Abstract
(1) Background: Predicting chronic low back pain (LBP) is of clinical and economic interest as LBP leads to disabilities and health service utilization. This study aims to build a competitive and interpretable prediction model; (2) Methods: We used clinical and claims data of 3837 participants of a population-based cohort study to predict future LBP consultations (ICD-10: M40.XX-M54.XX). Best subset selection (BSS) was applied in repeated random samples of training data (75% of data); scoring rules were used to identify the best subset of predictors. The rediction accuracy of BSS was compared to randomforest and support vector machines (SVM) in the validation data (25% of data); (3) Results: The best subset comprised 16 out of 32 predictors. Previous occurrence of LBP increased the odds for future LBP consultations (odds ratio (OR) 6.91 [5.05; 9.45]), while concomitant diseases reduced the odds (1 vs. 0, OR: 0.74 [0.57; 0.98], >1 vs. 0: 0.37 [0.21; 0.67]). The area-under-curve (AUC) of BSS was acceptable (0.78 [0.74; 0.82]) and comparable with SVM (0.78 [0.74; 0.82]) and randomforest (0.79 [0.75; 0.83]); (4) Conclusions: Regarding prediction accuracy, BSS has been considered competitive with established machine-learning approaches. Nonetheless, considerable misclassification is inherent and further refinements are required to improve predictions.
Highlights
Non-specific low back pain (LBP) is one of the leading health problems and associated with substantial individual burdens, health service utilization and indirect costs worldwide [1,2,3]
The results of our study suggest that many participants with LBP were able to cope with LBP without seeking medical care, and did not become patients
In subgroup analyses of our study, we found that a considerable proportion of participants (n = 740, 19.3%) with severe LBP had previously consulted a physician for LBP, but these participants did not seek physicians’ care again during follow-up
Summary
Non-specific low back pain (LBP) is one of the leading health problems and associated with substantial individual burdens, health service utilization and indirect costs worldwide [1,2,3]. LBP has been portrayed as having a good prognosis with a high proportion of patients recovering spontaneously within few weeks, with no or minimal interventions This is true for most, a substantial proportion of patients will develop chronic or recurrent back pain [4,5,6]. For health care providers and health insurances, the identification of such patients is, of great importance, since this is a prerequisite for the rational allocation of limited specialized services and for offering targeted interventions to those who are most affected This would, for example, be of relevance to optimize disease management programs for chronic low back pain, which are mandated by the statutory health insurance in Germany
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call