Decary S.1, Frémont P.2, Pelletier B.3, Fallaha M.4, Martel-Pelletier J.3, Pelletier J.-P.3, Feldman D.1, Sylvestre M.-P.5, Vendittoli P.-A.4, Desmeules F.1
1Université de Montréal, School of Rehabilitation, Montreal, Canada, 2University Laval, School of Rehabilitation, Quebec, Canada, 3Université de Montréal, Medicine, Montréal, Canada, 4Université de Montréal, Surgery, Montréal, Canada, 5Université de Montréal, Public Health, Montréal, Canada
Background: Although patellofemoral pain (PFP) is a commonly encountered disorder in physiotherapy, the evidence is currently limited concerning the optimal combination of patients' history elements and physical examination tests to establish a valid diagnosis of PFP.
Purpose: The objective of this study is to establish the diagnostic validity of combining selected history elements and physical examination tests, part of a standardized musculoskeletal examination (ME), to confirm or exclude a diagnosis of PFP.
Methods: This is a prospective diagnostic study where a physiotherapist and physicians (three orthopaedic surgeons and two sports medicine physicians) independently assessed and diagnosed consecutive participants consulting for a new knee complaint. ME by the physiotherapist, including history elements and physical examination tests was compared to the reference standard: a physicians composite diagnosis including both ME and imaging tests. Penalized logistic regression (LASSO) was used to identify predictors yielding maximal area under the curve (AUC) and recursive partitioning was used to identify valid diagnostic clusters from these predictors. Sensitivity (Se) and specificity (Sp) as well as positive and negative likelihood ratios (LR+/-) with associated 95% confidence intervals were calculated to assess the validity of the diagnostic clusters to include or to exclude PFP.
Results: Two hundred seventy-nine participants were evaluated for a total of 359 primary and secondary diagnoses including: osteoarthritis (n=129), meniscal tears (n=80), anterior cruciate ligament (ACL) tears (n=43) and other diagnoses (n=32). Seventy-five participants had a diagnosis of PFP (26.9%). Compared to other participants, those with PFP were significantly younger (PFP: 38.3±13.5 years old, others: 53.1±14.8 years old, p 0.01) and had a lower body mass index (PFP: 26.8±5.9, others: 30.4±6.5 Kg/m2, p 0.01), but the proportion of females was not significantly different (PFP: 61.3%, others: 56.4%, p=0.46). Twenty-three variables were associated with the diagnosis of PFP and yielded a maximal AUC of 0.87±0.01. Five of these variables formed predicting diagnostic clusters including: age, presence of anterior knee pain, difficulty descending stairs, patellar facets tenderness and patellar maltracking (J-sign) during knee extension. Individuals aged 40 years old with anterior knee pain or medial patellar facet tenderness or individuals aged between 40 and 58 years old with patellar tenderness of any facets, positive J-sign and light to moderate difficulty descending stairs had a high likelihood of having PFP (LR+: 10.25; 95%CI: 5.91-17.79). Individuals aged >58 years old or aged 58 years old with other locations than anterior knee pain and negative J-sign or without patellar facets tenderness had a low likelihood of having PFP (LR-: 0.09; 95%CI: 0.04-0.22).
Conclusion(s): Age, presence of anterior knee pain, difficulty descending stairs, patellar facets tenderness and patellar maltracking (J-sign) during knee extension used in combination lead to two diagnostic clusters to accurately diagnose PFP and one cluster to accurately exclude PFP.
Implications: These clusters may be used by physiotherapists to confirm or exclude a diagnosis of PFP and initiate efficient conservative management. Further research will externally validate these clusters in more individuals consulting in primary care.
Funding acknowledgements: This project has received funding from the Canadian Institute of Health Research and Fonds de Recherche en Santé du Québec.
Topic: Musculoskeletal: peripheral
Ethics approval: The study was approved by Maisonneuve Rosemont Hospitals ethics comity and all participants signed a consent form prior to consultation.
All authors, affiliations and abstracts have been published as submitted.
