Knee Injury and Osteoarthritis Outcome Score

Abstract

The Knee Injury and Osteoarthritis Outcome Score (KOOS) is an internationally recognized, patient-reported outcome measure (PROM) designed to quantitatively evaluate short- and long-term knee-related symptoms, functional limitations, and health-related quality of life. Developed by Ewa M. Roos and colleagues in 1998 as an extension of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), the KOOS was specifically engineered to capture clinically meaningful status changes across a broader patient demographic, spanning young, physically active individuals sustaining acute joint injuries—such as anterior cruciate ligament (ACL) tears or meniscus tears—to older adults experiencing post-traumatic or primary osteoarthritis.

The instrument comprises 42 self-administered items distributed across five clinically distinct subscales: Pain (9 items), other Symptoms (7 items), Function in Daily Living (ADL; 17 items), Function in Sport and Recreation (Sport/Rec; 5 items), and Knee-related Quality of Life (QOL; 4 items). Each item is evaluated using a 5-point Likert scale scored from 0 to 4. Each subscale yields an independently calculated, normalized percentage score ranging from 0 (representing extreme knee problems or total disability) to 100 (representing completely normal joint function and zero symptoms), explicitly precluding the calculation of a single aggregated global score to prevent diagnostic obfuscation.

Extensive psychometric investigations have established high internal consistency across all subscales (Cronbach’s alpha typically ranging from 0.74 to 0.93) and robust test-retest reliability (intraclass correlation coefficients [ICC] ranging between 0.75 and 0.95). Structural and convergent validity have been corroborated through systematic comparisons with the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36), the Lysholm Knee Score, and radiographic indicators of joint degeneration (Kellgren-Lawrence grading). The KOOS exhibits superior longitudinal responsiveness in evaluating surgical interventions (such as ACL reconstruction, meniscectomy, and total knee arthroplasty) and conservative physical therapy regimens.

Keywords

Knee Injury and Osteoarthritis Outcome Score, KOOS, patient-reported outcome measure, psychometrics, osteoarthritis, anterior cruciate ligament reconstruction, structural validity, internal consistency, physical functioning, joint-related quality of life.

Authors

The Knee Injury and Osteoarthritis Outcome Score was conceptualized, developed, and validated through an international academic collaboration led by clinical researchers in orthopedics, physical therapy, and epidemiology:

• Ewa M. Roos, PT, PhD — Professor of Muscle and Joint Health, Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark; formerly at the Department of Orthopaedics, Clinical Sciences, Lund University, Sweden.
• Harald P. Roos, MD, PhD — Associate Professor, Department of Orthopaedics, Institute of Clinical Sciences, Lund University Hospital, Lund, Sweden.
• L. Stefan Lohmander, MD, PhD — Senior Professor, Department of Clinical Sciences Lund, Orthopaedics, Lund University Faculty of Medicine, Lund, Sweden.
• Dutch Translation and Adaptation Authors: I. B. de Groot, M. Favejee, and M. Reijman (2008), Department of Orthopaedics, Erasmus University Medical Center, Rotterdam, The Netherlands.

Purpose

The primary clinical and psychometric objective of the KOOS is to provide an empirically robust, multidimensional metric capable of longitudinally tracking knee health from immediate post-traumatic insult to late-stage structural degeneration. Standard orthopedic assessment instruments historically suffered from severe ceiling or floor effects depending on the patient cohort; legacy instruments like the WOMAC were optimized for geriatric populations with advanced, end-stage tri-compartmental osteoarthritis and exhibited profound ceiling effects when administered to athletic cohorts with ligamentous injuries. Conversely, performance metrics designed for sports medicine, such as the International Knee Documentation Committee (IKDC) Subjective Knee Form or the Lysholm score, lacked sensitivity to the slow, insidious functional decline characteristic of early-onset osteoarthritis.

The theoretical rationale underpinning the KOOS directly bridges this longitudinal divide. Acute joint trauma, particularly high-energy disruption of the anterior cruciate ligament or intra-articular fibrocartilaginous meniscus tears, markedly elevates the lifetime relative risk of developing post-traumatic osteoarthritis (PTOA), often manifesting 10 to 20 years post-injury. Consequently, a single, standardized, self-administered questionnaire was required to follow patients across decades—capturing both high-demand physical capacities (e.g., cutting, pivoting, rapid deceleration) during young adulthood and basic activities of daily living (e.g., stair ambulation, sit-to-stand transitions, personal hygiene) in middle and older age.

In clinical trials and comparative effectiveness research, the KOOS functions as a standardized primary or secondary endpoint to benchmark the efficacy of pharmacological interventions, biological therapies (e.g., platelet-rich plasma, autologous chondrocyte implantation), physical therapy regimens, and surgical procedures (including arthroscopic debridement, partial/total knee arthroplasty, and complex ligamentous reconstructions). In routine orthopedic and physical therapy practice, the KOOS provides granular diagnostic clarity, identifying whether a patient’s primary impairment stems from acute mechanical symptoms, localized pain, functional deficits in basic or athletic domains, or psychological distress regarding joint vulnerability.

Psychological Construct

The KOOS operationalizes a complex, multidimensional biopsychosocial construct: knee-specific functional health status and impairment-induced quality of life. Rather than treating joint impairment as an isolated anatomical deficit, the instrument assesses how structural intra-articular pathology interacts with physical self-efficacy, behavioral modifications, pain perception, and psychosocial functioning. The construct is decomposed into five distinct operational dimensions:

1. Pain (9 items)

This subscale evaluates the magnitude and temporal frequency of knee pain during various mechanical joint loadings. It captures baseline background discomfort (pain frequency, nocturnal pain while recumbent, pain during static sitting or standing) as well as dynamic, load-dependent nociceptive responses provoked by knee flexion, terminal extension, axial loading during level walking, stair navigation, and twisting or pivoting motions. Conceptually, this dimension operationalizes nociception and peripheral mechanical sensitization within the joint capsule and surrounding periarticular soft tissues.

2. Symptoms (7 items)

This dimension assesses non-painful mechanical and physiological manifestations of joint pathology. It quantifies localized intra-articular swelling (joint effusion), acoustic or palpable crepitus (grinding and clicking sensations indicative of cartilage fibrillation), mechanical instability or internal derangement (catching and pseudo-locking), anatomical range-of-motion deficits (inability to attain terminal extension or full flexion), and post-inactivity stiffness. The latter incorporates both morning stiffness—a classic hallmark of degenerative or inflammatory joint conditions—and dynamic gel-phenomenon stiffness occurring after prolonged sitting or recumbency.

3. Function in Daily Living (ADL; 17 items)

Covering basic functional mobility, this dimension assesses the degree of physical difficulty experienced when executing non-demanding, routine physical tasks required for independent living. The construct aligns directly with personal autonomy, evaluating lower-extremity closed-chain tasks (ascending and descending stairs, rising from a seated or supine position, standing unsupported), personal self-care (putting on and taking off footwear/socks, entering and exiting a bathtub, utilizing the toilet), and household mobility (getting in and out of an automobile, performing light versus heavy domestic chores). This subscale exhibits high overlap with the physical function subscale of the legacy WOMAC.

4. Function in Sport and Recreation (Sport/Rec; 5 items)

This subscale captures higher-demand physical activities requiring complex neuromuscular control, rapid joint acceleration/deceleration, deep articular flexion, and multidirectional torsional stability. It specifically measures functional performance during deep squatting, running, dynamic jumping, rapid twisting/pivoting over the injured extremity, and sustained kneeling. This dimension eliminates the ceiling effect inherent in basic ADL measures when evaluating younger, athletic, or high-functioning individuals recovering from ligamentous or meniscal trauma.

5. Knee-Related Quality of Life (QOL; 4 items)

This psychosocial dimension measures the psychological and behavioral footprint of chronic knee pathology. It probes cognitive awareness of the joint pathology (how frequently the patient is reminded of their knee condition), lifestyle and behavioral adaptations (conscious modification or cessation of recreational, occupational, or social activities to prevent joint damage), psychological confidence in the mechanical integrity of the joint (fear of the knee giving way or failing), and global subjective difficulty. This subscale directly operationalizes constructs related to kinesiophobia, internalized illness burden, and loss of somatic self-confidence.

Theoretical Framework

The KOOS is firmly rooted in modern health status measurement theory and aligns with the International Classification of Functioning, Disability and Health (ICF) promulgated by the World Health Organization (WHO). Under the ICF framework, human functioning and disability are conceived not as static biological traits, but as dynamic, bidirectional interactions between underlying health conditions (pathological joint anatomy), bodily functions and structures, activity execution, participation in societal roles, and personal/environmental contextual factors.

Historically, orthopedic outcomes relied predominantly on physician-administered objective metrics, such as passive range of motion measured via goniometry, anterior-posterior drawer or Lachman tests for ligamentous laxity, and radiographic staging of joint space narrowing via the Kellgren-Lawrence classification. However, psychometric research repeatedly demonstrated marked discordance between objective structural pathology and subjective patient well-being; patients with severe radiographic osteoarthritis occasionally report minimal functional impairment, while individuals with minor intra-articular lesions may experience catastrophic disability and vocational termination. The KOOS was designed to capture this subjective experiential domain directly, establishing patient experience as the primary metric of therapeutic success.

Furthermore, the KOOS is theoretically grounded in Classical Test Theory (CTT) and multi-attribute utility measurement. Because different pathological states disproportionately compromise distinct dimensions of functioning, the KOOS intentionally avoids aggregating raw item responses into a single composite score. Compressing pain, daily functioning, athletic performance, and emotional quality of life into a single aggregate index conceals critical clinical patterns—for example, an athlete whose severe functional impairment in pivoting maneuvers is completely obscured by flawless scores in basic daily living. By preserving five autonomous subscale vectors, the theoretical architecture of the KOOS preserves clinical specificity across diverse trajectory profiles.

Validity

The psychometric validity of the KOOS has been rigorously scrutinized and corroborated across dozens of distinct linguistic adaptations, clinical populations, and research contexts:

Content and Face Validity

Content validity was established during the instrument’s initial development through iterative focus groups and structured cognitive debriefing involving patients presenting with anterior cruciate ligament ruptures, meniscus pathology, and radiographic osteoarthritis, alongside expert panels of orthopedic surgeons, physical therapists, and epidemiologists. Items were systematically evaluated for linguistic clarity, cognitive interpretability, and ecological relevance. Retaining the complete item set of the validated WOMAC index within the KOOS (specifically embedded across the Pain, Symptoms, and ADL dimensions) ensured immediate legacy content validity while adding vital sport and psychosocial items.

Construct and Convergent Validity

Construct validity has been repeatedly verified via hypothesized correlations with established criterion standards:

• Convergent Validity: The KOOS ADL and Pain subscales exhibit strong, statistically significant correlations ($r = 0.70$ to $0.85$) with the physical functioning and bodily pain domains of the SF-36, as well as with the original WOMAC subscales ($r > 0.90$). The Sport/Rec and QOL subscales correlate moderately to strongly ($r = 0.60$ to $0.78$) with the Lysholm Knee Score and the Tegner Activity Scale in post-ACL reconstruction populations.
• Divergent (Discriminant) Validity: Subscales measuring physical performance and mechanical symptoms show weak correlations ($r < 0.35$) with divergent constructs such as the Mental Health, Vitality, and Emotional Role Functioning subscales of the SF-36, proving that the physical dimensions of the KOOS specifically reflect lower-extremity biomechanical function rather than generalized psychological distress.

Predictive and Longitudinal Construct Validity

Longitudinal construct validity—frequently evaluated in terms of responsiveness to clinical change—has been documented using standardized effect sizes (SES) and Guyatt’s Responsiveness Index. Following arthroscopic ACL reconstruction, the KOOS Sport/Rec and QOL subscales demonstrate exceptionally large effect sizes ($SES > 1.20$ to $1.80$) at 6 and 12 months post-operatively, substantially outperforming legacy measures that lack high-demand functional items. In cohorts undergoing total knee arthroplasty, the KOOS Pain and ADL subscales demonstrate monumental responsiveness ($SES > 2.0$), while the QOL subscale captures substantial, progressive psychological recovery across the rehabilitation timeline.

Reliability

The reliability of the KOOS has been evaluated extensively across acute, subacute, and chronic musculoskeletal conditions, meeting the stringent quality criteria set forth by the COSMIN (COnsensus-based Standards for the selection of health Measurement INstruments) initiative.

Internal Consistency

Internal consistency evaluates the degree of inter-item correlation within each hypothesized subscale dimension. Across multiple international validation trials (including Swedish, English, Dutch, French, German, and Japanese cohorts), Cronbach’s alpha coefficients consistently satisfy classical standards for group-level and individual-level measurement:

• Pain (9 items): $\alpha = 0.85 – 0.93$
• Symptoms (7 items): $\alpha = 0.74 – 0.84$
• Function in Daily Living (ADL; 17 items): $\alpha = 0.92 – 0.97$
• Function in Sport and Recreation (Sport/Rec; 5 items): $\alpha = 0.86 – 0.92$
• Knee-Related Quality of Life (QOL; 4 items): $\alpha = 0.78 – 0.90$

These values demonstrate excellent internal homogeneity without redundant item content (rarely exceeding the $\alpha > 0.95$ threshold that suggests excessive item duplication, except occasionally within the expansive 17-item ADL subscale).

Test-Retest Reliability

Stability across repeated administrations under invariant clinical conditions has been evaluated at test-retest intervals ranging from 48 hours to 2 weeks. Intraclass Correlation Coefficients (ICC), utilizing two-way random-effects models for absolute agreement, demonstrate high temporal stability:

• Pain: $\text{ICC} = 0.80 – 0.93$
• Symptoms: $\text{ICC} = 0.75 – 0.86$
• ADL: $\text{ICC} = 0.84 – 0.95$
• Sport/Rec: $\text{ICC} = 0.81 – 0.92$
• QOL: $\text{ICC} = 0.83 – 0.93$

Measurement Error and Clinical Thresholds

The Standard Error of Measurement (SEM) and Minimal Detectable Change at the 95% confidence level ($MDC_{95}$) vary by subscale and patient population. In general, the $MDC_{95}$ ranges from 6 to 12 points across the five subscales. Methodological consensus identifies a difference or change score of 8 to 10 points as the clinically meaningful threshold, representing the Minimal Clinically Important Difference (MCID) indicative of true, clinically perceptible improvement or deterioration.

Factor Analysis

The underlying dimensionality and latent trait architecture of the KOOS have been comprehensively investigated using both Classical Test Theory—via Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA)—and Modern Test Theory, notably Item Response Theory (IRT) and Rasch measurement models.

Exploratory and Confirmatory Factor Analysis

Initial structural examinations confirmed that the 42 items load meaningfully onto five primary latent factors corresponding to the established subscales. In CFA studies testing oblique five-factor configurations, model fit parameters generally demonstrate acceptable to good structural fit (Root Mean Square Error of Approximation $[RMSEA] < 0.07$, Comparative Fit Index $[CFI] > 0.92$, and Tucker-Lewis Index $[TLI] > 0.90$). Factor loadings for individual items onto their designated target latent variables consistently exceed the standard $0.50$ threshold, with the majority falling between $0.68$ and $0.88$.

However, CFA studies frequently identify very high inter-factor correlations between the Pain and ADL latent dimensions (often exceeding $r = 0.85$). This high collinearity reflects the biomechanical reality that, in degenerative conditions, execution of daily mobility tasks is intrinsically constrained by nociceptive signaling. Despite this overlap, preserving separate dimensions remains clinically essential, particularly when tracking post-operative interventions that rapidly decouple pain perception from long-term neuromuscular functional recovery.

Rasch and Item Response Theory (IRT) Insights

Modern Rasch polytomous rating scale analyses have revealed nuanced structural characteristics:

• Unidimensionality of Subscales: When evaluated individually, each of the five subscales demonstrates acceptable fit to the Rasch model, confirming that each dimension operationalizes a single, coherent latent continuum.
• Category Disordering: Occasional category threshold disordering has been identified within the 5-point Likert response options, particularly in the middle response categories (“Mild” vs. “Moderate”). In certain demographic subgroups, patients struggle to psychometrically differentiate intermediate impairment levels.
• Item Infit/Outfit Statistics: Mean square (MnSq) infit and outfit statistics for the vast majority of items fall within the acceptable psychometric range ($0.6$ to $1.4$). Certain items within the Symptoms subscale (notably acoustic crepitus/clicking and morning stiffness duration) occasionally exhibit borderline outfit values, reflecting the fact that mechanical noises can occur in otherwise asymptomatic knees or resolve independently of other functional improvements.

Instrument / Measurement Tool

• Test Type: Patient-Reported Outcome Measure (PROM); self-administered standardized health questionnaire.
• Administration Format: Self-completed via paper-and-pencil questionnaire, digital web interface, or electronic clinical assessment portal.
• Completion Time: Approximately 10 to 15 minutes.
• Target Population: Adolescents and adults (typically aged 14 and older) presenting with acute joint trauma (ACL tears, collateral ligament strains, meniscal lesions), chondral defects, post-traumatic osteoarthritis, or primary knee osteoarthritis.
• Number of Items: 42 items.
• Subscale Architecture:
  • Pain: 9 items (P1–P9)
  • Symptoms: 7 items (S1–S7)
  • Function in Daily Living (ADL): 17 items (A1–A17)
  • Function in Sport and Recreation (Sport/Rec): 5 items (SP1–SP5)
  • Knee-Related Quality of Life (QOL): 4 items (Q1–Q4)
• Response Format: 5-point Likert scale (scored 0 to 4): None (0), Mild (1), Moderate (2), Severe (3), Extreme (4) [or equivalent anchors: Never (0) to Always (4); None (0) to Totally (4)].
• Scoring Methodology:
  • Each item is assigned an integer value ranging from 0 to 4.
  • Each of the five subscale scores is calculated independently. A single global composite score is never calculated.
  • Subscale transformation formula:
    
    $$\text{Normalized Subscale Score} = 100 – \left( \frac{\text{Mean Raw Score of Subscale Items}}{4} \times 100 \right)$$
    
    Equivalently, using the sum of raw scores:
    
    $$\text{Normalized Subscale Score} = 100 – \left( \frac{\sum \text{Raw Item Scores}}{\text{Item Count} \times 4} \times 100 \right)$$
  • Scale Directionality: Transformed scores range from 0 to 100, where 100 indicates zero impairment (completely healthy knee) and 0 indicates extreme impairment (severe disability).
  • Missing Data Handling: If one or two items are missing within a subscale, the mean of the remaining completed items in that specific subscale can be imputed. If more than two items are omitted, the subscale score is deemed invalid and should be treated as missing.

Permissions & Fee and Test Year

The Knee Injury and Osteoarthritis Outcome Score was originally published in 1998 by Dr. Ewa M. Roos and her research colleagues at Lund University, Sweden. The KOOS was explicitly developed to serve the global medical and scientific community and is maintained as an open-access, public-domain instrument.

No licensing fees or royalties are charged for non-commercial academic research, epidemiological studies, or routine clinical healthcare operations. Users may access downloadable PDF versions, official multi-language translations, scoring spreadsheets, and computerized adaptive testing implementations directly via official clinical repositories and the international KOOS registry (hosted at www.koos.nu). Commercial entities or sponsored pharmaceutical and device clinical trials wishing to incorporate the KOOS into proprietary electronic clinical outcome assessment (eCOA) platforms typically do not pay copyright royalties to the originators, but are expected to respect instrument integrity by using officially validated linguistic translations and adhering precisely to established scoring procedures without modifying item stems, response categories, or recall periods.

References

• Collins, N. J., Misra, D., Felson, D. T., Crossley, K. M., & Roos, E. M. (2011). Measures of knee function: International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form, Knee Injury and Osteoarthritis Outcome Score (KOOS), Knee Injury and Osteoarthritis Outcome Score Physical Function Short Form (KOOS-PS), Knee Outcome Survey Activities of Daily Living Scale (KOS-ADL), Lysholm Knee Scoring Scale, Oxford Knee Score (OKS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), Activity Rating Scale (ARS), and Tegner Activity Scale (TAS). Arthritis Care & Research, 63(S11), S208–S228. https://doi.org/10.1002/acr.20632
• de Groot, I. B., Favejee, M. M., Reijman, M., Verhaar, J. A., & Terwee, C. B. (2008). The Dutch version of the Knee Injury and Osteoarthritis Outcome Score: A validation study. Health and Quality of Life Outcomes, 6(1), Article 16. https://doi.org/10.1186/1477-7525-6-16
• Engelhart, L., Nelson, L., Lewis, S., Mordin, M., McClelland, S., & Roos, E. M. (2012). Validation of the Knee Injury and Osteoarthritis Outcome Score subscales for patients undergoing anterior cruciate ligament reconstruction. The American Journal of Sports Medicine, 40(12), 2764–2772. https://doi.org/10.1177/0363546512461431
• Roos, E. M., & Lohmander, L. S. (2003). The Knee injury and Osteoarthritis Outcome Score (KOOS): From joint injury to osteoarthritis. Health and Quality of Life Outcomes, 1(1), Article 64. https://doi.org/10.1186/1477-7525-1-64
• Roos, E. M., Roos, H. P., Ekdahl, C., & Lohmander, L. S. (1998). Knee injury and Osteoarthritis Outcome Score (KOOS)—Validation of a Swedish version. Scandinavian Journal of Medicine & Science in Sports, 8(6), 439–448. https://doi.org/10.1111/j.1600-0838.1998.tb00465.x
• Roos, E. M., Roos, H. P., Lohmander, L. S., Ekdahl, C., & Beynnon, B. D. (1998). Knee Injury and Osteoarthritis Outcome Score (KOOS)—Development of a self-administered outcome measure. Journal of Orthopaedic & Sports Physical Therapy, 28(2), 88–96. https://doi.org/10.2519/jospt.1998.28.2.88

Items of the Scale