Premium
Risk Stratification for 4,837 Individuals with Knee Pain Who Receive Physical Therapy Treatment
Author(s) -
Salamh Paul A.,
Reiman Michael,
Cleland Joshua,
Mintken Paul,
Rodeghero Jason,
Cook Chad E.
Publication year - 2017
Publication title -
musculoskeletal care
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.628
H-Index - 28
eISSN - 1557-0681
pISSN - 1478-2189
DOI - 10.1002/msc.1150
Subject(s) - medicine , psychological intervention , physical therapy , risk stratification , univariate , multivariate analysis , cohort , retrospective cohort study , knee pain , univariate analysis , multivariate statistics , multinomial logistic regression , alternative medicine , osteoarthritis , statistics , mathematics , pathology , machine learning , psychiatry , computer science
Risk stratification is a modelling method that is designed to target interventions toward patients with specific needs. The objective of the present study was to identify predictive characteristics related to patients with knee impairments who had a high risk of a bad prognosis (exceptional non‐responders) as well as those who were at low risk of a bad prognosis (exceptional responders). A cohort of 4,837 patients with knee pain seen for physical therapy was retrospective analysed using univariate and multivariate multinomial regression analyses. Modelling was used to identify characteristics associated with those who were exceptional responders and those who were exceptional non‐responders. Exceptional non‐responders were significantly associated with older age, female gender, longer duration of symptoms, surgical history, lower functional status at baseline and a payer type. Exceptional responders were significantly associated with younger age, no previous surgical history, higher functional status at baseline and a payer type. Findings may be used for managing processes involving intensity of care service and in understanding probable prognoses for each patient. Future research should continue to examine variables predictive of treatment response in patients with knee pain. Copyright © 2016 John Wiley & Sons, Ltd.