emergency between April 1, 2014, and December 31, 2018, were selected from the Diagnosis Procedure Combination database and divided into intensive care unit/coronary care unit (ICU) and general ward (GW) groups according to the hospitalization type on admission day 1. The propensity score was calculated to create matched cohorts where admission style (intensive care unit/coronary care unit admission) was independent of measured baseline confounding factors, including ADL at admission. The primary outcome was ADL performance level at discharge (post-ADL) defined according to the Barthel index. Secondary outcomes included length of stay and total hospitalization cost (expense). Overall, 12231 patients were eligible, and propensity score matching created 2985 pairs. After matching, post-ADL was significantly higher in the ICU group than in the GW group [mean (standard deviation), GW vs. ICU: 71.5 (35.3) vs. 78.2 (31.2) points, P<0.001; mean difference: 6.7 (95% confidence interval, 5.1–8.4) points]. After matching, length of stay was significantly shorter and expenses were significantly higher in the ICU group than in the GW group. Stratified analysis showed that the patients with low ADL at admission (Barthel index score <60) were the most benefited from early intensive care unit/coronary care unit admission. Thus, early intensive care unit/coronary care unit admission was associated with improved post-ADL in patients with emergency acute decompensated heart failure admission.

show abstract

Subgroup analysis and interaction tests: why they are important and how to avoid common mistakes

Cited by 18 publications

References 2 publications

A systematic review of subgroup analyses in randomised clinical trials in cardiovascular disease

A systematic review of subgroup analyses in randomised clinical trials in cardiovascular disease

Research and scholarly methods: Subgroup analysis

Association between early intensive care or coronary care unit admission and post-discharge performance of activities of daily living in patients with acute decompensated heart failure

Contact Info

Product

Resources

About