STROBE Statement and Observational Study Reporting

The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) Statement is a reporting guideline for the three main analytical observational designs — cohort, case-control, and cross-sectional studies. Its checklist specifies the information such studies should report so that readers can understand what was done, judge the threats to validity, and assess generalisability.

Definition

STROBE is an expert-consensus reporting guideline consisting of a checklist of items that authors of cohort, case-control, and cross-sectional studies should address when describing the study's design, setting, participants, variables, statistical methods, results, and limitations.

Scope

This topic covers the purpose and structure of STROBE, the items it asks observational studies to report, how it parallels CONSORT for trials, and its role in the critical appraisal of non-randomised evidence. STROBE addresses reporting rather than conduct; this entry is reference-educational and is not clinical guidance.

Core questions

• What should an observational study report so that it can be appraised?
• How does STROBE handle the distinct features of cohort, case-control, and cross-sectional designs?
• Why is reporting of confounding, bias, and participant selection emphasised for observational studies?
• How does STROBE differ from CONSORT, and why are separate guidelines needed?

Key concepts

• Reporting checklist for observational designs
• Confounding and its reporting
• Selection of participants and eligibility
• Bias and its sources (reported items)
• Cohort, case-control, and cross-sectional design distinctions
• Reporting completeness versus study validity

Mechanisms

STROBE provides a checklist whose items map onto the parts of an observational study report — title and abstract, background, methods (including study design, setting, participants, variables, data sources, bias, study size, statistical methods), results (participant flow, descriptive data, outcome data, main results), and discussion (von Elm 2007). Several items are tailored to design, because cohort, case-control, and cross-sectional studies handle eligibility, follow-up, and sampling differently. The companion Explanation and Elaboration paper supplies the rationale, examples, and methodological background for each item (Vandenbroucke 2007). By making confounding, selection, and bias explicit in the report, STROBE equips a reader to appraise threats to validity that are central when randomisation is absent (Greenhalgh 1997). It deliberately parallels CONSORT in form, extending the reporting-guideline approach from trials to observational research (Schulz 2010).

Clinical relevance

Much of the evidence on risk factors, prognosis, and harms comes from observational studies, and STROBE makes those studies transparent enough to be appraised and used responsibly. It describes how observational evidence should be reported; it characterises evidence and does not provide diagnostic or treatment recommendations for individuals.

Evidence & guidelines

STROBE is endorsed by many epidemiology and general medical journals and is coordinated through the EQUATOR Network. The core guideline is the 2007 STROBE Statement and its Explanation and Elaboration (von Elm 2007; Vandenbroucke 2007), with later extensions for specific subfields such as molecular and genetic epidemiology. As with CONSORT for trials, STROBE adherence reflects completeness of reporting and is distinct from a study's risk of bias, which is judged separately during appraisal.

History

STROBE was developed by a 2004 international collaboration of epidemiologists, methodologists, and journal editors who recognised that observational studies lacked the reporting standards already emerging for trials. The resulting statement and its detailed Explanation and Elaboration were published in 2007 (von Elm 2007; Vandenbroucke 2007), consciously modelled on the CONSORT approach (Schulz 2010) and subsequently incorporated into the EQUATOR Network's library of reporting guidelines, with design- and topic-specific extensions following.

Debates

Reporting standard, not a quality scale

STROBE is intended to improve how observational studies are reported and explicitly states it should not be used as a checklist to score study quality or to judge the validity of a study, a caution sometimes overlooked in practice.

Key figures

• Jan Vandenbroucke
• Erik von Elm
• Matthias Egger
• Douglas Altman

Seminal works

• vonelm-2007-strobe
• vandenbroucke-2007-strobe-ee

Frequently asked questions

Which study designs does STROBE cover?

STROBE covers the three main analytical observational designs: cohort studies, case-control studies, and cross-sectional studies. Some checklist items are tailored to the differences between these designs.

Can STROBE be used to grade how good an observational study is?

No. Its authors state that STROBE is a reporting guideline and should not be used as a quality score or to judge a study's validity. Appraising risk of bias is a separate task done with dedicated appraisal tools.