Validity, reliability, and responsiveness of a low-cost activity monitor for assessing walking in patients with lumbar spinal stenosis.

BACKGROUND CONTEXT: Reduced walking capacity is the hallmark of lumbar spinal stenosis (LSS). Currently, walking capacity in LSS is assessed with self-reported questionnaires. The Self-Paced Walking Test (SPWT) has been recommended as an objective capture of walking capacity, but takes up to 30 minutes to complete, preventing its widespread use by clinicians. Consumer-based activity monitors and smartphones have been suggested as an alternative but have not been validated in the LSS population.

PURPOSE: To determine the measurement properties of criterion validity, reliability and responsiveness of a low-cost consumer-based wrist activity monitor, smartphone walking app, SPWT and 2 patient-reported outcomes, the Oswestry Disability Index (ODI) walk item and the Physical Function Scale of the Swiss Spinal Stenosis Questionnaire (PFS) within an LSS population.

STUDY DESIGN/

SETTING

Measurement properties study utilizing observational cohort design.

PATIENT SAMPLE: A total of 124 subjects with LSS (confirmed by a spinal surgeon from a combination of imaging and clinical symptoms) were recruited prospectively from the private consulting rooms of 3 spinal surgeons between May 2019 and August 2023.

OUTCOME MEASURES: Wrist activity monitor distance, smartphone walking app distance, SPWT, ODI walk item, and PFS from the Swiss Spinal Stenosis Questionnaire.

METHODS

Walking capacity was assessed on 3 occasions, with the first/baseline assessment before surgery. The first assessment was conducted to determine the criterion validity of the wrist activity monitor, phone app, ODI walk item and PFS by comparison against the SPWT. The reliability of all 5 walking capacity measures was assessed 1-3 weeks following initial assessment. Responsiveness was assessed in a subset of participants following their spinal procedure. Responsiveness was calculated using both the SPWT and a self-perceived recovery scale as the external anchor. COSMIN criteria was used to determine adequate and inadequate results for each measurement property.

RESULTS

Criterion validity of the activity monitor (n=123) and smartphone app (n=89) was strong with r=0.96 and r=0.93, respectively. There was only a moderate correlation between the SPWT and subjective walking measures ODI (n=121) and PFS (n=123) with r=-0.33 and r=-0.39 respectively. Objective measures of walking capacity demonstrated excellent reliability based on the intraclass correlation coefficient (ICC) results. The activity monitor, smartphone application and SPWT showed an ICC of 0.96, 0.97, and 0.96, respectively. The PFS showed good reliability (ICC=0.76) while the reliability of the ODI walk item was poor (ICC=0.45). When responsiveness was assessed based on self-perceived recovery, only the PFS demonstrated acceptable responsiveness with an Area Under the Curve (AUC) value of 0.73. Responsiveness anchored on SPWT distance was excellent for the activity monitor (AUC=1.00) and the smartphone application (AUC=1.00), but neither the ODI walking (AUC=0.69) nor PFS (AUC=0.50) demonstrated acceptable responsiveness.

CONCLUSIONS

A low-cost wrist activity monitor and phone walking app are valid, reliable, and responsive measures of walking capacity in LSS and can be recommended in clinical and research assessment. These measures can overcome the existing barriers in the use of objective measures. These measures may also support a move away from single time-point measures of LSS, that could allow clinicians and researchers to capture the fluctuation in walking capacity evident in people with LSS.