MCID, PASS, SCB and MDC, explained

MCID — minimal clinically important difference — is the smallest change in score that a patient would recognise as meaningful. It answers "has this patient noticeably improved," and it is defined on the change between two time points, not on a single score.

PASS — patient acceptable symptom state — is a single-timepoint threshold. It asks "is the patient’s current state good enough," independent of where they started. A patient can exceed the MCID and still sit below the PASS: improved, but not yet at a state they would call acceptable.

SCB — substantial clinical benefit — is a higher bar than MCID. Where MCID captures the smallest change a patient notices, SCB captures a change large enough that the patient would call it a substantial improvement, not just a perceptible one.

MDC — minimal detectable change — is not a clinical judgement at all. It is a statistical one: the smallest change that exceeds the instrument’s own measurement error, given its test-retest reliability. A change smaller than the MDC cannot be distinguished from noise, whatever the MCID says.

Every one of these four values is derived in a specific patient population, at a specific follow-up interval, and using a specific comparison. A KOOS MCID measured six months after ACL reconstruction in athletes in their twenties is not the same number as a KOOS MCID measured a year after total knee arthroplasty in patients in their seventies — and both figures circulate under the same three letters.

This is why every threshold in the database on this site is published with the population and follow-up interval it came from, rather than as a single bare number. Using a threshold from the wrong population is a common and avoidable error.

Suppose a patient’s KOOS Pain subscale rises from 55 to 68 after total knee arthroplasty at the one-year mark. The published MCID for that population and interval is roughly 8 to 10 points, so a 13-point rise clears it — the patient has changed enough to notice. Whether 68 also clears the PASS threshold for that population (roughly 78) is a separate question, and in this example it does not: the patient has genuinely improved but has not yet reached what patients in that cohort call an acceptable state. Both statements can be true at once, which is exactly why the two figures are reported separately rather than folded into one.