International Knee Documentation Committee Subjective Knee Form

Mixed Conditions: (Irrgang et al., 2001; n = 533; mean age = 37.5 (16.2) years; range 6.2–86.6 years; 92.3% white; 76.1% participated in sports; mixed knee diagnoses including ligament injury (n = 150), meniscal injury (n = 108), patellofemoral pain (n = 93), and osteoarthritis (n = 92))

• SEM based on the test-retest reliability coefficient: 4.6 points
• SEM based on coefficient alpha (internal consistency): 5.4 points

Mixed Conditions: (Irrgang et al., 2001)

• True score change (based on test-retest SEM) ± 9.0 points; changes greater than 9 points represent true change, while changes less than 9 points represent measurement error 
• MDC95 (calculated using test-retest SEM of 4.6 as 1.96 x SEM x √2): ±12.8 points

Mixed Conditions: (Irrgang et al., 2006; n = 207; mean age = 40.5 (16.7) years; 47.1% male; mean follow-up = 19.0 (2.9) months; mixed knee diagnoses including ligament injury (n = 67), meniscal injury (n = 69), and osteoarthritis (n = 76))

• Change score of 11.5 points had the highest sensitivity (0.82) for distinguishing patients who perceived improvement from those who did not (specificity = 0.64)
• Change score of 20.5 points had the highest specificity (0.84) for distinguishing patients who perceived improvement from those who did not (sensitivity = 0.64)
• Receiver operating characteristic curve area under the curve = 0.78 (95% Cl, 0.72-0.85)

Mixed Conditions (general U.S. population, non-institutionalized adults): (Anderson et al., 2006; n = 2,670 respondents [5,246 knees]; mean age = 39 (14) years; range = 18–65 years; stratified random sample matched to U.S. Census data)

• Overall sample mean score: 82 (22); 26% of respondents scored the maximum of 100
• Men: 18–24 years: 89.1 (17.5); 25–34 years: 88.9 (16.3); 35–50 years: 84.9 (19.3); 51–65 years: 77.4 (23.3)
• Women: 18–24 years: 85.7 (19.1); 25–34 years: 86.0 (18.5); 35–50 years: 79.9 (22.6); 51–65 years: 70.9 (26.0)
• Men scored significantly higher than women overall (84.4 vs 80.1, p < 0.0001); score correlated negatively with age (r = -0.25, p < 0.0001), with age-related differences emerging after age 35
• Scores were significantly lower for respondents reporting a current knee problem, current treatment, or history of knee surgery on the involved knee compared to those without (e.g. right knee problem: mean = 56.9 (20.7)) vs no problem: mean = 90.8 (12.8))

Mixed Conditions: (Irrgang et al., 2001; test-retest subsample n = 33; mean age = 37.2 (10.5) years; 39.4% female; mean interval between administrations = 49.7 (24.2) days, range 4–92 days)

• Excellent test-retest reliability: (ICC[2,1] = 0.94; 95% CI: 0.88–0.97)

Mixed Conditions: (Irrgang et al., 2001; n = 533)

• Excellent internal consistency for the total scale: (Cronbach’s α = 0.92*)

Mixed Conditions: (Higgins et al., 2007; n = 1,517; 58.2% male; 73.0% Caucasian; mean age = 37.5 (14.9) years; age range = 6–100 years; patients at a sports medicine clinic in the southeastern US with a variety of knee disorders including osteoarthritis, ligamentous tears, general knee pain, and pre-/post-surgery status)

• Excellent internal consistency for both subscales using a two-factor structure: symptom and knee articulation subscale (11 items) Cronbach’s α = 0.87; activity level subscale (4 items) Cronbach’s α = 0.88 Note: these subscale values reflect the two-factor model of Higgins et al. and are distinct from the single-scale α = 0.92* reported by Irrgang et al. (2001)

*Scores higher than 0.9 may indicate redundancy in the scale questions.

Concurrent Validity:

Mixed Conditions: (Irrgang et al., 2001)

• Poor to excellent correlations with SF-36 subscales (r = 0.16–0.66), with the strongest associations observed for physical-function domains.

Mixed Conditions: (Higgins et al., 2007)

• Both two-factor subscales correlated more strongly with SF-12 physical than mental component scores.

Convergent Validity:

Mixed Conditions: (Irrgang et al., 2001)

• Item-response theory analyses indicated the instrument functioned similarly across all age (young vs old; r = 0.95), sex (male vs female; r = 0.97), and diagnosis subgroups: ligament injury vs no ligament injury (r = 0.94), patellofemoral problems vs none (r = 0.95), meniscal injury vs none (r = 0.99), and osteoarthritis vs none (r = 0.95); only 3 of 18 items (pain frequency, ascending stairs, descending stairs) functioned somewhat differently for patients with vs without ligament injury 
• Factor analysis revealed a single dominant component (eigenvalue = 9.03) accounting for 50.2% of total variance, supporting combination of items into a single total score 

Discriminant Validity:

Mixed Conditions: (Higgins et al., 2007)

• Exploratory and confirmatory factor analysis supported a two-factor structure (symptom and knee articulation, 11 items; activity level, 4 items), retaining 15 of the 18 items; 3 items (locking/catching, running straight ahead, jumping and landing) were factorially complex and recommended for exclusion. This differs from the single-factor structure reported by Irrgang et al. (2001) (see Considerations)
• Graded-response item response theory indicated that most items discriminated well across levels of the knee-function latent trait, supporting item-level validity

Mixed Conditions: (Irrgang et al., 2001; n = 533)

• Items were developed by an international panel of knee experts (the International Knee Documentation Committee) based on review of the original IKDC form, the MODEMS Lower Limb Instrument, and the Knee Outcome Survey Activities of Daily Living and Sports Activity Scales, and were refined through 2 rounds of pilot testing (n = 144 and n = 222) prior to selection of the final 18-item version from an initial pool of 41 items

Mixed Conditions: (Irrgang et al., 2001; n = 533)

• Excellent floor effect (0% of patients scored the minimum possible score)
• Adequate ceiling effect (0.2% [1 of 533] scored the maximum possible score of 100

Mixed Conditions: (Irrgang et al., 2006; n = 207; mean follow-up = 19.0 (2.9) months, range = 6–28 months)

• Large Change and Responsiveness: effect size = 1.13; standardized response mean = 0.94
• Mean score improved from 44.8 (19.4) at baseline to 66.8 (22.9) at follow-up (mean change = 22.0 (23.4) points, p < .0001)
• Mean change increased systematically with patients’ global rating of change, from −15.1 points (greatly worse) to 38.6 points (greatly better)
• Subgroup – Ligament injury (ACL, n = 50): mean change = 27.7 (27.7) points (baseline 49.5 to follow-up 77.2)
• Subgroup – Osteoarthritis (n = 76): mean change = 14.2 (16.8) points (baseline 37.3 to follow-up 51.5)

Pediatric and Congenital Conditions: (calculated from Kocher et al., 2011; n = 589; mean age = 14.6 (2.5) years; age range = 10.0-18.9 years; 48.9% male; variety of knee disorders including ligament injury, meniscal injury, patellofemoral injury, and osteochondritis dissecans)

• SEM (calculated using overall mean score SD of 22 and test-retest ICC of 0.91): 6.6

Pediatric and Congenital Conditions: (calculated from Kocher et al., 2011)

• MDC95 (calculated as 1.96 x SEM x √2, using calculated SEM of 6.6): 18.3)

Pediatric and Congenital Conditions (general population): (Nasreddine et al., 2016; n = 2000 (11% in each age from 10-18 years); female = 50%, 68% white, 86% non-Hispanic; 7% (n = 136) reported prior surgery on one or both knees, with 4% of the surgeries on the index knee)

• Age-specific 5th-percentile values (smoothed) can serve as a lower limit of normal: a score below the age-specific 5th percentile may be considered outside the normal range (e.g., a 13-year-old scoring below 52.5)
• See Normative Data below for the full smoothed-percentile table by age

Pediatric and Congenital Conditions (general population): (Nasreddine et al., 2016)

Mean (SD) Scores on Pedi-IKDC and Smoothed Percentiles by Age^a 

^aPedi-IKDC: Pediatric International Knee Documentation Committee Subjective Knee Evaluation Form

Pediatric and Congenital Conditions: (Kocher et al., 2011)

• Excellent test-retest reliability: (ICC = 0.91)

Pediatric and Congenital Conditions: (Kocher et al., 2011)

• Excellent internal consistency (Cronbach’s α = 0.91*)

*Scores higher than 0.9 may indicate redundancy in the scale questions.

Concurrent Validity:

Pediatric and Congenital Conditions: (Kocher et al., 2011)

• Excellent correlations between scores on the Pedi-IKDC and the Child Health Questionnaire (CHQ) domains of Physical Functioning (r = 0.65) and Bodily Pain (r = 0.61)
• Adequate correlations between scores on the Pedi-IKDC and CHQ domains: Emotional Limitations (r = 0.35), Physical Limitations (r = 0.45), Self-Esteem (r = 0.32), and Family Activities (r = 0.37)
• Poor correlations between scores on the Pedi-IKDC and CHQ domains: Mental Health (r = 0.30) and General Health Perceptions (r = 0.20)

Pediatric and Congenital Conditions (clinical sample): (Kocher et al., 2011)

• All 11 hypothesized relationships were supported, demonstrating construct validity

Discriminant Validity:

Pediatric and Congenital Conditions (general population): (Nasreddine et al., 2016; n = 2,000)

• Significant ability of Pedi-IKDC scores to discriminate between known groups: children with a history of prior surgery in the index knee (n = 79; mean = 69.9 (20.4); median = 70.7) and children with recent (4-week) activity limitation in the index knee (n = 234; mean = 71.0 (18.5); median = 70.7) scored approximately 25 points lower than those without these histories (both p < 0.0001)
• By contrast, variation across race, ethnicity, geographic region, and sports-participation level was minor (all median ranges < 4.5 points)

Pediatric and Congenital Conditions: (Kocher et al., 2011)

• The Pedi-IKDC was developed as a modification of the adult IKDC Subjective Knee Form, adapting wording and content for readability and developmental relevance in patients aged 10–18 years

Pediatric and Congenital Conditions (clinical sample): (Kocher et al., 2011; n = 589)

• Excellent floor effect for the overall score (0%)
• Adequate ceiling effect for the overall score (6%)

Pediatric and Congenital Conditions (general/healthy population): (Nasreddine et al., 2016; n = 2,000)

• Poor overall ceiling effect: 34% of scores reached the maximum value of 100
• Adequate ceiling effects within clinically relevant subgroups (6% among those with prior index-knee surgery; 5% among those with recent index-knee activity limitation), consistent with the overall ceiling effect being driven by healthy respondents

Pediatric and Congenital Conditions: (Kocher et al., 2011; n = 98; mean age = 14.9 (2.3) years; 48% male; assessed before and a mean of 7.6 months after a variety of knee surgical procedures including ACL reconstruction, meniscal procedures, and patellar procedures)

• Large Change and Responsiveness: effect size = 1.39; standardized response mean = 1.35
• Significant improvement in mean score from 52.7 to 81.8 (mean change = 29.1 (21.5) points, p < 0.0001); similar improvement across male, female, and age (10–14 vs 15–18 years) subgroups

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007; n = 31; mean age = 50.6 years, age range = 19 to 73 years; patients with meniscus pathology of the knee requiring treatment; knees with ligament pathology or chondral defect greater than Outerbridge grade 2 were excluded; test-retest subgroup, retested within 4 weeks)

• SEM = 3.19

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007)

• Minimum detectable change = 8.8 points

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007)

• Excellent test-retest reliability for the overall IKDC score (ICC = 0.95)

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007)

• Adequate: Cronbach’s alpha = 0.773

Concurrent Validity:

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007)

• Excellent correlation between the total IKDC score and the physical component of the SF-12 scale (r = 0.60)

Discriminant Validity:

Musculoskeletal Conditions (Meniscal Injury): (Crawford et al., 2007)

• Significant differences (p < 0.05) in IKDC scores observed among known-groups for all constructs tested: lower IKDC scores were associated with lower patient activity levels, greater difficulty with activities of daily living, more difficulty with sports, abnormal knee function, and complex/degenerative meniscus tears

Musculoskeletal Conditions (Meniscal Injury): (Shepard et al., 2023, citing Crawford et al., 2007)

• Excellent floor effect (0%) and excellent ceiling effect (0%) reported for overall IKDC score

Musculoskeletal Conditions (Meniscal Injury): (Shepard et al., 2023, systematic review and meta-analysis of 257 studies overall [28,612 patients]; responsiveness analysis subset of 66 studies reporting IKDC scores, n = 3261 menisci; pooled mean age = 38.6 years; pooled mean BMI = 26.3; minimum 1-year follow-up required for inclusion in responsiveness analysis)

• Large Change: pooled effect size = 1.94, the largest of 10 patient-reported outcome measures compared in this meta-analysis (Lysholm = 1.56; KOOS Overall = 1.64; KOOS Quality of Life = 1.62; KOOS Pain = 1.32; visual analog scale = -1.38; KOOS Sport/Recreation = 1.14; KOOS Activities of Daily Living = 1.03; KOOS Symptoms = 0.99; Tegner Activity Scale = 0.57)
• Pooled score improved from 46.7 (15.2) preoperatively to 76.2 (18.3) postoperatively
• IKDC demonstrated greater relative efficiency (comparative responsiveness) than the Lysholm Score (relative efficiency = 1.03), Tegner Activity Scale (3.90), and KOOS Activities of Daily Living (1.12); KOOS Quality of Life demonstrated somewhat greater relative efficiency than IKDC (1.45)
• IKDC was the second most commonly used PROM in the meniscal surgery literature overall, appearing in 132 of 257 studies (51.4%), behind only the Lysholm Score (75.1%)