Oxford Hip Score

Oxford Hip Score in Hip Replacement Outcomes and Economic Evaluation

The Oxford Hip Score is the hip-specific questionnaire that English NHS hospitals and joint replacement studies use to record how much a hip replacement has relieved pain and restored everyday activity. Dawson and colleagues at the University of Oxford published it in 1996 for patients having total hip replacement, and it has since become the condition-specific outcome of the national patient-reported outcome programme in England. For a health economist the score matters in two ways. It is the health-gain measure used in national PROMs comparisons of providers. Because its total carries no preference weights, the score cannot enter a QALY calculation directly, so analyses that hold only Oxford Hip Score data need a mapping step before a cost-utility analysis can use them.

How the twelve hip questions were developed and are scored

The questionnaire was built for patients having one operation. Oxford University Innovation, which licenses the questionnaire, describes the measure as a short 12-item patient-reported instrument designed and developed to assess function and pain in patients undergoing hip replacement surgery.

Dawson and colleagues tested the new questionnaire in 220 patients assessed before total hip replacement and again six months afterwards. Patients also completed the SF-36 and, in some cases, the Arthritis Impact Measurement Scales, while a surgeon rated each hip on the Charnley hip score. The single derived score showed high internal consistency and satisfactory test-retest reproducibility, and correlated as expected with the comparison measures, and its sensitivity to change was assessed from the differences between pre-operative and post-operative scores. The authors concluded that the questionnaire was short, practical, reliable, valid and sensitive to clinically important change.

The questionnaire asks patients about their hip during the past four weeks, and each item offers five ordered responses. The 1996 paper scored each item 1 to 5 with 1 as the best outcome, giving totals of 12 to 60 in which a lower score meant a better hip. Murray and colleagues later proposed a reversed method, now the one the developers recommend, in which each item scores 0 to 4 with 4 as the best outcome. The total is the simple sum.

$$\text{OHS} = \sum_{i=1}^{12} x_i, \qquad S_{\text{new}} = 60-S_{\text{old}}$$

where $x_i$ is the score of item $i$ on the current 0 to 4 scale, OHS is the current total on the 0 to 48 scale, $S_{\text{old}}$ is a total on the original 12 to 60 scale and $S_{\text{new}}$ is the same total expressed on the current scale. A score of 0 means the most severe symptoms and 48 the fewest.

The developers' scoring guide sets rules for incomplete forms. If one or two items are unanswered, the mean of the patient's other answers may fill the gaps, and if more than two are missing no total should be calculated. The guide adds that studies of the questionnaire's measurement properties should not impute at all.

Pain and function inside a single total

The Oxford Hip Score gives one number, but its items cover pain and function. Separating them matters when a treatment might ease pain without restoring function.

Harris and colleagues examined pre-operative scores of 97,487 hip replacement patients in the NHS data set in 2014, and the number of factors depended on the extraction method. When two factors were extracted they formed a pain subscale (items 1, 8, 9, 10, 11 and 12) and a function subscale (items 2 to 7), with Cronbach's alpha of 0.855 and 0.861. The two-factor model fitted better, although no one-factor or two-factor model was rejected, and the authors concluded that the subscales can be reported in addition to the summary score.

The Oxford Hip Score in NHS PROMs

England collects the score for NHS-funded hip replacements before and after surgery. NHS England uses the results to compare providers.

NHS England's 2017 programme guidance describes the Oxford Hip Score as 12 questions on activities of daily living that assess function and residual pain in total hip replacement, collected alongside the three-level EQ-5D that the programme then used. Patients complete a pre-operative questionnaire before surgery and a post-operative questionnaire after a minimum follow-up of six months. NHS England Digital's final 2024/25 publication, released on 2 April 2026 as management information and not as official statistics, counted about 90,000 hip replacement episodes, with about 60,000 pre-operative and 30,000 post-operative responses, and 85.8 per cent of hip patients said they felt much better after surgery. With half as many post-operative as pre-operative responses, linked before and after analyses describe only the patients who returned both forms.

Turning Oxford Hip Score data into QALYs

The total is an ordinal summary of symptoms, not a value of a health state, so the Oxford Hip Score belongs with the non-preference-based measures. A gain of 10 points has no defined meaning on the scale that runs between death at 0 and full health at 1, and the score cannot be multiplied by time to give QALYs. Studies that recorded only the Oxford Hip Score therefore rely on a mapping function estimated in a mapping study that collected both measures.

Pinedo-Villanueva and colleagues estimated such functions in 2013 from Oxford Hip Score and EQ-5D data recorded before and six months after total hip replacement, using data on 1,759 operations. They compared four models, an ordinary least squares regression on the total score, a regression with the twelve items as categorical predictors, a two-part model and a response mapping model. All four estimated the mean EQ-5D score within 0.005 of the observed value. The continuous total-score model was the most accurate and the categorical-item model the most consistent, and adding age, gender and deprivation did not improve them. The authors concluded that when EQ-5D data are missing, Oxford Hip Score answers can estimate a group's mean EQ-5D with high accuracy.

NICE's technology appraisal manual (PMG36, section 4.3.9) allows EQ-5D values to be estimated by mapping other health-related quality-of-life measures when EQ-5D data are not available. Under the interim methods statement PMG51, published on 27 August 2026, such measures should be mapped to utility values based on the UK EQ-5D-5L value set, which may mean mapping onto the 3L value set first. A hip score algorithm estimated against EQ-5D-3L values would therefore need that further step for QALY estimates in a NICE reference case, and because the authors' conclusion concerned group means, mapped values suit cohort models better than patient-level decisions.

Kang reported in 2021, using linked NHS PROMs data on 181,424 primary hip and 191,379 primary knee patients, that both Oxford scores showed great univariate and multivariate responsiveness, while the EQ-5D-3L showed moderate effects by Cohen's thresholds.

Thresholds for a meaningful change in hip score

A difference in Oxford Hip Score has to be read against a threshold suited to the question being asked. Beard and colleagues estimated several thresholds in 2015 from anchor-based and distribution-based analyses.

These figures relate to the minimal clinically important difference concept and depend on the anchor and the population studied, so a trial powered on one threshold should state which it used.

Illustrative worked example of converting and costing a hip score result

All numbers in this example are illustrative. An older cohort study reported a mean pre-operative score of 44 and a mean six-month score of 20 on the original 1996 scale. A new trial compares two hip implants on the current scale, and the new implant costs GBP 300 more per patient.

1. Convert the legacy scores. On the current scale the pre-operative score is 60 − 44 = 16 and the six-month score is 60 − 20 = 40.

2. Compute the cohort change. The improvement is 40 − 16 = 24 points, well above the cohort threshold of about 11 points. On the original scale the same change appears as 20 − 44 = −24, a fall that means improvement.

3. Compare the trial arms. The mean six-month change is 24 points with the new implant and 21 points with the comparator, a difference of 24 − 21 = 3 points. That is below the between-group threshold of about 5 points, so the trial alone does not show an important difference in hip outcome.

4. Map to utilities and compute the cost per QALY. Suppose a mapping function turns the score difference into a mean utility gain of 0.02 for one year.

$$\text{ICER} = \frac{\Delta C}{\Delta U \times T} = \frac{300}{0.02 \times 1} = 15000$$

where $\Delta C$ is the incremental cost per patient in GBP, $\Delta U$ is the mapped mean utility difference and $T$ is the duration of the difference in years, giving GBP 15,000 per QALY.

5. Test the mapping error. If the mapped gain were overstated by 0.005, the error size Pinedo-Villanueva and colleagues reported for group means, the true gain would be 0.02 − 0.005 = 0.015 and the ratio would be 300 / 0.015 = 20,000, or GBP 20,000 per QALY.

An error that is small for a mean utility is large relative to a small utility difference, and here the error moves the ratio by a third. A sensitivity analysis over the mapping function belongs in any appraisal that rests on mapped Oxford Hip Score data.

How the Oxford Hip Score differs from neighbouring measures

The score sits among joint-specific questionnaires and generic instruments, and each comparison has a practical consequence. The Oxford Knee Score was developed by the same group with its own items and validation for knee replacement, so the two totals are not interchangeable even though both run 0 to 48. The WOMAC is a separate osteoarthritis index used for both hips and knees, with different items and scoring, so it cannot be substituted for the Oxford Hip Score without a validated crosswalk. The Charnley hip score, against which Dawson and colleagues validated the questionnaire, is rated by a surgeon, not the patient.

EQ-5D is a generic, preference-based measure scored with a value set, which is why the national programme collects it alongside the Oxford Hip Score. The two answer different questions, one about the value of a health-related quality of life state and one about the hip itself.

Limitations and common misreadings

Most misuses of the Oxford Hip Score in pooled analyses and economic models come from mismatched scales or from reading points as utility. A ceiling at the top of the scale adds a further limit.

• Mixed scoring direction. Studies before and after the 2007 change use opposite directions, and pooling a 12 to 60 score with a 0 to 48 score without conversion reverses the sign of some effects.
• Ceiling. Clement and colleagues reported in 2025, in 7,871 primary total hip replacements, that 21.8 per cent of patients reached the maximum score at one year and 26.6 per cent at two years. Female sex, younger age and better pre-operative scores were each associated with reaching the maximum. The authors judged that this ceiling effect did not stop patients achieving a meaningful change, but a patient at 48 cannot record any further gain, which limits comparisons among patients who do very well.
• Treating points as utility. Adding Oxford Hip Score points to QALYs, or treating a 10-point gain as a fixed share of full health, has no basis without mapping.

Sources

• Beard DJ, Harris K, Dawson J, Doll H, Murray DW, Carr AJ, Price AJ. Meaningful changes for the Oxford hip and knee scores after joint replacement surgery. Journal of Clinical Epidemiology. 2015;68(1):73-79. doi:10.1016/j.jclinepi.2014.08.009
• Clement ND, Jones S, Qaddoura B, Afzal I, Kader DF. The Oxford hip score demonstrates moderate ceiling effects at one and two years after total hip arthroplasty: which patients are at risk and does it matter? European Journal of Orthopaedic Surgery and Traumatology. 2025;35:54. doi:10.1007/s00590-024-04155-7
• Dawson J, Fitzpatrick R, Carr A, Murray D. Questionnaire on the perceptions of patients about total hip replacement. Journal of Bone and Joint Surgery (British volume). 1996;78(2):185-190. doi:10.1302/0301-620X.78B2.0780185
• Harris K, Price A, Beard D, Fitzpatrick R, Jenkinson C, Dawson J. Can pain and function be distinguished in the Oxford Hip Score in a meaningful way? An exploratory and confirmatory factor analysis. Bone and Joint Research. 2014;3(11):305-309. doi:10.1302/2046-3758.311.2000313
• Kang S. Assessing responsiveness of the EQ-5D-3L, the Oxford Hip Score, and the Oxford Knee Score in the NHS patient-reported outcome measures. Journal of Orthopaedic Surgery and Research. 2021;16:18. doi:10.1186/s13018-020-02126-2
• Murray DW, Fitzpatrick R, Rogers K, et al. The use of the Oxford hip and knee scores. Journal of Bone and Joint Surgery (British volume). 2007;89(8):1010-1014. doi:10.1302/0301-620X.89B8.19424
• NHS England. National Patient Reported Outcome Measures (PROMs) Programme Guidance. 2017. https://www.england.nhs.uk/wp-content/uploads/2017/09/proms-programme-guidance.pdf
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). 2022, updated 2026. Section 4.3.9. https://www.nice.org.uk/process/pmg36/chapter/economic-evaluation-2
• National Institute for Health and Care Excellence. Interim methods statement: implementing the EQ-5D-5L value set (PMG51). Published 27 August 2026. Sections 1.10 and 1.11. https://www.nice.org.uk/process/pmg51/chapter/statement-on-using-the-eq-5d-5l-value-set
• NHS England Digital. Patient Reported Outcome Measures (PROMs) in England, Final 2024/25 data. Published 2 April 2026. https://digital.nhs.uk/data-and-information/publications/statistical/patient-reported-outcome-measures-proms/final-2024-25-data
• Oxford University Innovation. The Oxford hip score: a guide to the (new) scoring system. 2018. https://innovation.ox.ac.uk/wp-content/uploads/2014/09/OHS-Scoring-System_2018-01-23.pdf
• Oxford University Innovation. The Oxford Hip Score (OHS). Licensing and outcome measure page, accessed 2026. https://innovation.ox.ac.uk/outcome-measures/oxford-hip-score-ohs/
• Pinedo-Villanueva R, Turner D, Judge A, Raftery JP, Arden NK. Mapping the Oxford hip score onto the EQ-5D utility index. Quality of Life Research. 2013;22(3):665-675. doi:10.1007/s11136-012-0174-y