PROMs in Clinical Practice

For Medical Students

What It Tests

A patient reported outcome measure records the patient's own account of their symptoms and function, so it measures a different thing from the examination rather than a better or worse version of it. 4.18.6 covers functional limitation.

The examination localises and grades, and the questionnaire describes the burden and the effect on life. A decision usually needs both. 4.1.4 covers the structured approach.

It matters because the two can disagree, and a disagreement is information rather than an error. 4.2.10 covers severity.

How It Is Performed

Choose the instrument for the population and the question rather than by habit. 4.22.5 covers the Oxford Knee Score.

Give it to the patient to complete themselves, before the consultation.

Collect a baseline, because every one of these instruments is interpreted as a change. 4.1.8 covers examination documentation.

Collect it at defined points rather than opportunistically, so the series is comparable.

Check completeness, and leave an incomplete questionnaire unscored.

Record it with the examination rather than in place of it. 4.1.4 covers the structured approach.

Record the activity level alongside a symptom score, since the score is reported against it. 4.22.4 covers the Tegner scale.

What a Positive Means

A change from a baseline is the result these instruments are built to produce.

A score that improves while the examination is unchanged means the patient's experience changed, which is usually the point of the treatment. 4.18.6 covers functional limitation.

A score that does not improve after a technically successful operation needs explaining rather than discounting. 4.20.8 covers the replaced knee.

A disagreement between the examination and the score identifies something neither has described, which is where the consultation goes next. 4.2.10 covers severity.

An incomplete or unbaselined score is not interpreted. 4.1.8 covers examination documentation.

MRCS Level

How It Is Performed

In a viva this is the question about outcome measurement, and the expected answer places these instruments relative to the examination.

The expected points are the choice of instrument, the self completion, the baseline, the defined collection points, the completeness, the paired examination record and the activity level.

Say that the instrument is chosen for the population. 4.22.1 covers the IKDC.

Say that a baseline is required, and give the reason. 4.1.8 covers examination documentation.

Say that a score does not replace the examination. 4.1.4 covers the structured approach.

What a Positive Means

The reasoning expected is that structural severity, examination findings and patient reported burden are three different quantities that correlate loosely. 4.1.6 covers that relationship.

A treatment is aimed at one of those, so the outcome is measured on the one it was aimed at. 4.18.7 covers the arthroplasty assessment.

Scores are ordinal constructions, so the distance between two values is not a physical quantity and arithmetic on them is a convention. 4.2.10 covers severity.

Expectation shapes the reported score, which makes it a variable to record rather than a bias to dismiss. 4.21.8 covers return to play decision making.

How Good It Is

These instruments are reproducible for the same patient over time and that is the use they support.

Their best contribution is making the patient's experience recordable, comparable and auditable, which no examination finding is. 4.1.8 covers examination documentation.

They do not localise, diagnose or grade, and they are not interchangeable with each other. 4.22.3 covers the Lysholm score.

Imaging, examination and questionnaire answer three questions, and none substitutes for another. 4.1.6 covers that relationship.

No figures appear for any accuracy comparison on this page, under decision 023, and the comparison is the claim being made. No psychometric property of any instrument is stated anywhere in this chapter, because each is a published quantity this environment cannot check.

FRCS Level

How Good It Is

At decision level these instruments support audit, registry comparison, research and the individual before and after record.

A service collecting a baseline and a follow up has evidence of its results, and one that does not has impressions. 4.1.8 covers examination documentation.

A registry comparison between services is only as good as the comparability of their populations, which the scores do not describe. 4.22.5 covers the Oxford Knee Score.

An individual's change is the use with the least interpretive risk, because the patient is their own control. 4.18.6 covers functional limitation.

The choice of instrument fixes what can later be asked of the data, which makes it a decision rather than a formality. 4.22.1 covers the IKDC.

Where It Misleads

The commonest error is collecting scores without a baseline, which makes the whole collection uninterpretable. 4.1.8 covers examination documentation.

Scores are compared across instruments as if they shared a scale. 4.22.2 covers the KOOS.

A population difference between services is read as a quality difference. 4.22.5 covers the Oxford Knee Score.

A ceiling effect is read as an absence of change. 4.22.4 covers the Tegner scale.

An unvalidated instrument is used as if it were validated. 4.22.7 covers this site's own score.

Fellowship Level

Where It Misleads

At fellowship level the harder problems are about what these scores support rather than how to collect them.

A single number aggregates domains, so two patients with the same score have different problems and a mean score across a service describes nobody. 4.22.2 covers the KOOS.

Importance thresholds are published quantities that differ by instrument and by population, and none is stated in this chapter. 4.2.10 covers severity.

Expectation predicts the reported outcome strongly enough that a comparison between populations with different expectations is not a comparison of treatment. 4.21.8 covers return to play decision making.

Missing data in a collected series is not missing at random, since the patients who stop returning questionnaires are not the same as those who continue. 4.1.8 covers examination documentation.

An instrument validated in one population and applied in another produces numbers whose meaning is unknown rather than approximate, which is the distinction this chapter has tried to hold throughout. 4.22.7 covers this site's own score.