How to report and discuss subgroup analyses in clinical practice guidelines? Evaluation procedure of the clinical and statistical relevancy.

Abstract

The results of subgroup analyses of clinical trials are important reference information when considering the generalizability of a study treatment, i.e., providing the best treatment for each individual patient. The results of subgroup analyses are often presented in publications, etc. as forest plots focusing on patient backgrounds. However, it is important to fully understand and grasp some of the issues involved in subgroup analyses and to interpret the results carefully to apply them in clinical practice. Although the literature includes some reports on how subgroup analyses should be evaluated and handled for the purpose of establishing medical practice guidelines, most of the papers have mainly evaluated the reliability of subgroup analyses from a statistical perspective; few of them have incorporated clinical importance in their evaluations. Therefore, in December 2019, we established a Subgroup Analysis Review Committee consisting of oncologists specializing in lung cancer treatment and statistical experts among the members of the Guidelines Review Committee of the Japanese Lung Cancer Association, with the aim of appropriately reflecting subgroup analysis in Japanese lung cancer practice guidelines. We developed a new evaluation strategy to incorporate clinical aspects as well as reliability assessment. Specifically, on the basis of a clinical and statistical review of the problems with subgroup analyses presented as clinical trial results, we developed criteria and procedures to ensure consistency and fairness in the citation of clinical guidelines.