To the Editor,
We read with great interest the recent study by Ziga et al. [1], which demonstrated that smartphone-derived step counts continue to improve up to 6 months after lumbar surgery for sciatica, diverging from the earlier plateau observed in traditional patient-reported outcome measures and objective functional tests. This study underscores the value of real-world activity monitoring in capturing recovery processes that extend beyond clinic-based assessments.
However, several methodological and interpretive aspects merit further discussion to refine the translational utility of digital phenotyping in this field.
First, the magnitude of the step count increase warrants cautious interpretation. In the cohort of Ziga et al. [1], the mean daily step count increased from 4,602 preoperatively to 9,858 at 6 months. While this may reflect genuine functional restoration, it also raises the possibility of behavioral modulation.
Compared with published data from other orthopedic populations, such as total joint arthroplasty patients (~5,200 steps at 6 months) [2] or hospitalized lumbar surgery patients (<8,000 steps in the early postoperative period) [3]. the observed trajectory appears relatively steep. Although this is speculative, factors such as a rebound increase in activity after pain relief or heightened self-monitoring awareness (similar to the Hawthorne effect) could contribute. Importantly, because the step count data in the original study were extracted retrospectively, real-time behavioral reactivity may be less likely.
Nevertheless, future prospective designs with control groups would help disentangle physiological recovery from behavioral amplification.
Second, reliance on raw step counts alone may underutilize the analytical potential of smartphone sensors. Modern devices can capture high-resolution accelerometer and gyroscope data, enabling metrics such as gait symmetry, cadence variability, and postural transition dynamics [4]. Sciatica and lateral recess stenosis (LRS) often alter gait mechanics through pain-avoidant strategies and subtle motor deficits.
Incorporating these digital biomarkers in future studies could provide a more nuanced view of neural and biomechanical recovery beyond what cumulative step counts alone can reveal. However, it should be noted that the extraction of advanced gait metrics from consumer-grade smartphones is not without limitations. These include variations in proprietary algorithms across device manufacturers (e.g., iOS vs.
Android), differences in sensor hardware specifications, and the ongoing need for external validation against gold-standard motion capture systems in clinical populations.
Third, the aggregation of heterogeneous surgical pathologies may obscure distinct recovery kinetics. Although Ziga et al. appropriately excluded fusion cases, their cohort included both patients who underwent discectomy for lumbar disc herniation (LDH) and those who underwent decompression for LRS. Emerging evidence suggests that even among decompression procedures, underlying pathologies may exhibit different mobility trajectories [5].
LDH patients typically experience a rapid resolution of radicular pain and corresponding functional improvement shortly after discectomy, whereas LRS patients often present with chronic, progressive compression leading to more complex neuromuscular adaptations; consequently, the latter group may demonstrate a slower, more protracted recovery phenotype. Future studies with larger samples could explore whether digital recovery phenotypes differ between LDH and LRS, potentially informing tailored rehabilitation strategies.
In summary, Ziga et al. [1] provided important evidence that real-world activity monitoring extends our understanding of postoperative recovery. By considering behavioral confounders, integrating advanced digital biomarkers, and exploring pathology-specific trajectories, subsequent research can further enhance the clinical relevance of smartphone-based outcome assessments.