Disease Severity and Outcome Measures

Severity is best judged by keeping several domains separate and then weighing them together. Symptoms describe the burden the patient experiences. Function and participation describe what the athlete can do and how far that falls short of their sport. Examination findings add objective signs such as the size of an effusion or the degree of lost range. Structural findings come from imaging and validated grading systems, for example the Kellgren-Lawrence grading of osteoarthritis on radiographs or the graded classification of muscle injuries. Prognosis, the likely course and time to recovery, is a separate domain again. A structural grade is not the same thing as an overall severity grade.

Two cautions matter. First, grading systems should only be used where their reliability, validity and clinical purpose are understood, because not every muscle or ligament grading system predicts recovery accurately. Second, imaging severity and symptom severity frequently diverge: marked degenerative change can be painless, and a normal scan does not exclude a disabling problem, so structural grading informs rather than overrides the clinical picture. In SEM the severity that usually matters most is functional, which keeps the assessment relevant and sets up meaningful measurement of change.

Outcome measures sit within a wider family called clinical outcome assessments, which are conventionally divided into four types. Patient-reported outcome measures (PROMs) come directly from the patient without interpretation by anyone else. Clinician-reported outcome (ClinRO) measures require a trained clinician to exercise judgement in interpreting an observable sign, such as rating the severity of an effusion or a neurological sign. Observer-reported outcome measures come from someone who observes the patient in daily life, such as a parent, and involve no clinical judgement. Performance outcome (PerfO) measures are obtained when the patient completes a standardised, repeatable task. This last category is commonly misplaced: goniometric range of movement, dynamometry and hop tests are performance-based measurements rather than clinician-reported outcomes.

Patient-reported measures are then subdivided by scope. Generic measures such as the EuroQol five dimension questionnaire (EQ-5D) describe health broadly and allow comparison across conditions. The Patient-Reported Outcomes Measurement Information System is not a single questionnaire but a system of domain-specific item banks, short forms and computer-adaptive tests. Region-specific measures cover an anatomical area and include the Knee injury and Osteoarthritis Outcome Score, the Disabilities of the Arm, Shoulder and Hand questionnaire and the Foot and Ankle Ability Measure. Condition-specific measures target one disorder, such as the Victorian Institute of Sport Assessment scores for Achilles and patellar tendinopathy. Sport-specific instruments are often needed to avoid ceiling effects in high-functioning athletes. Measures also vary in length, from single-item tools such as the Single Assessment Numeric Evaluation and the visual analogue scale for pain to detailed questionnaires. Choosing well means matching the measure to the condition, the population and the question, and using it only within the group in which it was validated.

Three ideas are needed, and the terminology repays precision. The minimal detectable change (MDC), sometimes called the smallest detectable change, concerns measurement error: it is the smallest change that lies beyond the error of the instrument. The minimal important change (MIC) concerns meaningful within-person change, the smallest change a patient perceives as beneficial, while a minimal important difference usually refers to a between-group difference. The term minimal clinically important difference is widely used but applied inconsistently across the literature, which is why the more precise terms are preferred. The patient acceptable symptom state (PASS) is different in kind: it is a threshold describing a state in which the patient considers themselves well enough, and depending on how the instrument is scored the acceptable state may lie above or below the cut-off.

Two practical points follow. The MDC and the MIC are not necessarily in a fixed order, and importance and measurement precision can genuinely conflict: a change may matter to a patient yet still fall within the instrument's measurement error, so the MIC does not by itself prove a change is real. And these thresholds are context-specific, varying with the population, the baseline severity and whether an anchor-based method tied to patient-rated change or a distribution-based method tied to statistical spread was used. Rather than asking only whether a result is statistically significant, interpret the effect estimate and its confidence interval against an appropriate important-difference threshold, remembering that a group mean falling below a threshold does not prove that no individual benefited.

In practice a measure is chosen at the start, repeated at sensible intervals and recorded at discharge, so that the trajectory, not just the endpoint, is visible. Interim measurement can identify an unexpectedly poor trajectory and prompt a change of plan while there is still time to act; observational data in some rehabilitation populations, notably non-specific low back pain, report an association between interim measurement and better discharge outcomes, which is a reason to measure along the way rather than proof of a causal effect.

Return-to-sport decisions should never rest on a single score. They combine symptoms and patient-reported measures, objective performance testing, psychological readiness assessed with tools such as the Anterior Cruciate Ligament Return to Sport after Injury scale, graduated sport-specific exposure, tissue and clinical status, and a shared discussion of risk. Several pitfalls recur. Modifying a validated questionnaire by changing wording, dropping items or rescaling compromises its measurement properties, and the modified version requires fresh validation before published scoring and thresholds can be reused. Ceiling effects are common in young, high-functioning athletes, where a general measure may show a top score despite a genuine deficit. Using a measure outside the population in which it was validated, or aggregating scores in ways the developers did not intend, produces misleading numbers. Used carefully, outcome measures support shared decision-making, service evaluation and research.

• •Keep severity domains separate: symptoms, function and participation, examination, structural findings and prognosis. A structural grade is not an overall severity grade.
• •Imaging severity and symptom severity often diverge, so structural grading informs rather than overrides the clinical picture.
• •There are four clinical outcome assessment types: patient-reported, clinician-reported, observer-reported and performance-based.
• •Goniometry, dynamometry and hop tests are performance-based, not clinician-reported.
• •The minimal detectable change concerns measurement error; the minimal important change concerns meaningful within-person change. They are not in a fixed order.
• •The patient acceptable symptom state is a final-state threshold, not an amount of change.