Imaging progress and function are two different clocks

Combining scan appearance and capacity into one recovery percentage adds two measures that do not move together.

Halfway through a rehabilitation, two things are being tracked. What the tissue looks like on a scan, and what the limb can do. They are frequently reported as a single percentage, and that single percentage is an addition of two quantities on different scales moving at different rates.

Structure lags. A repaired tendon can look untidy on imaging for many months after it has stopped hurting and started tolerating load, and there is no rule saying the picture must catch up before the function does. Capacity can also lead in the other direction: a limb that looks decent and cannot produce force, usually because the muscle around it has not been asked to work in ten weeks. Two clocks, running independently, occasionally in opposite directions.

When you merge them into one figure you inherit the worst property of both. A good scan can carry a weak limb over the line, and a limb that is working fine can be held back by an image that was always going to look like that. Neither error is visible in the merged number, because merging is exactly the operation that removes the disagreement, and the disagreement is the clinically interesting part.

There is also a weighting question nobody answers out loud. How many per cent of the composite is the scan worth? Whatever number is chosen, it was chosen by somebody, and it is doing more to the final figure than most of the measurements are. I have never seen those weights written down. They live in the head of whoever built the spreadsheet, and they change when that person leaves.

Keep the two lines separate on the same chart. Let them disagree in public. When the picture is behind the capacity, that is a conversation about how much risk the tissue can carry; when the capacity is behind the picture, it is a conversation about training. Those are different conversations with different people and a composite score prevents either from starting.

The pull towards one number comes from wanting a single answer to a single question, which is whether he plays. Fair enough as a demand.

But the answer to that question is a judgement made by people who have looked at both clocks, and dressing the judgement up as an arithmetic output does not make it more defensible. It only makes it harder to argue with.