Patient-reported outcome measure

Oxford Shoulder Score

The Oxford Shoulder Score is a 12-item patient-reported questionnaire about shoulder pain and everyday function over the past four weeks, each item scored 0 to 4 and summed to a total of 0 to 48 in which 48 represents the best possible shoulder — it is licensed by Oxford University Innovation, so the item wording is not reproduced here.

Licensed instrument12 items~5 min0–48, higher is better

This instrument is not reproduced here

The Oxford Shoulder Score is copyright the University of Oxford and is administered by Oxford University Innovation, which operates an active licensing programme. Non-commercial academic and clinical use is generally granted on registration, and commercial use — including use inside a commercial product, app or calculator — requires a paid licence. Because we cannot licence the items on a reader’s behalf, the questionnaire itself is not reproduced on this page and no calculator is offered. Obtain a licensed copy directly from the rights holder; the scoring rule below is published openly and is reproduced in full so that a licensed copy can be scored from this page.

Obtain a licence from the rights holder

What the Oxford Shoulder Score means

Oxford Shoulder Score scores run from 0 to 48, where a higher score is better. The bands below are conventional reporting categories, not diagnostic cut-offs.

Severe shoulder problem0–19
These groupings are the ones most commonly used in practice and in the United Kingdom national reporting. Check the wording in the current Oxford University Innovation user manual, which is supplied with the licence, before quoting them in a publication.
Moderate to severe20–29
Mild to moderate30–39
Satisfactory shoulder function40–48

How the score is calculated

Stated in full so the calculation can be independently checked, and reproduced by hand if needed.

Each of the twelve items offers five ordered response options. Under the current scoring, the best option scores 4 and the worst scores 0, and the twelve item scores are summed to give a total from 0 to 48, with 48 the best possible shoulder. The direction of the total is the opposite of the SPADI and the same as the ASES: a higher Oxford Shoulder Score is a better shoulder. The original 1996 publication scored each item 1 to 5 for a total of 12 to 60, in which a LOWER score was better; the developers restated the scoring in 2009 as 0 to 48 with a higher score better, and that is the version in use today. This is the single most important thing to check when reading an Oxford Shoulder Score in the literature or in a database — the two conventions run in opposite directions, and a total of 20 means a good shoulder under the old scoring and a poor one under the new.

Missing data

The developers’ rule is that up to two missing items may be replaced by the mean of the items the patient did answer, and the total then computed as usual. With three or more items missing the questionnaire is not scored. The user manual supplied with the licence is the authority for this; confirm it there before applying it to study data.

Instrument profile

Developer: Jill Dawson, Ray Fitzpatrick and Andrew Carr
Institution: University of Oxford, United Kingdom
First published: 1996
Custodian: Oxford University Innovation (formerly Isis Innovation), University of Oxford
Items: 12
Response format: Twelve items, each with five ordered response options scored 0 to 4
Score range: 0 to 48
Direction: Higher score = better
Recall period: The last four weeks
Completion time: About 5 minutes
Age range: Adults
Regions: Shoulder
Conditions: Rotator cuff tear, Subacromial pain syndrome, Shoulder osteoarthritis, Adhesive capsulitis, Cuff tear arthropathy
Procedures: Total shoulder arthroplasty, Reverse shoulder arthroplasty, Rotator cuff repair, Subacromial decompression
Languages: English, German, Dutch, French, Italian, Spanish, Swedish, Norwegian, Danish, Turkish, Chinese, Japanese, Korean, Brazilian Portuguese
Test–retest (ICC): High over short intervals in stable patients, consistently reported above 0.9 in the validation literature
Internal consistency: Cronbach’s alpha consistently above 0.9 across languages, reflecting a single underlying construct
Validity: Correlates strongly with the Constant score, the SPADI and the shoulder items of the SF-36, and is one of the two shoulder measures collected in the English national patient-reported outcome programme, which gives it an unusually large comparison dataset.
Responsiveness: Responsive after shoulder arthroplasty and rotator cuff repair. Published minimal important change estimates vary with the population and the anchor used, and we have not verified a single figure well enough to quote it here — take the threshold from a study whose population matches yours rather than from a general number.
Ceiling effect: Marked in recovered patients, who cluster at 44 to 48.
Floor effect: Uncommon; few patients score below 5.

Normative values

Reference values from published cohorts, for comparison against the score you calculated.

Population	Age	Mean	SD
Patients listed for shoulder arthroplasty, before surgery	—	Commonly in the high teens to low twenties	—
Asymptomatic shoulders	—	At or close to 48, with a marked ceiling effect	—

When to use the Oxford Shoulder Score

Well suited to

Shoulder arthroplasty and rotator cuff outcome reporting in the United Kingdom and Europe, where it is the expected measure and large comparison datasets exist.
Short, entirely patient-reported follow-up by post or online, where a clinician-administered score is impractical.
Any setting where a joint-specific rather than a general shoulder-and-arm measure is wanted — it is shorter and more specific than the DASH.

Limitations and poor fits

Shoulder instability in young athletes, where the items do not capture apprehension or recurrence — use the WOSI.
Without a licence. This is not a formality: the items are copyright and the rights holder actively administers them.
Comparing totals across studies without first establishing which scoring direction each used, because the 1996 and 2009 conventions run opposite ways.
Detecting further improvement in patients already scoring in the mid-forties, because of the ceiling effect.

Oxford Shoulder Score compared with other instruments

Instruments measuring a related construct, with the practical difference between them.

Instrument	Items	Range	What it adds
SPADI	13	0–100	The SPADI is a 13-item patient-reported questionnaire that measures shoulder pain and disability over the past week on two subscales — 5 pain items and 8 disability items — each rated 0–10 and transformed to a 0–100% score in which a higher score means more pain and greater disability.
ASES Score	11	0–100	The ASES score is a patient-reported shoulder index running from 0 to 100 in which a single 0–10 pain rating supplies 50 points and ten activities of daily living rated 0–3 supply the other 50, so that 100 represents a pain-free shoulder with no functional difficulty.
Constant Score	10	0–100	The Constant-Murley score is a 100-point clinician-administered shoulder assessment combining 35 subjective points for pain and daily activities with 65 objective points for measured range of motion and abduction strength, where 100 represents a normal, pain-free shoulder.
UCLA Shoulder Score	5	2–35	The UCLA Shoulder Rating Scale is a 35-point shoulder assessment scoring pain and function out of 10 each, active forward flexion and forward-flexion strength out of 5 each, and patient satisfaction as either 0 or 5, where a higher total is better and 34–35 is graded excellent.
SST	12	0–12	The Simple Shoulder Test asks twelve yes-or-no questions about everyday shoulder functions and scores the number answered yes, giving a total from 0 to 12 where 12 means all twelve functions are retained.
QuickDASH	11	0–100	The QuickDASH is an 11-item shortened form of the DASH that measures physical function and symptoms across the whole upper limb on a single 0 to 100 scale, where a higher score means greater disability.

Copyright and permitted use

Published openly because it is the question every clinician and researcher asks before using an instrument, and almost nobody answers it.

Licensed instrumentRights holder: Oxford University Innovation, University of Oxford

Rights holder’s own page

References

Dawson J, Fitzpatrick R, Carr A. Questionnaire on the perceptions of patients about shoulder surgery. Journal of Bone and Joint Surgery (British Volume). 1996.
Dawson J, Rogers K, Fitzpatrick R, Carr A. The Oxford shoulder score revisited. Archives of Orthopaedic and Trauma Surgery. 2009.

Editorial status

Not yet reviewedQueued for clinical review by Sam Tuffun, PT, DPT. Sign-off has not taken place — this is a placeholder assignment, not a completed review.