Reading a paediatric trial result for a shared decision — OSCE

Station brief (8–10 minutes)

Help the parents decide whether to start the preventive therapy by reading the abstract result honestly and weighing it for their child. Address the online 'significantly shortens flare duration' figure, the small effect size, the confidence interval that crosses zero, the limits of a p-value on its own, and the fact that the study was small. Do not invent jurisdiction-specific thresholds or specific drug doses. [4]

Tasks for the candidate

1. Read the effect size and its confidence interval before the p-value, and explain what an interval crossing zero means. [9] [5]
2. Reframe the 'significant' figure with the absolute effect size and its precision, and explain why the result is inconclusive rather than positive. [4] [9]
3. Weigh whether a small, underpowered study should change practice, defining the relevant errors and power. [3]
4. Integrate the appraisal with the family's values through shared decision-making, and arrange follow-up. [4] [5]

Expected performance

Must hit. Effect size and 95 percent confidence interval read before the p-value; an interval crossing zero identified as not statistically significant at the 0.05 level, compatible with both benefit and no effect; the small sample identified as underpowered so the result is inconclusive, not negative; the absolute effect size given in plain terms; Type I and Type II errors and power named correctly; the decision integrated with the family's values; written summary and follow-up arranged. [9] [4] [3]

Merit. Acknowledges the online figure without dismissing the parent; distinguishes statistical significance from clinical importance explicitly; explains that absence of evidence is not evidence of absence; uses teach-back and frames the choice as preference-sensitive given the uncertainty; names the plan to revisit the decision as larger evidence appears. [5] [3]

Fail. Quotes the p-value of 0.04 alone as proof the therapy works; reads the point estimate while ignoring the null-crossing interval; declares the therapy useless because the result is non-significant; conflates significance with importance; gives no absolute effect size, no shared decision, and no follow-up. [4] [9]

Sample candidate structure

"Thank you both for coming in. The figure you read said the therapy 'significantly shortens flare duration,' and I can see why that sounds reassuring, but I need to read the whole result with you before we decide. The headline number behind that word 'significant' is small — on average about half a day off a flare — and the range of plausible truth around it, the confidence interval, is wide and runs from a real benefit all the way down to essentially no effect. That means the study cannot actually tell us the therapy works; it was too small to be sure. So rather than a clear yes or no, what we have is genuine uncertainty, and that makes this your decision to share with me. Let me lay out the absolute benefit, the burden of a daily treatment, and what we'd watch for, and then I'd like to hear what matters most to you and to your child." [4] [9] [3]