[STandardized Reporting Of Secondary data Analyses (STROSA)—a recommendation]

Abstract

Secondary data analyses will play an increasingly important role in health services research. But to date, there is no guideline for the systematic, transparent and complete reporting of secondary data. We investigated whether the STROBE statement, i.e., the recommendations for reporting observational studies, satisfies the specific characteristics of secondary data analyses and whether any specifications/modifications and extensions are necessary. For the majority of the 22 STROBE criteria, specifications and extensions are needed to meet the requirements of systematic, transparent and complete reporting of secondary data analysis. Seven aspects of secondary data analysis not covered by STROBE (legal aspects, data flow, protocol, unit of analysis, internal validations/definitions, advantages of secondary data utilisation, role of data owners) should be considered as a specific complement to STROBE. The so called STROSA (STandardized Reporting Of Secondary data Analyses) checklist therefore includes 29 items that relate to the title/abstract, introduction, methods, results and discussion sections of articles. The STROSA checklist is intended to support authors and readers in the critical appraisal of secondary data analyses. This proposal will now be subject to continued scientific discussions.

Keywords: Berichtsstandard; Gute Praxis Sekundärdatenanalyse; RECORD Statement; RECORD statement; Reporting guideline; STROBE; STROSA; good practice secondary data analysis.