Clinical course and prognostic models for the conservative management of cervical radiculopathy: a prospective cohort study

Abstract

Purpose To describe the clinical course and develop prognostic models for poor recovery in patients with cervical radiculopathy who are managed conservatively. Methods Sixty-one consecutive adults with cervical radiculopathy who were referred for conservative management were included in a prospective cohort study, with 6-and 12-month follow-up assessments. Exclusion criteria were the presence of known serious pathology or spinal surgery in the past. Outcome measures were perceived recovery, neck pain intensity and disability level. Multiple imputation analyses were performed for missing values. Prognostic models were developed using multivariable logistic regression analyses, with bootstrapping techniques for internal validation. Results About 55% of participants reported to be recovered at 6 and 12 months. All multivariable models contained 2 baseline predictors. Longer symptoms duration increased the risk of poor perceived recovery, whereas the presence of paresthesia decreased this risk. A higher neck pain intensity and a longer duration of symptoms increased the risk of poor relief of neck pain. A higher disability score increased the risk of poor relief of disability, and larger active range of rotation toward the affected side decreased this risk. Following bootstrapping, the explained variance of the models varied between 0.22 and 0.30, and the median area under the curve varied between 0.75 and 0.79. Conclusions The clinical course of cervical radiculopathy appears to be long, with most of the reduction in symptoms occurring within the first 6 months. All prognostic models showed an adequate predictive performance with modest diagnostic accuracy and explained variance. Graphical abstract These slides can be retrieved under Electronic Supplementary Material.

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

“…Paradoxically, better balance interacted with better cervical ROM, to predict worse neck pain at 12 th months using MuARS. This was in contrast to prior research which reported greater recovery in individuals with CR with better cervical ROM 11,37 . Based on the predictor of "h(54-AROM_RR) * h(Romberg-12)", a 1 s increase in Romberg timing and a 1° increase in right cervical rotation, increased neck pain by 0.127 points, only in individuals with poorer balance (< 12 s) but with better right cervical rotation (> 54°).…”

“…A limitation of the present study is the small sample size relative to the number of predictors included, which precluded splitting our data into training and validation sets, the latter for independent validation 6 . In defence, the present study represents one of the largest prospective clinical investigations of individuals with CR, compared to previous research 10,11 . A previous simulation study reported that machine learning methods are "data hungry" -in that they may need 10 times as many events per predictor to achieve stable prediction within the classification framework 43 .…”

“…Standard post-surgical rehabilitation (weeks [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]. Participants in this group were treated in accordance with the Swedish standard post-surgical care of individuals with CR.…”

“…The present findings were supported by the literature which reported up to one-third of post-operative individuals with CR present with deficits in neck muscle strength and endurance at a 12 th month follow up, compared to healthy controls 35 . Previous studies have only reported the following baseline variables to be predictors of 12 month disability: pain intensity and psychological distress 12 ; disability, axial cervical ROM, pain intensity, sex, hand grip strength 13 ; axial cervical ROM and disability 11 . Considering the outcome of 24 months disability, the following predictors were selected in previous studies: worker's compensation case and neurological sensory function 36 ; pain intensity and cervical sagittal ROM 37 ; disability, pain intensity, cervical sagittal and axial ROM, sex and hand-grip strength 13 ; sex, and number of operated levels on the cervical spine 26 .…”

Probing the mechanisms underpinning recovery in post‐surgical patients with cervical radiculopathy using Bayesian networks

“…Yet, no studies have investigated the role of psychological features as mediators of pain‐related disability in individuals with CR. In a recent prognostic study on conservatively managed CR, factors that predicted poor recovery of disability after 1 year were the cervical ROM towards the affected side, and the baseline neck disability score (Sleijser‐Koehorst et al, ). Earlier prognostic studies also reported that cervical ROM, hand grip strength and measures of depression with psychosomatism were important predictors of long‐term disability, albeit in post‐surgical patients with CR (Peolsson & Peolsson, ; Peolsson, Vavruch, & Öberg, ).…”