Oxford Hip Score

Abstract

The Oxford Hip Score (OHS) is a widely recognized, joint-specific patient-reported outcome measure (PROM) designed to evaluate pain intensity and physical functional disability in patients undergoing total hip replacement (arthroplasty) or presenting with advanced hip osteoarthritis. Developed in 1996 by Dawson, Fitzpatrick, Carr, and Murray at the University of Oxford, the questionnaire was engineered to address the critical need for a brief, self-administered, reliable, and valid instrument that directly reflects the patient’s subjective experience without surgeon-observer bias. The OHS comprises 12 categorical items assessing the status of the hip over the preceding four weeks: six items focus on joint pain and symptoms (including nocturnal pain, pain duration during ambulation, and sudden sharp pain episodes), while six items evaluate specific functional limitations in activities of daily living (such as climbing stairs, entering/exiting vehicles, putting on socks or stockings, and personal hygiene).

Originally scored from 1 to 5 per item (yielding a total score range of 12 to 60, where 60 indicated the most severe symptoms and disability), the scoring system was modernized in 2007 to an intuitive scale of 0 to 4 per item, producing a cumulative score from 0 (indicating severe hip impairment) to 48 (representing optimal hip function). Across international clinical trials and nationwide joint registries, the OHS demonstrates exceptional psychometric characteristics, including high internal consistency (Cronbach’s alpha typically ranging between .84 and .93), robust test-retest reliability (intraclass correlation coefficients > .90), strong responsiveness to surgical intervention (standardized response mean > 1.50), and clear construct validity against generic health metrics such as the SF-36 and EQ-5D.

Keywords

Oxford Hip Score, Total Hip Arthroplasty, Patient-Reported Outcome Measures, Hip Osteoarthritis, Psychometrics, Joint-Specific Functional Disability, Pain Assessment, Health-Related Quality of Life, Construct Validity, Clinical Orthopaedics

Authors

The Oxford Hip Score was conceptualized, designed, and psychometrically validated by an interdisciplinary team of researchers and clinical orthopaedic surgeons based at the University of Oxford, United Kingdom:

• Jill Dawson, MSc, DPhil: Associate Professor and Health Services Researcher, Department of Public Health and Primary Care (now the Nuffield Department of Population Health), University of Oxford.
• Ray Fitzpatrick, PhD: Professor of Public Health and Primary Care, Department of Public Health, University of Oxford; Emeritus Fellow of Nuffield College.
• Andrew Carr, ChM, FRCS, FMedSci: Nuffield Professor of Orthopaedic Surgery, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences (NDORMS), University of Oxford.
• David W. Murray, MD, FRCS: Professor of Orthopaedic Surgery, NDORMS, University of Oxford; Consultant Orthopaedic Surgeon, Nuffield Orthopaedic Centre, Oxford.

Primary International Adaptations: The official Dutch validation was developed and published by Dr. Ton Gosens and colleagues in 2005 (Department of Orthopaedic Surgery, St. Elisabeth Hospital, Tilburg, Netherlands). Intellectual property rights and translation distributions are governed by Oxford University Innovation (Clinical Outcomes team, Oxford, United Kingdom).

Purpose

Before the introduction of the Oxford Hip Score in 1996, clinical evaluations of surgical hip interventions predominantly relied upon clinician-administered composite scores, most notably the Harris Hip Score (HHS) and the Charnley Hip Scale. While these traditional instruments provided objective physiological data such as passive range of motion, limb deformity, and radiographic alignment, empirical investigations identified substantial limitations: high inter-observer variability, marked discrepancies between clinical physical assessments and actual patient satisfaction, and a pervasive risk of surgeon bias when assessing their own surgical outcomes. Clinicians routinely underestimated the subjective burden of persistent chronic pain while overestimating the functional utility of marginal gains in passive joint range of motion.

The primary clinical and research purpose of the Oxford Hip Score was to create a concise, disease- and joint-specific outcome instrument completed entirely by the patient. The design prioritizes minimal respondent burden (requiring less than five minutes to complete) while maximizing sensitivity to clinically meaningful changes induced by total hip replacement (THR) or hip resurfacing. The rationale acknowledges that from the patient’s perspective, successful joint reconstruction is defined by the attenuation of debilitating pain and the restoration of practical autonomy in everyday life, rather than technical radiographic indicators.

In contemporary orthopaedic practice, the OHS serves several critical functions:

• National Joint Registries: The instrument is a core metric utilized by statutory surgical registries, such as the National Joint Registry (NJR) of England and Wales and the New Zealand Joint Registry, to benchmark surgeon performance, track long-term prosthetic implant survivorship, and monitor population-level health trends.
• Clinical Trial Endpoints: It serves as a primary endpoint in randomized controlled trials assessing novel prosthetic implant designs, surgical techniques (e.g., direct anterior approach vs. posterior approach), robotic-assisted surgery, and accelerated rehabilitation protocols.
• Individualized Clinical Decision Support: Preoperative OHS scores assist clinical teams in identifying candidates who will derive the greatest therapeutic benefit from surgical intervention, while postoperative trajectory tracking aids in the early detection of surgical failure, infection, or aseptic loosening.
• Health Economics and Quality Audits: It provides empirical inputs for health-related quality of life calculations, contributing to cost-effectiveness models and quality-adjusted life year (QALY) analyses.

Psychological Construct

The Oxford Hip Score measures the latent construct of joint-specific physical functioning and pain severity attributable to pathology of the hip joint. Rather than evaluating psychological distress or systemic health in isolation, the instrument captures the behavioral, sensory, and functional manifestations of musculoskeletal impairment as perceived by the individual in their natural environment. The construct is conceptualized along two primary, interacting domains:

1. Sensory and Affective Hip Pain

Pain associated with advanced hip osteoarthritis or structural joint degeneration is multidimensional, encompassing continuous baseline aching, mechanical loading pain, intermittent sharp distress, and sleep interruption. Within the OHS, pain is evaluated across five specific items reflecting distinct sensory and temporal characteristics:

• Baseline Pain Severity (Item 1): Assesses the typical or baseline intensity of pain experienced over the previous four weeks, capturing the continuous nociceptive burden.
• Walking-Induced Pain Threshold (Item 6): Evaluates the temporal and mechanical threshold at which ambulation precipitates severe pain, linking joint loading capacity to sensory thresholds.
• Paroxysmal/Neuropathic-Type Pain (Item 10): Assesses the frequency of sudden, severe episodes—described in the questionnaire as shooting, stabbing, or spasmic pain—which carry high affective distress and often evoke fear-avoidance behaviors.
• Pain Interference with Work and Domestic Activity (Item 11): Captures the functional disability directly mediated by pain severity, bridging sensory experience and societal role performance.
• Nocturnal Pain (Item 12): Assesses sleep disruption caused by joint pain, a major marker of severe inflammation or mechanical intra-osseous pressure and an established driver of depressive symptoms and emotional distress in chronic arthritic populations.

2. Functional Limitations in Activities of Daily Living

Physical functional capacity represents the degree to which joint stiffness, biomechanical instability, and pain prevent the individual from executing fundamental daily tasks. This domain is evaluated across seven functional items:

• Lower-Extremity Flexibility and Dressing (Item 4): Putting on socks, stockings, or tights requires pronounced hip flexion, internal rotation, and adduction. In severe joint capsule contracture, this movement is invariably compromised.
• Independent Hygiene (Item 2): Washing and drying oneself all over assesses complex multi-planar joint mobility and balance.
• Community Mobility and Transport (Item 3 & Item 5): Getting in and out of a vehicle or utilizing public transit (Item 3) and performing household shopping autonomously (Item 5) reflect advanced functional independence, outdoor balance, and community participation.
• Stair Negotiation and Locomotor Demands (Item 7 & Item 8): Climbing a flight of stairs (Item 7) requires dynamic quadriceps and gluteal contraction combined with significant axial loading through the hip joint. Similarly, rising from a seated position after a meal (Item 8) evaluates joint stiffness (“start-up pain”) and dynamic extensor strength.
• Gait Impairment and Limping (Item 9): Assesses antalgic gait patterns or Trendelenburg limping caused by gluteus medius insufficiency or joint pain.

Theoretical Framework

The Oxford Hip Score operates at the intersection of several complementary theoretical models in psychometrics, rehabilitation sciences, and health psychology:

The Biopsychosocial Model and the WHO ICF Framework

The OHS is theoretically grounded in George Engel’s Biopsychosocial Model and aligns directly with the World Health Organization’s International Classification of Functioning, Disability and Health (ICF). Within the ICF framework, human functioning is conceptualized across three interconnected levels:

1. Body Functions and Structures (Impairments): Pathological joint changes, cartilage erosion, osteophyte formation, and localized nociception.
2. Activities (Activity Limitations): Difficulties an individual experiences in executing tasks (e.g., walking, climbing stairs, dressing).
3. Participation (Participation Restrictions): Problems an individual experiences in involvement in life situations (e.g., occupational duties, domestic shopping, social transport).

Traditional orthopaedic scales focused heavily on Body Structures (range of motion, joint deformity). The OHS shifts the measurement locus upward to Activity Limitations and immediate Participation Restrictions, acknowledging that physiological preservation without functional utility does not equate to health restoration.

Wilson and Cleary Model of Health-Related Quality of Life

The Wilson and Cleary conceptual model links biological and physiological variables to symptom status, functional status, general health perceptions, and overall quality of life. The Oxford Hip Score concentrates specifically on the link between symptom status (pain intensity, spasmodic episodes) and functional status (locomotor and self-care capacity). By measuring both domains within a concise, coherent framework, the OHS captures how primary joint symptoms directly generate daily functional disability.

Psychometric Measurement Theory: CTT and Modern Rasch Modeling

From a psychometric measurement perspective, the development of the OHS relied initially on Classical Test Theory (CTT), prioritizing high internal consistency, low respondent burden, and high construct validity. Subsequent psychometric research has evaluated the scale through the lens of modern Item Response Theory (IRT) and Rasch measurement models. These analyses demonstrate that the 12 items form an ordered continuum of difficulty, tracking patients from profound incapacity prior to joint replacement through the transition to asymptomatic restoration.

Validity

The construct, criterion, and cross-cultural validity of the Oxford Hip Score have been demonstrated across hundreds of independent orthopaedic trials and validation cohorts globally.

Construct Validity (Convergent and Discriminant)

In the original validation study by Dawson et al. (1996), involving 224 patients undergoing total hip replacement, construct validity was evaluated against the generic Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). The OHS demonstrated strong convergent validity, displaying high correlation with the SF-36 Bodily Pain dimension (Pearson’s r = 0.76) and Physical Functioning subscale (r = 0.67). Conversely, divergent (discriminant) validity was verified through markedly lower correlations with domains capturing non-musculoskeletal concepts, such as the SF-36 Mental Health dimension (r = 0.28) and Role-Emotional scale (r = 0.26).

Subsequent psychometric investigations confirmed strong convergence between the OHS and other legacy or disease-specific joint instruments:

• WOMAC Hip Index: Correlation coefficients between the OHS composite score and the WOMAC Pain and Physical Function subscales consistently exceed r = 0.80 to 0.88, demonstrating that the OHS measures equivalent musculoskeletal constructs while imposing a lower respondent burden (12 items vs. 24 items).
• Harris Hip Score (HHS): The OHS correlates strongly with the HHS (r = 0.68 to 0.78), while avoiding the floor effects frequently observed in the HHS among patients with severe physical debility.
• EuroQol-5D (EQ-5D): Pre- and postoperative OHS scores correlate solidly with the EQ-5D health utility index (r = 0.60 to 0.72), confirming its relevance to general health economic evaluations.

Predictive Validity and Responsiveness to Change

Responsiveness—the capacity to detect genuine clinical change following an effective intervention—is a central psychometric strength of the OHS. Dawson et al. (1996) reported a Standardized Response Mean (SRM) of 1.76 at six months post-surgery, substantially outperforming generic health surveys (e.g., SF-36 Physical Functioning SRM = 0.99; SF-36 Bodily Pain SRM = 1.34). Subsequent large-scale registry analyses have observed SRMs exceeding 2.0 at one year following primary total hip arthroplasty.

The Minimal Clinically Important Difference (MCID) or Minimally Important Change (MIC) on the 0–48 scale has been established via anchor-based and distribution-based methods at approximately 5 to 9 points (equivalent to a change of roughly 6 to 11 points on the original 12–60 system). Improvements exceeding this threshold correspond to meaningful patient-perceived recovery.

Cross-Cultural Validity and Global Adaptations

The instrument has undergone formal linguistic and psychometric adaptation worldwide, including Dutch (Gosens et al., 2005), German, Swedish, French, Japanese, and Turkish cohorts. In the Dutch validation study by Gosens et al. (2005), 114 patients undergoing primary total hip arthroplasty were assessed pre- and postoperatively. The Dutch version displayed structural equivalence to the original English questionnaire, high correlation with the Dutch-validated Harris Hip Score (r = 0.74), and an SRM of 1.84, confirming cross-cultural measurement invariance.

Reliability

Extensive psychometric investigations have established that the Oxford Hip Score has high reliability across divergent surgical indications, ages, and demographic profiles.

Internal Consistency

Internal consistency evaluates the degree to which all individual items within the instrument measure the same overarching construct. In the initial validation cohort (Dawson et al., 1996), Cronbach’s alpha was calculated at 0.84 preoperatively and 0.89 at six months postoperatively. In the Dutch cross-cultural adaptation by Gosens et al. (2005), Cronbach’s alpha reached 0.89 preoperatively and 0.92 postoperatively. Contemporary registry validations across tens of thousands of surgical recipients report coefficients consistently hovering between 0.89 and 0.93, indicating strong scale homogeneity without excessive item redundancy.

Test-Retest Reliability

The reproducibility of the instrument when administered to clinically stable patients over short test-retest intervals (ranging from 24 hours to two weeks) has been confirmed in multiple investigations:

• Dawson et al. (1996) established an initial test-retest reliability coefficient of r = 0.92 among stable prospective candidates assessed over a two-week interval.
• Gosens et al. (2005) reported an Intraclass Correlation Coefficient (ICC) of 0.93 (95% CI: 0.88–0.96) for the complete translated scale.
• Kemp et al. (2013) demonstrated an ICC exceeding 0.95 in younger populations suffering from non-arthritic hip disease (e.g., femoroacetabular impingement, labral tears).

Standard Error of Measurement and Smallest Detectable Change

The Standard Error of Measurement (SEM) of the modernized OHS (0–48 scale) has been quantified across the literature at approximately 1.5 to 2.2 points. Consequently, the Smallest Detectable Change (SDC), reflecting the minimum point deviation necessary to guarantee with 95% statistical confidence that the observed difference is not attributable to random measurement error, ranges between 4.1 and 5.5 points at the individual patient level.

Factor Analysis

The structural dimensionality of the Oxford Hip Score has been the subject of extensive psychometric investigation, focusing on whether the scale reflects a strictly unidimensional construct or a correlated two-factor model distinguishing joint pain from physical functional performance.

Exploratory Factor Analysis (EFA)

In original exploratory analyses, principal component analyses with varimax rotation routinely identify two distinct eigenvalues exceeding Kaiser’s criterion (> 1.0), accounting for approximately 60% to 68% of the total cumulative variance:

• Factor 1: Pain and Symptom Severity: Items loading heavily on this factor include Item 1 (overall typical pain; loading > 0.75), Item 6 (walking duration before severe pain; loading > 0.70), Item 10 (sudden shooting pain; loading > 0.65), Item 11 (interference with daily work; loading > 0.78), and Item 12 (pain at night; loading > 0.68).
• Factor 2: Functional Limitations in Locomotion and Self-Care: Items loading on this factor include Item 2 (washing and drying; loading > 0.72), Item 3 (car and public transport; loading > 0.68), Item 4 (socks/stockings; loading > 0.74), Item 5 (household shopping; loading > 0.70), Item 7 (stairs; loading > 0.73), Item 8 (standing from chair; loading > 0.62), and Item 9 (limping; cross-loading between 0.45 and 0.58).

Despite these distinct factor loadings, the inter-factor correlation between Pain and Function is high (Pearson’s r typically between 0.72 and 0.82), indicating substantial shared variance.

Confirmatory Factor Analysis (CFA) and Fit Indices

Confirmatory factor analytic investigations have compared single-factor, correlated two-factor, and higher-order hierarchical or bifactor models. While a correlated two-factor model provides slightly superior statistical fit indices compared to a strict unidimensional model (e.g., Comparative Fit Index [CFI] = 0.96 vs. 0.92; Root Mean Square Error of Approximation [RMSEA] = 0.062 vs. 0.088), the high correlation between factors provides strong psychometric justification for combining both sub-dimensions into a single composite score.

Bifactor structural analyses indicate that the general factor accounts for more than 75% to 80% of the common variance, yielding an Omega Hierarchical coefficient (ωh) > 0.80. This demonstrates that the total composite score accurately reflects general joint dysfunction, even if subtle multidimensionality exists at the subscale level.

Rasch Analysis and Item Response Theory (IRT)

Modern psychometric assessments using Rasch measurement models have examined item difficulty ordering, category threshold functioning, and Differential Item Functioning (DIF):

• Item Invariance and Difficulty Hierarchy: Rasch analyses reveal an ordered continuum of physical limitation. Item 4 (putting on socks) and Item 7 (climbing stairs) represent high-difficulty items (requiring high functional capacity to achieve maximum points), whereas Item 8 (rising from a chair) and Item 2 (washing/drying) represent low-difficulty items.
• Differential Item Functioning (DIF): Minor uniform DIF has been identified across biological sex for Item 4 (putting on socks/tights) and Item 5 (shopping), as well as across age cohorts for high-demand locomotor items. However, the magnitude of this DIF is clinically negligible and does not alter the clinical interpretation of the composite score.
• Threshold Disordering: The original 1996 scoring system (1–5) exhibited slight threshold disordering in intermediate response categories. The revised 2007 scoring system (0–4) corrected these issues, demonstrating ordered category thresholds across the scale items.

Instrument / Measurement Tool

• Instrument Name: Oxford Hip Score (OHS)
• Alternative Titles: Oxford heup score (Dutch); Oxford-Hip-Score (German)
• Measurement Paradigm: Patient-Reported Outcome Measure (PROM); health-related quality of life; joint-specific physical disability and pain
• Target Population: Adult and elderly individuals presenting with hip osteoarthritis, degenerative hip disease, femoral head osteonecrosis, femoroacetabular impingement, or those undergoing elective total hip arthroplasty (THR) / revision hip surgery
• Administration Format: Self-administered paper-and-pencil questionnaire, electronic clinical interface (e-PROM), or telephone interview
• Completion Time: 2 to 5 minutes
• Item Count: 12 categorical items
• Recall Period: The preceding four weeks (“during the past 4 weeks”)
• Response Scale (Mandatory Psychometric Structure):
  • 5-point categorical response scale (originally scored 1 to 5, total score 12 to 60 with 60 being worst; updated in 2007 to 0 to 4 per item, total score 0 to 48 with 48 being best outcome).
• Scoring Transformations and Historical Systems:
  • Original 1996 Scoring System: Each item was scored from 1 to 5, where 1 represented the least symptoms/disability and 5 represented the most severe symptoms/disability, producing a total score between 12 (best possible) and 60 (worst possible).
  • Modernized 2007 Scoring System (Murray et al., 2007): Each item is scored from 0 to 4, where 0 represents the worst possible outcome/severe disability and 4 represents the best possible outcome/no symptoms. Item scores are summed to generate an overall score ranging from 0 (most severe symptoms, worst outcome) to 48 (no symptoms, best outcome).
• Clinical Stratification and Interpretation Guidelines (0–48 Modern Scale):
  • Score 0 to 19 (Severe Hip Arthritis): Indicates severe physical impairment and unremitting pain. May indicate the need for surgical evaluation or urgent orthopaedic intervention.
  • Score 20 to 29 (Moderate to Severe Hip Impairment): Indicates substantial joint pain and significant functional limitation in daily activities. Warranting active orthopaedic management, physical therapy, pharmacological intervention, or surgical assessment.
  • Score 30 to 39 (Mild to Moderate Hip Impairment): Indicates moderate symptoms manageable with non-surgical conservative therapies, lifestyle modifications, or baseline postoperative stabilization.
  • Score 40 to 48 (Satisfactory Joint Function): Represents optimal hip health, minor or absent functional restrictions, and minimal to no pain. Typical expected outcome following successful, uncomplicated total hip arthroplasty.
• Missing Data Imputation Protocol: If one or two items are omitted, the mean value of the completed items may be imputed to substitute the missing response. If more than two items (three or more) are unanswered, the questionnaire is psychometrically invalid and cannot be scored.

Permissions & Fee and Test Year

The Oxford Hip Score was originally developed and published in 1996 by Dawson, Fitzpatrick, Carr, and Murray. The modernized, standardized scoring system (0 to 48) was established in 2007 by Murray et al. The Dutch cross-cultural adaptation was published in 2005 by Ton Gosens and associates.

Copyright and Licensing Regulations: Intellectual property rights for the Oxford Hip Score (including all official validated translations and electronic adaptations) are owned by the University of Oxford and managed by Oxford University Innovation (OUI), Clinical Outcomes team (formerly Isis Innovation Ltd), Oxford, United Kingdom.

• Academic and Non-Funded Clinical Use: The scale is typically accessible free of charge for non-commercial academic research, routine public healthcare delivery, individual clinical practice, and university audits, provided that permission is formally requested and authorized via an Oxford University Innovation user agreement.
• Commercial and Industry-Funded Trials: Pharmaceutical companies, medical device manufacturers, clinical research organizations (CROs), and commercial entities conducting funded clinical studies are required to pay a licensing fee per study or per administration.
• Modification Notice: The wording, item structure, order, and response options of the OHS must not be modified, amended, or adapted without explicit, written authorization from Oxford University Innovation to preserve standard psychometric validity and cross-study comparability.

References

• Beard, D. J., Harris, K., Dawson, J., Doll, H., Murray, D. W., Carr, A. J., & Fitzpatrick, R. (2015). Meaningful changes for the Oxford Hip and Knee Scores after joint replacement surgery. Journal of Clinical Epidemiology, 68(1), 73–79. https://doi.org/10.1016/j.jclinepi.2014.08.009
• Dawson, J., Fitzpatrick, R., Carr, A., & Murray, D. (1996). Questionnaire on the perceptions of patients about total hip replacement. The Journal of Bone and Joint Surgery. British volume, 78-B(2), 185–190. https://doi.org/10.1302/0301-620X.78B2.0780185
• Gosens, T., Hoefnagels, N. H. M., de Vet, H. C. W., Dhert, W. J. A., van Langelaan, E. J., Bulstra, S. K., & Geesink, R. G. T. (2005). The “Oxford Heup Score”: De Nederlandse versie van de Oxford Hip Score. Nederlands Tijdschrift voor Orthopedie, 12(3), 89–96.
• Kemp, J. L., Collins, N. J., Roos, E. M., & Crossley, K. M. (2013). Psychometric properties of patient-reported outcome measures for hip and groin injury: A systematic review. British Journal of Sports Medicine, 47(16), 1040–1048. https://doi.org/10.1136/bjsports-2013-092784
• Murray, D. W., Fitzpatrick, R., Rogers, K., Pandit, H., Beard, D. J., Carr, A. J., & Dawson, J. (2007). The use of the Oxford Hip and Knee Scores. The Journal of Bone and Joint Surgery. British volume, 89-B(8), 1010–1014. https://doi.org/10.1302/0301-620X.89B8.19424
• Ostendorf, M., van Stel, H. F., Buskens, E., Schrijvers, A. J. P., Marting, L. N., Verbout, A. J., & Dhert, W. J. A. (2004). Patient-reported outcome in total hip replacement: A comparison of five instruments of health status. The Journal of Bone and Joint Surgery. British volume, 86-B(6), 801–808. https://doi.org/10.1302/0301-620X.86B6.14660

13. Items of the Scale (Questionnaire)