Translating an effect estimate into a shared decision — OSCE

Station brief (8–10 minutes)

Help the parents decide whether to start the preventive therapy by computing the trial's effect honestly and applying it to their five-year-old. Address the online "50 percent" figure, compute the absolute benefit and the number needed to treat from the event rates, read the confidence interval for the preschool subgroup, and weigh applicability to the child. Do not invent jurisdiction-specific thresholds or specific survival statistics. [1]

Tasks for the candidate

1. From the event rates, compute the relative risk reduction, the absolute risk reduction, and the number needed to treat, and contrast the absolute figure with the online "50 percent". [2] [3]
2. Read the confidence interval whose lower bound crosses unity in preschool children, and state what that means for the certainty of the benefit in this child. [5] [4]
3. Assess the applicability of the trial's number needed to treat to a five-year-old, and explain how it would change with a different baseline risk. [2] [3]
4. Integrate the evidence with the family's values through shared decision-making, and arrange follow-up. [6] [1]

Expected performance

Must hit. Relative risk reduction computed as 50 percent, absolute risk reduction computed as 10 percent, and number needed to treat computed as 10, each shown with working; the online figure reframed honestly as a relative reduction that hides the baseline; the null-crossing confidence interval in preschool children read as indeterminate rather than positive; the number needed to treat tailored to the child's own baseline risk; the decision integrated with the family's values; a written summary and follow-up arranged. [2] [3]

Merit. Acknowledges the online figure without dismissing the parent; states the null value of 1 for ratio measures and reads the interval against it; recomputes the number needed to treat for the child's baseline risk rather than borrowing the trial average; places the single trial in the totality of the evidence and rates the certainty; uses teach-back and a decision aid for the preference-sensitive choice. [5] [6]

Fail. Quotes the 50 percent relative figure alone and calls the therapy transformative; ignores the null-crossing confidence interval in preschool children and reports a positive result; applies the trial's number needed to treat to the child without weighing baseline risk; gives no absolute measure, no certainty rating, no shared decision, and no follow-up. [4] [6]

Sample candidate structure

"Thank you both for coming in. The '50 percent' you read online is a relative figure, which means the therapy reduced the complication by half compared with not taking it, but on its own it looks bigger than the benefit your child will actually feel. Let me show you the absolute numbers from the trial. Out of 1,000 children who did not take the medicine, 200 had the complication, and out of 1,000 who did take it, 100 did. So for every ten children treated, one complication is prevented — that is the number needed to treat. The confidence interval around that benefit crosses the line of no effect in preschool children like yours, which means the data cannot rule out that the benefit is smaller, or even absent, at this age. Let us weigh that honestly against your values and the daily burden of the medicine, and decide together."