Or perhaps it should be ‘What would Mohammed want?’, or Alyx, or Devine, or Vladimir … when we’re asking clinical questions, do we really think what outcome our patient and family want, or do we just think what we think they would want? To put it another way – how close are the outcomes that patients and families want and the outcomes we proxy for them?

The early 1980s saw rheumatologists starting to grasp the problems of outcomes, starting with the need to have similar outcomes from their studies so they could understand what different trials meant, and think about comparisons. It was only in 2002 that patients made their way into this, most forward thinking, group of trial-minded clinicians. But how far have paediatricians come to thinking this way?

Two years ago now, a group from Liverpool published a systematic review of studies that examined how to decide which outcomes were important for clinical trials. (As an aside, it’s still taking a very long time to translate knowledge into practice, as an eminent paediatrician blogs on here.) They found 25 studies that had reported on the process – of these, only three included parents in the decision making and none included children. None. We have never studied how to include child-important outcomes in our research, have barely done it for parents and have only tickled the surface of how to involve clinicians.

Yet we know that proxies can give very different results for quality of life measures, and arthritis indices so we should be cautious about just believing we know what folk want. Perhaps – if we’re framing questions for specific situations – we might even want to ask our patients and their families themselves, and see if our PICO can actually answer their question.