Predicting recovery in patients with acute low back pain: A Clinical Prediction Model

Area of Science:

• Orthopedics
• Pain Management
• Clinical Prediction Modeling

Background:

• Acute low back pain (LBP) prognosis varies significantly among individuals.
• Identifying patients likely to recover at specific time points is crucial for effective treatment decisions.
• Predicting recovery at the 1-week mark after initial care offers valuable clinical insights.

Purpose of the Study:

• To develop and validate a clinical prediction model for recovery from acute LBP.
• To estimate the probability of recovery at 1-week, 1-month, and 3-months post-initial care for persistent LBP patients.
• To identify key prognostic factors influencing recovery timelines in acute LBP.

Main Methods:

• A cohort of 1070 patients with acute LBP and pain scores ≥2 at 1-week follow-up was studied.
• Multivariable Cox regression analysis was employed to identify prognostic factors.
• Recovery was defined as days to pain resolution.

Main Results:

• The final model incorporated: duration of episode, prior episodes, depressive symptoms, 1-week pain intensity, and pain change in the first week.
• Predicted recovery probabilities at 1-week, 1-month, and 3-months ranged from 4-59%, 19-91%, and 30-97%, respectively.
• The model demonstrated good predictive performance with a C-statistic of 0.758 and strong calibration.

Conclusions:

• A simple clinical prediction model using five easily obtainable variables can forecast recovery likelihood in persistent acute LBP.
• This model provides valuable prognostic information for guiding treatment strategies in acute LBP management.