Shoulder PROMs in Athletes: Which Outcome Measures Actually Hold Up

Ask ten physiotherapists which shoulder outcome measure they trust for an overhead athlete and you'll get ten different answers, usually followed by a caveat. That caveat is worth taking seriously. Most of the questionnaires clinicians reach for in shoulder rehab were built and validated in populations that look nothing like a county-level swimmer, a club cricketer, or a semi-professional thrower returning from a rotator cuff problem. Understanding where these tools hold up and where they don't changes what you should actually be measuring in clinic.

The evidence base wasn't built with athletes in mind

The clinical practice guidelines that underpin most shoulder rehabilitation decisions were synthesised from trials in general orthopaedic and occupational shoulder pain populations, not competitive athletes managing sport-specific load (Philadelphia, 2001). That matters because the functional demands, baseline capacity, and recovery expectations of a 34-year-old office worker with subacromial pain and a 19-year-old overhead thrower are not comparable, yet the outcome tools applied to both are frequently the same.

The problem compounds when you look across the trial literature for a single shoulder condition. Reviews assessing therapeutic trials for conditions like adhesive capsulitis routinely find inconsistent selection and reporting of outcome measures between studies (Fares et al., 2023). If the trials underpinning your evidence base can't agree on what to measure, comparing your own patient's trajectory against a published benchmark is shakier ground than it looks on paper.

Reliability problems show up before validity ones

Before asking whether a shoulder outcome measure captures the right construct, ask whether it produces the same answer twice. This is where a lot of routine shoulder assessment quietly breaks down. Systematic review evidence on specific physical examination tests for shoulder pathology shows wide variation in diagnostic reliability depending on the test, the rater, and the protocol used (Lange et al., 2017). Even measurement approaches that look objective on the surface, such as acromiohumeral distance measurement in rotator cuff tendinopathy, lack a single reliable, clinically applicable method that's been consistently validated for routine use (McCreesh et al., 2015).

The practical takeaway is uncomfortable but important: a shoulder outcome is only as defensible as the measurement process sitting underneath it. If two clinicians in the same clinic can't reproduce each other's findings on the same patient, the PROM or performance score attached to that assessment inherits the same instability, regardless of how well-validated the questionnaire itself is.

• Standardise positioning, instruction, and scoring criteria before you trust a shoulder measure across sessions
• Treat any single-session physical exam finding as provisional until it's reproduced
• Prioritise measures with published reliability data specific to the population you're testing, not just validity data from a different cohort

Performance-based functional tests close a gap questionnaires leave open

Self-report questionnaires tell you how a patient feels about their shoulder. They don't tell you what the shoulder can actually do under load. That distinction matters most in athletes, where the return-to-sport decision hinges on capacity, not comfort. A systematic review and meta-analysis of performance-based functional tests in shoulder pathology found these measures provide psychometric information that self-report tools don't capture on their own, supporting their use as a complement to, not a replacement for, questionnaire-based outcomes (Nazari et al., 2020).

This is echoed in return-to-sport specific testing. A shoulder and arm return-to-sport test battery designed for athletic populations has shown acceptable reliability across its component tests, which means it can be trusted to track genuine change rather than measurement noise as an athlete progresses through rehab (Olds et al., 2019). For time-pressed clinicians, the message is straightforward: pair a self-report score with a performance-based capacity test, and don't rely on either one in isolation to clear an athlete for return to sport.

Why a generic questionnaire score misses what matters to the athlete in front of you

Generic shoulder scales are built around items that apply to the average shoulder pain patient: reaching overhead, sleeping on the affected side, lifting light objects. An athlete whose problem only appears at end-range external rotation under high angular velocity, or at the top of a throwing motion, can score close to the ceiling on a generic scale while still being nowhere near ready to compete. The scale isn't wrong, it's just answering the wrong question for that patient.

This is why an individualised approach to patient-reported function performs better for athletic populations than a fixed generic item list. Asking a patient to identify and rate their own specific functional limitations, rather than scoring them against a standard set of everyday tasks, captures the demand that actually matters to them. Pain also needs more nuance than a single 0-10 rating at rest. Frequency, intensity, and duration behave differently across a rehab timeline, and averaging the three gives a more complete picture of where an athlete's pain experience is actually heading, rather than a single number that can mask whether pain is becoming less frequent, less intense, or shorter-lived.

Building a shoulder assessment protocol that holds up under scrutiny

None of this means throwing out PROMs. It means being deliberate about which ones you use and pairing them with objective performance data that's reliable enough to trust session to session. A defensible shoulder protocol for an athletic caseload looks something like this:

1. Patient-specific function. Let the athlete identify their own limiting activities rather than scoring them against a generic item bank.
2. Three-dimensional pain tracking. Capture frequency, intensity, and duration separately rather than a single resting pain score.
3. Range of movement. Track shoulder range with any inclinometer tool, measured the same way every session.
4. Strength. Use handheld dynamometer data benchmarked against peer-reviewed normative literature, expressed as a percentage of expected score for age, sex, and weight, rather than a raw number with no context.
5. Power. For throwing athletes, a medicine ball throw distance measured with a tape measure gives a low-cost, technology-agnostic read on upper-limb power output that a questionnaire simply can't provide.
6. Longitudinal tracking. Plot PROM and performance data against each other over time. A rising PSFS score alongside a flat power output is a different clinical picture than both moving together.

The common thread across the reliability and psychometric literature is that measurement quality, not the sophistication of the tool, is what determines whether an outcome measure holds up. A simple test performed the same way every time beats a validated questionnaire applied inconsistently.

Making the shift from questionnaire-only to measured practice

Most clinics already collect some form of shoulder PROM. Far fewer pair it consistently with reliable, standardised performance data, which is exactly the gap the evidence above points to. Benchmark PS was built to close that gap without requiring expensive equipment: patient-specific function and three-dimensional pain scoring alongside range of movement, strength, and power testing, all tracked longitudinally against normative expectation. If your shoulder caseload includes athletes returning to sport, that combination gives you a far more defensible basis for the return-to-play decision than a questionnaire score on its own.

References

1. Philadelphia Panel. Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for shoulder pain. Physical therapy. 2001;81(10):1719-30. PubMed
2. Fares MY, Koa J, Abboud JA. Assessment of therapeutic clinical trials for adhesive capsulitis of the shoulder. JSES international. 2023;7(3):412-419. PubMed
3. Lange T, Matthijs O, Jain NB et al. Reliability of specific physical examination tests for the diagnosis of shoulder pathologies: a systematic review and meta-analysis. British journal of sports medicine. 2017;51(6):511-518. PubMed
4. McCreesh KM, Crotty JM, Lewis JS. Acromiohumeral distance measurement in rotator cuff tendinopathy: is there a reliable, clinically applicable method? A systematic review. British journal of sports medicine. 2015;49(5):298-305. PubMed
5. Nazari G, Lu S, MacDermid JC et al. Psychometric Properties of Performance-Based Functional Tests in Patients With Shoulder Pathologies: A Systematic Review and Meta-analysis. Archives of physical medicine and rehabilitation. 2020;101(6):1053-1063. PubMed
6. Olds M, Coulter C, Marant D et al. Reliability of a shoulder arm return to sport test battery. Physical therapy in sport : official journal of the Association of Chartered Physiotherapists in Sports Medicine. 2019;39:16-22. PubMed