Appraising a Body That Disagrees With Itself

In practice

The pressure-injury search returns fifty-one records. Rapid critical appraisal is phase one rather than the whole job: each study is asked whether it is valid, whether its results are reliable, and whether it applies to these patients. Nineteen survive as keeper studies. Phase two evaluates them together, pulling the same elements from each into one table — design, sample, effect, quality. The disagreement becomes visible and specific. Two trials show a reduction in incidence, four observational studies show none, and one shows harm through disturbed sleep. Phase three synthesises rather than counts. Level and quality together determine strength, so two well-executed trials are not outvoted by four studies that can only show association. Consistency is examined rather than demanded: the studies that found nothing used discretionary repositioning as their comparison, which is what the unit already does, while the trials that found an effect compared against no schedule at all. The two bodies were answering different questions. Magnitude is read alongside direction. Phase four states the recommendation with its confidence: a documented schedule is supported for patients at high risk, the certainty is moderate, and the harm signal is real enough to monitor.