Trajectories of acute low back pain: a latent class growth analysis

Objectives: The objective of this study was to investigate the different recovery trajectories after an episode of acute low back pain, and to explore individual characteristics that are associated with each trajectory. Methods: Participants were adults consulting general practice with acute low back pain, recruited for a placebo controlled trial of paracetamol. Pain numerical rating scale data (score range 0‐10) was collected at baseline, weeks 1, 2, 4 and 12. Latent Class Growth Analysis (LCGA) was used to identify pain trajectories, based on trichotomized pain scores (0‐1 = pain recovery, 2‐4 = moderate pain, ≥ 5 = high pain). Multinomial logistic regression analysis was used to explore patient characteristics at baseline that were associated with individual cluster membership. Results: In total, 1,477 participants who provided pain scores at all five collection points were analyzed. Six distinct clusters were identified including: ‘rapid pain recovery’ (n=518), ‘rapid pain decrease, recovery at 12 weeks’ (n=403), ‘gradual pain decrease, recovery at 12 weeks’ (n=180), ‘incomplete recovery’ (n=180), ‘persistent high pain’ (n=103) and ‘fluctuating pain’ (n=93). Higher pain, increased pain duration, and perceived risk of persistence were associated with delayed recovery trajectories when compared to rapid recovery. In addition, a higher number of previous episodes, workers' compensation, and decreased quality of life were associated with non-recovery trajectories when compared to rapid recovery. Conclusion: Overall, 25% of primary care attendees with acute low back pain show early tendency for pain persistence, recurrence or only partial recovery. Identification of patient characteristics that can influence recovery offers potential targets to improve patient outcomes.