How do I interpret a Centre for Reviews and Dissemination (CRD) report?

How to Interpret a CRD Report

A Centre for Reviews and Dissemination (CRD) report should be interpreted as a systematic review that synthesizes evidence on healthcare interventions, requiring you to critically assess its methodology, risk of bias, quality of included studies, and the certainty of conclusions before applying findings to clinical practice.

Understanding What a CRD Report Is

A CRD report represents a systematic review—a structured synthesis of research evidence that includes:

• A comprehensive literature search across multiple databases 1
• Explicit eligibility criteria for study inclusion 1
• Assessment of risk of bias in individual studies 1
• Systematic presentation and synthesis of findings 1
• Meta-analysis when appropriate (though not all systematic reviews include this) 1

Key Sections to Evaluate

Methods Section - The Foundation of Quality

Examine the search strategy and study selection process first, as inadequate methods here undermine all subsequent conclusions. 1

• Verify that multiple databases were searched with explicit dates of coverage 1
• Check if the full electronic search strategy is provided for at least one database, allowing reproducibility 1
• Confirm the study selection process is clearly described, including screening and eligibility assessment 1
• Look for the PRISMA flow diagram showing numbers of studies screened, assessed, excluded (with reasons), and included 1

Risk of Bias Assessment - Critical for Validity

The credibility of conclusions depends entirely on whether bias was properly assessed in included studies. 1

• Identify which tool was used to assess risk of bias (e.g., Cochrane Risk of Bias tool for RCTs) 2
• Check if bias assessment was performed at the study or outcome level 1
• Verify that results of bias assessments are clearly presented for each included study 1
• Assess whether high-risk studies were handled appropriately in the synthesis 1

Results and Synthesis - Where Evidence Meets Interpretation

Focus on the certainty of evidence ratings (often using GRADE), not just statistical significance, as this determines how confidently you can apply findings. 1

• Review the characteristics of included studies (populations, interventions, comparisons, outcomes) 1
• Examine forest plots if meta-analysis was performed, noting effect sizes and confidence intervals 1
• Check for heterogeneity measures (I² statistic) when studies were pooled 1
• Look for assessment of publication bias or selective reporting 1
• Prioritize the certainty of evidence rating (high, moderate, low, very low) over p-values alone 1

Common Pitfalls to Avoid

Conflicts of Interest Can Bias Recommendations

Be particularly cautious when authors have financial conflicts of interest, as these are associated with more favorable recommendations regardless of the actual evidence. 1

• Check the funding sources and author disclosures 1
• Financial conflicts of interest are associated with biased conclusions in systematic reviews 1
• Non-financial conflicts (e.g., specialty interests) may also influence interpretation 1

Distinguishing Primary from Secondary Outcomes

Do not give equal weight to all reported outcomes—secondary outcomes require more cautious interpretation, especially without correction for multiple testing. 1

• Identify which outcome was pre-specified as primary 1
• Secondary outcomes should be interpreted more conservatively 1
• Be wary of selective reporting of only positive secondary outcomes 1

Quality of Included Studies Matters More Than Quantity

A systematic review including many low-quality studies provides less reliable evidence than one with fewer high-quality studies. 2

• Assess whether included studies used appropriate designs for the question 2
• Check if observational studies were necessary or if RCTs exist 3
• Note whether the review authors appropriately downgraded certainty for methodological limitations 1

Applying Findings to Clinical Practice

Match the Evidence to Your Patient Population

• Verify that included studies enrolled patients similar to yours in age, disease severity, and comorbidities 1
• Check if interventions were delivered in settings comparable to your practice 1
• Assess whether outcomes measured align with what matters to your patient (mortality, morbidity, quality of life) 1

Consider the Certainty of Evidence

High certainty evidence warrants confident application; low or very low certainty evidence requires shared decision-making acknowledging substantial uncertainty. 1

• High certainty: Further research very unlikely to change confidence in effect estimate 1
• Moderate certainty: Further research likely to have important impact 1
• Low certainty: Further research very likely to have important impact 1
• Very low certainty: Any estimate of effect is very uncertain 1

Look for Subgroup Analyses Relevant to Your Patient

• Check if effects differ by patient characteristics, intervention intensity, or duration 1
• Be cautious with subgroup findings unless they were pre-specified 1
• Recognize that post-hoc subgroup analyses are hypothesis-generating, not confirmatory 1

Red Flags Indicating Lower Quality

• Missing or incomplete search strategy 1
• No assessment of risk of bias in included studies 1
• Pooling of studies with substantial clinical heterogeneity without justification 1
• Conclusions that overstate the certainty of findings 1
• No discussion of limitations 1
• Failure to assess publication bias when appropriate 1