Risk prediction model for knee pain in the Nottingham Community: a Bayesian modeling approach

Fernandes, Gwen Sascha and Bhattacharya, Archan and McWilliams, Daniel F. and Ingham, Sarah Louise and Doherty, Michael and Zhang, Weiya (2017) Risk prediction model for knee pain in the Nottingham Community: a Bayesian modeling approach. Arthritis Research and Therapy, 19 . p. 59. ISSN 1478-6354 (Unpublished)

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (806kB) | Preview

Abstract

Background: 25% of the British population over the age of 50 experience knee pain. It can limit physical ability, cause distress and bears significant socioeconomic costs. Knee pain, not knee osteoarthritis (KOA) is the all to common malady. The objectives of this study were to develop and validate the first risk prediction model for incident knee pain in the Nottingham community and validate this internally within the Nottingham cohort and externally within the Osteoarthritis Initiaitve (OAI) Cohort.

Methods: 1822 participants at risk for knee pain from the Nottingham community were followed up for 12 years. Of this cohort, 2/3 (n=1203) were used to develop the risk prediction model and 1/3 (n=619) were used to validate the model. Incident knee pain was defined as pain on most days for at least one month in the past 12 months. Predictors were age, gender, body mass index (BMI), pain elsewhere, prior knee injury and knee alignment. Bayesian logistic regression model was used to determine the probability of an odds ratio >1. The Hosmer-Lemeshow x2 statistic (HLS) was used for calibration and receiver operator characteristics (ROC) was used for discrimination. The OAI cohort was used to examine the performance of the model in a secondary care population.

Results: A risk prediction model for knee pain incidence was developed using a Bayesian approach. The model had good calibration with HLS of 7.17 (p=0.52) and moderate discriminative abilities (ROC 0.70) in the community. Individual scenarios are given using the model. However, the model had poor calibration (HLS 5866.28, p<0.01) and poor discriminative ability (ROC 0.54) in the OAI secondary care dataset.

Conclusion: This is the first risk prediction model for knee pain, irrespective of underlying structural changes of KOA, in the community using a Bayesian modelling approach. The model appears to work well in a community-based population but not in a hospital derived cohort and may provide a convenient tool for primary care to predict the risk of knee pain in the general population.

Item Type: Article
Keywords: Knee pain ; Bayesian statistics ; Prediction modelling ; Musculoskeletal epidemiology
Schools/Departments: University of Nottingham, UK > Faculty of Medicine and Health Sciences > School of Medicine > Division of Rheumatology, Orthopaedics and Dermatology
Identification Number: 10.1186/s13075-017-1272-6
Depositing User: Eprints, Support
Date Deposited: 23 Mar 2017 13:18
Last Modified: 24 Mar 2017 15:10
URI: http://eprints.nottingham.ac.uk/id/eprint/41186

Actions (Archive Staff Only)

Edit View Edit View