Physical activity and patient-reported outcomes after decompressive endoscopic lumbar spine surgery

Although patients often report improvements in pain and functional capacity after endoscopic lumbar spine surgery, objectively measured real-life physical activity may differ from that reported. This multicenter prospective study aimed to assess the correlations between physical activity and patient-reported outcomes (PROs).

All adult patients undergoing endoscopic lumbar spine surgery were offered enrollment in the SPINEhealthie smartphone app to assess physical activity. Estimated daily step counts (SCs) and serial PROs, including visual analog scale (VAS) leg pain, VAS back pain, and Oswestry Disability Index (ODI) scores, were collected.

Of the 289 patients with pre- and postoperative SC data (mean follow-up 11.4 months), > 70% of patients achieved minimal clinically important difference (MCID) for VAS back and leg pain at 2 weeks and remained stable at 1 year (p < 0.05). Although patients demonstrated substantial improvement in VAS leg (-3.4, p < 0.001) and VAS back (-3.1, p < 0.001) pain scores at 2 weeks, the mean SC decreased significantly at 2 weeks (-632.4, p < 0.001). SC began to increase at 3 months (+265.7, p = 0.004). VAS scores demonstrated minimal further improvement beyond 2 weeks, but SCs continued to improve at all subsequent time points (p < 0.001), as did ODI (-12.2 at 3 months and -13.6 at 1 year, p < 0.001 for both). Preoperative SC was highly predictive of 1-year SC (r = 0.86, p < 0.001) and mildly predictive of ODI at 1 year (r = -0.31, p = 0.006). No preoperative PRO was correlated with PROs or SC at 1 year (r ≤ 0.10, p > 0.2 for all). Quartile analysis of SC revealed that pain scores did not correlate with functional measures. The patient quartiles' recovery patterns remained distinct (p < 0.05 between quartiles at 1 year). ODI improvement was greater in patients with higher baseline SCs (-15.6 vs -13.8 in the lowest quartile at 1 year, p = 0.047). Stratification of patients by baseline scores did not distinguish improvement in SC, or ODI, VAS leg pain, or VAS back pain (p > 0.05 for all) scores, indicative that baseline scores did not predict outcomes.

Patients demonstrated immediate reduction in pain but delayed improvement in SC and ODI score. Among preoperative assessments, only SCs were predictive of postoperative outcomes. Furthermore, pain scores did not correlate with disability or activity. SC demonstrated interquartile stability, suggesting that SC may provide a reliable and independent perspective and may be predictive of outcomes.