Clinical predictive modelling of post-surgical recovery in individuals with cervical radiculopathy: a machine learning approach

Abstract

Prognostic models play an important role in the clinical management of cervical radiculopathy (CR). No study has compared the performance of modern machine learning techniques, against more traditional stepwise regression techniques, when developing prognostic models in individuals with CR. We analysed a prospective cohort dataset of 201 individuals with CR. Four modelling techniques (stepwise regression, least absolute shrinkage and selection operator [LASSO], boosting, and multivariate adaptive regression splines [MuARS]) were each used to form a prognostic model for each of four outcomes obtained at a 12 month follow-up (disability—neck disability index [NDI]), quality of life (EQ5D), present neck pain intensity, and present arm pain intensity). For all four outcomes, the differences in mean performance between all four models were small (difference of NDI < 1 point; EQ5D < 0.1 point; neck and arm pain < 2 points). Given that the predictive accuracy of all four modelling methods were clinically similar, the optimal modelling method may be selected based on the parsimony of predictors. Some of the most parsimonious models were achieved using MuARS, a non-linear technique. Modern machine learning methods may be used to probe relationships along different regions of the predictor space.

Assessment of Neuromuscular and Psychological Function in People with Recurrent Neck Pain during a Period of Remission: Cross-Sectional and Longitudinal Analyses

“…The RMSE score to predict NDI is similar to a model generated in people with cervical radiculopathy [ 82 ], with an RMSE of about 8.2% (NDI 0–100% scale). However, this comparison should be interpreted with caution due to the different populations.…”

“…R statistical software was used to conduct this analysis. The functions, packages, and codes that were used to analyse this data have been described elsewhere [ 82 ].…”

Designing Social Care Plans by Grouping Services and Patients in Mixed Cohorts: A Study using Regression versus Neural Nets

“…In [20] the ARMA models apply to likely linearly related year series. In [21] a step-wise regression is suggested as better over traditional LR or ARMA methods and this is also proven in this work empirically using specific ad-hoc ranges for LR orders and ARMA lags (delays). In [22] it is advocated that CC needs to be coupled with ML to reveal specific relationships across data.…”

Machine learning versus logistic regression for prognostic modelling in individuals with non-specific neck pain

“…The complexity between the predictors and outcome in previous studies [29,30] may be too low for ML to have a meaningful benefit over logistic regression. A previous study in cervical radiculopathy found nonlinear relationships between baseline selfreported predictors and 12 months clinical outcomes of neck and arm pain and disability [31]. The nonlinear relationship between baseline and outcomes may not be surprising given that previous studies reported different nonlinear rates of recovery in disability with different baseline neck disability scores in individuals with whiplash-associated disorders (WAD) [32].…”