We thank the authors of the letter for their insightful comments and interest in our study, “Physical Performance Continues to Improve After Surgery for Sciatica, Exceeding Recovery Periods of Physical Capacity and Patient-Reported Outcomes: Multicenter Prospective Observational Study.” [1] The points raised offer an excellent opportunity to clarify the methodological underpinnings and clinical interpretations of our findings.
1. Magnitude of Step Count Trajectories and Cohort Characteristics
The authors note that the increase in mean daily steps from 4,602 preoperatively to 9,858 at 6 months is relatively steep compared with that reported in the orthopedic and spinal literature. We agree that recovery in this context is a complex interplay of clinical improvement and behavioral adaptation.
Crucially, our study followed recovery as a sequential process to demonstrate that patients first feel better (perceived pain relief and stabilized patient-reported outcome measures [PROMs] at 6 weeks) but perform better only at a later stage as they reintegrate physical movement into daily routines. This unique temporal dissociation was illustrated specifically to suggest the ceiling effect of standardized subjective outcome measures rather than to directly quantify pathology-specific functional performance.
The authors contrast our findings with those of the cohort studied by Kim et al. [2] in 2019. However, a closer look at this study reveals a considerably different clinical scenario:
• Acute Inpatient Assessment: Kim et al.2 prospectively evaluated 22 patients strictly during their immediate, short-term hospitalization period (postoperative days 1 through 7).
• Pathology vs. Timeline: In contrast, our multicenter study captured long-term, free-living ambulatory data across a 6-month timeline postdischarge in a younger, highly active, and strictly nonfused microsurgical sciatica cohort.
Therefore, the cohorts and observation phases are fundamentally distinct. While the post hoc extraction of passively collected data minimizes active behavioral reactivity (the Hawthorne effect), we completely agree that definitive assessment of physiological recovery versus behavioral amplification must be addressed in future larger cohorts utilizing dedicated control groups.
2. Utilization of High-Resolution Biomarkers Versus Raw Step Counts
The suggestion to integrate complex digital biomarkers—such as gait symmetry, cadence variability, and postural transitions—is a highly valid point. These metrics will undoubtedly be pivotal in building a deeper, multidimensional understanding of walking patterns across various spinal pathologies.
However, extracting and clinically validating high-resolution biomechanical data requires substantial technical infrastructure, standardized data pipelines, and extensive institutional resources. The primary objective of our study was pragmatic: we aimed to demonstrate that clinicians can utilize basic smartphone step counts in a very simple, low-burden way to effectively measure objective functional impairment and long-term recovery. Demonstrating that a readily accessible metric such as daily step count can successfully expose the ceiling effects of PROMs serves as an important and scalable first step before introducing more resource-intensive parameters.
3. Aggregation of Surgical Pathologies and Recovery Kinetics
The authors point out that aggregating lumbar disc herniation (LDH) and lateral recess stenosis (LRS) might obscure distinct recovery patterns. We agree that different spinal pathologies can entail varied walking kinematics, a phenomenon previously addressed in the literature (e.g., Chauhan et al.3). Intuitively, one can predict that patients requiring instrumentation or fusion due to underlying structural instability will exhibit highly disparate recovery slopes.
To maximize cohort homogeneity and isolate sciatica as the primary driving impairment, our study deliberately implemented a strict surgical protocol:
• We exclusively included patients without mechanical instability or the need for spinal fusion.
• All patients presented with sciatica as their primary symptom and were operated on using the same microsurgical decompression technique.
By strictly excluding fusion cases and patients with structural instability, we minimized biomechanical confounding across our 50 patients. Nevertheless, we fully endorse the authors’ recommendation that subsequent, larger-scale trials should explicitly model and compare the long-term digital phenotypes of pure LDH and isolated LRS to guide tailored rehabilitation.
We believe that real-world passive digital phenotyping provides valuable insight into longitudinal recovery. Addressing behavioral confounders, refining simple consumer metrics, and mapping precise pathology-specific paths are the logical next steps for this field.