Cohort studies examining medical interventions face significant shortcomings in outcome reporting standards compared to randomised controlled trials, according to new analysis by Song and colleagues published in The BMJ. While these observational studies provide crucial evidence for populations underrepresented in clinical trials, they lack the robust registration and reporting requirements that govern randomised trials.
Registration Requirements Across Study Types
Mandatory registration policies by study design, 2025
Source: ICMJE Guidelines, BMJ 2025 | Georgian Medical Journal News
Growing Reliance on Non-Randomised Evidence
Cohort studies of interventions have become increasingly important as they complement randomised controlled trials by providing evidence for situations where randomisation is difficult or impossible. Song and colleagues note in their BMJ analysis that these studies are “sometimes the best or the only evidence about interventions and outcomes” for specific populations.
The International Committee of Medical Journal Editors mandates registration for all randomised controlled trials, requiring primary outcomes to be defined in funding applications and ethical approvals. This framework has significantly improved transparency and reduced selective reporting in clinical trials. For more insights on research methodology improvements, see our latest studies coverage.
Registration Gap Creates Transparency Concerns
Unlike randomised trials, cohort studies face no universal registration requirements despite their growing influence on clinical practice and policy decisions, according to Song and colleagues. This disparity creates opportunities for selective outcome reporting and publication bias that could mislead healthcare practitioners and policymakers.
The absence of prospective registration means researchers can modify outcome definitions, analysis plans, or reporting strategies after seeing results. The WHO International Clinical Trials Registry Platform primarily focuses on interventional studies, leaving observational research largely unregulated.
Implications for Evidence-Based Medicine
Song and colleagues emphasise that “policies and procedures to minimise reporting biases in randomised controlled trials have not been applied widely to observational studies.” This creates an evidence hierarchy where the most rigorous reporting standards apply to studies of populations that may not represent real-world clinical practice.
The US Centers for Disease Control and Prevention and other public health agencies increasingly rely on observational studies for policy recommendations, making transparent reporting essential for informed decision-making according to the BMJ analysis.
Methodological Advances Demand Better Standards
Song and colleagues note that methodological advances have enhanced the reliability of non-randomised designs, particularly cohort studies that emulate target trials using new users and active comparators. These approaches can produce robust causal estimates when properly conducted.
However, these methodological improvements are undermined when reporting standards lag behind design quality, according to the researchers. For related discussions on research quality, visit our quality and safety section.
Cohort studies are sometimes the best or only evidence about interventions and outcomes that are difficult to assess in randomised controlled trials or populations underrepresented in these trials, yet lack the registration requirements that ensure transparent reporting.
— Song and colleagues, The BMJ (2025)
Key takeaways
- Cohort studies lack the registration requirements mandated for randomised controlled trials, creating transparency gaps according to Song and colleagues
- These studies provide crucial evidence for populations underrepresented in clinical trials, including elderly and pregnant patients
- The BMJ analysis highlights that selective outcome reporting in cohort studies could mislead clinical practice and health policy decisions
Frequently asked questions
Why are cohort studies important if randomised trials provide better evidence?
According to Song and colleagues, cohort studies examine populations and outcomes often excluded from randomised trials, such as pregnant women, elderly patients with multiple conditions, and long-term safety outcomes. They provide essential real-world evidence that complements controlled trial findings.
What is outcome reporting bias and why does it matter?
Outcome reporting bias occurs when researchers selectively report favourable results while suppressing negative or null findings. The BMJ analysis notes that without prospective registration, studies can modify their primary outcomes after seeing results, potentially inflating treatment benefits or hiding harms.
How do registration requirements improve research quality?
Registration forces researchers to specify their primary outcomes, analysis methods, and study populations before seeing results. This prevents data dredging and selective reporting, ensuring that published findings reflect the original research questions rather than post-hoc analysis.
Song and colleagues conclude that as cohort studies become increasingly important for evidence-based medicine, particularly for vulnerable populations, the research community must develop registration and reporting frameworks that match their methodological advances. Without such improvements, the potential for bias may undermine the valuable contributions these studies make to clinical practice and health policy.
Source: Outcome reporting in cohort studies of interventions
