Development of Prediction Model for the Prognosis of Sick Leave Due to Low Back Pain
Lisa C Bosman1, Jos W R Twisk, Anna S Geraedts
1Amsterdam UMC, Department of Epidemiology and Biostatistics, Vrije Universiteit Amsterdam, Amsterdam Public Health Research Institute, Amsterdam (Ms Bosman, Dr Twisk, Dr Heymans); ArboNed Occupational Health Service, Utrecht (Ms Bosman, Dr Geraedts), The Netherlands.
Objective:
The aim of this study was to develop a prediction model for the prognosis of sick leave due to low back pain (LBP).
Methods:
This is a cohort study with 103 employees sick-listed due to non-specific LBP and spinal disc herniation. A prediction model was developed based on questionnaire data and registered sick leave data with follow up of 180 days.
Results:
At follow up 31 (30.1%) employees were still sick-listed due to LBP. Forward selection procedure resulted in a model with: catastrophizing, musculoskeletal work load, and disability. The explained variance was 27.3%, calibration was adequate and discrimination was fair with area under the ROC-curve (AUC) = 0.761 (interquartile range [IQR]: 0.755-0.770).
Conclusion:
The prediction model of this study can adequately predict LBP sick leave after 180 days and could be used for employees sick listed due LBP.
