Choose a synthesis when the reported numbers cannot support pooling

Separate the information available from the answer desired

Cochrane Chapter 12, checked 7 November 2026, rejects significance-based vote counting. It describes alternative methods with different data requirements and questions, including descriptive effect summaries when variances are unavailable. This lesson uses that distinction for an original business review, not as a clinical instruction.

List the approved comparison and the outcome the review intends to explain. Then inspect what each eligible study actually supplies. Preserve the original result location and identify which entries are extracted, calculated by an approved method or still unresolved. Do not treat an empty cell as a zero or a neutral finding. The purpose is to discover which synthesis is supportable before writing the headline.

Keep a request for additional published detail separate from the analysis record. A hoped-for reply cannot be treated as information already obtained. Set a practical review date and state how the team will proceed if the information remains unavailable.

Choose between genuinely different questions

Consider a fictional review with five instruction studies. Four give estimates on the same scale, but only two provide usable uncertainty information. The fifth provides a narrative statement without an extractable effect. An analyst should determine whether additional information can be recovered legitimately before the team chooses its final approach.

If that effort fails, distinguish describing the reported estimates from estimating a pooled effect. A descriptive display may help the business understand the range of reported results without pretending to supply the missing precision. Do not combine the two fully reported studies automatically and let the remaining evidence vanish from the discussion. Explain what any restricted analysis represents and what was left outside it.

Write the alternative questions in ordinary language. Which question would influence the decision, and does the available reporting answer it? A narrower useful answer is preferable to a broad answer constructed from unsupported arithmetic.

Build a display that exposes the limitation

Use a consistent row structure with the study identity, relevant comparison, outcome window, reported estimate where available and a plain explanation of missing information. Keep studies with limited reporting visible. A blank estimate position should not appear to be evidence of no effect. Distinguish an unavailable number from an outcome that the study did not assess.

Arrange the display according to an agreed substantive grouping rather than moving favorable results to the top. Add a brief statement explaining what the display permits the reader to compare. If a chart is used, have the analyst check its scale and labels. Avoid decorative precision, such as adding an invented decimal place to make heterogeneous source reporting look uniform.

Ask a reader unfamiliar with the review to state what conclusion they would draw. If they count attractive colors or assume an unreported result was unfavorable, change the display. The design must make the actual evidential limitation understandable without requiring the reader to inspect the extraction workbook.

Report the method choice and remaining need

Explain why the chosen synthesis was used, what information prevented a different approach and how the limitation affects the business conclusion. Keep this explanation beside the results rather than only in a methods appendix. The team should be able to distinguish an evidential constraint from an editorial preference for a simpler visual.

For the next update, name the exact information that would permit reconsideration. Do not promise that any additional study will solve the problem. A new report could repeat the same omission or address a different comparison. Preserve the current method choice and the trigger for reopening it so an update does not silently change the meaning of the summary.

Finish with a recommendation proportional to the answer actually obtained. The review may support a targeted experiment, a limited operational trial or a decision to defer a broad claim. Those are business judgments; state them separately from what the synthesis demonstrates.

Make the reasoning concrete.

A fictional editor prematurely drafts a decisive headline from the incomplete packet.

The editor holds the tally and commissions the supported display. The fictional packet is not evidence about an actual intervention.

Take the questions into your next brief.

Use these prompts in your own notes. Nothing is submitted from this page.

1. Information

   What is genuinely available for each eligible study?

2. Recovery

   What additional detail can be obtained without invention?

3. Question

   Which narrower question is answerable?

4. Display

   How will limited reporting remain visible?

5. Update

   What information would justify revisiting the method?

Watch for the wrong conclusion.

• Treating incomplete reporting as an unfavorable result.
• Hiding studies that cannot enter the preferred calculation.
• Choosing the headline before the synthesis method.

Sources & scope

Original adult business scenarios are explicitly fictional. Methodological statements and applications remain source-derived. Qualified analysts and review teams own formal assessment. No clinical guidance, actual research operations or publication occurs.

• Cochrane Handbook, Chapter 12: Synthesizing and presenting findings using other methods →Primary technical documentation · Supports: Supports method choice under reporting limitations and rejection of significance-based vote counting.