Medical research

Know which findings belong in the same comparison.

Identify reports from the same trial and incompatible outcome comparisons before they become inputs to a research synthesis.

Evidence-synthesis teams, information specialists, and methods reviewers

Evidence Comparison WorkbenchIllustrative research data

Trial family

HARBOR · 480 randomized

Main report · day 30

Extended follow-up · day 90

Registry fixture

One study · three retained sources

Check comparison

Day 30

24/240 vs 36/240

All randomized

Day 90

52/226 vs 56/229

Observed outcomes

Same trialDifferent follow-upDifferent analysis population

The question

“Keep trial families separate from publications, and flag different outcomes, follow-up periods, and analysis populations.”

This illustrative workflow uses fictional research sources, no patient records, and makes no clinical recommendation.

Compare studies, not only publications.

Link HARBOR’s main report, extended follow-up, and registry fixture without discarding any source or double-counting participants.

Useful result

Five sources resolve to three trial families, not four independent trials.

Find out what the disagreement really is.

Compare day 30 with day 90 and all-randomized denominators with observed-outcome denominators.

Useful result

The apparent conflict is not two independent estimates of the same endpoint.

Bring back a conclusion the team can examine.

Return candidate inputs, source excerpts, exclusions from this comparison, and missing evidence.

Useful result

HARBOR A-30 and INLET C-30 remain candidates; QUAY stays in the review but not this synthesis.

The evidence behind the result.

Numbers stay connected to their definitions, source records, and limits.

Sources

4 reports + 1 registry fixture

Trial families

HARBOR reports linked

HARBOR day 30

24/240 vs 36/240

All randomized

HARBOR day 90

52/226 vs 56/229

Observed outcomes

Inspect the source trail.

This demonstration uses internally consistent fictional fixtures. It is not a customer result or live external execution.

R-A2 p.2: follows the 480 participants randomized in HARBOR

R-A2 Table 3: N reflects observed outcomes at 90 days

R-B1: medication-related readmission, not all-cause readmission

The work moves. Responsibility stays clear.

Originates

Review lead

Contributes

Methods reviewer

Decides

Information specialist

Decision remains with people

Retain two candidate day-30 records for further methods review. Do not calculate a pooled effect.

Bring a real medical research workflow.

See how Prajvis could create useful functionality around the way your team already works.

Request a demo Get started