This prospective study was conducted between October 2023 and March 2024. We used two distinct analytical cohorts to validate the psychometric properties of K-FLS and K-ZCQ, viz., a cross-sectional cohort for reliability and validity and a longitudinal cohort for responsiveness.

Inclusion criteria were as follows: Participants were aged 50–80 years and had a diagnosis of LSS. Surgical candidacy was defined as a previous LSS surgery within the preceding 24 months or a planned operation at enrollment. Operational definition of LSS. LSS required lumbar magnetic resonance imaging (MRI) findings consistent with Schizas morphological grade ≥B together with ≥1 of the following: ambulation-limiting neurogenic claudication, sensory symptoms in the buttock and/or lower leg (pain, numbness, or tingling), distal lower limb motor weakness, or bowel/bladder dysfunction [14,15].

Exclusion criteria were as follows: To minimize confounding, we excluded individuals with serious systemic illness that may affect disability or general health (e.g., sepsis and active malignancy), peripheral polyneuropathy, myelopathic signs (including gait imbalance), acute osteoporotic vertebral compression fracture, gait limitation primarily attributable to lower extremity osteoarthritis (ankle, knee, or hip), ischemic heart disease or peripheral arterial occlusive disease consistent with vascular claudication, or any other neurological/musculoskeletal disorder that independently impairs gait. All patients presented with typical symptoms of LSS, which include neurogenic intermittent claudication and numbness and/or leg pain. In all patients, diagnosis was confirmed by more than one spine surgeon. The level(s) of stenosis were localized using MRI.

Of the initial 140 patients (84 women and 56 men), the following 11 were excluded: five with osteoporotic spine fractures, four with knee osteoarthritis, and two with parkinsonism. Finally, 129 patients (77 women and 52 men) were included in this study, of whom 123 (74 women and 49 men) responded to the second survey. The average age of the 123 participants was 68.8±6.7 years at the time of the surveys.

The first assessment was performed at the outpatient clinic using a self-administered questionnaire. The second assessment for test–retest reliability was conducted after 2 weeks. To maximize patient compliance and minimize loss to follow-up, the second survey was administered through a structured phone interview for patients who could not visit the clinic.

We used K-FLS, K-ZCQ, K-ODI, and VAS. LSS symptoms were measured using a LSS symptom–specific scale (FLS), which was designed by Sekiguchi et al. [7–9] and asks patients about their symptoms in specific situations. The K-FLS consists of 25 items, scored from 4 (worst) to 0 (best), and the score is calculated as the sum of answers to each question. Overall scores on the questionnaire range from 0 to 100, with higher scores indicating worse conditions. The K-FLS identified the following four major factors: physical pain (five items), activity limitations (11 items), physical activity (six items), and mental health (three items). The K-ZCQ consists of three domains, each rated on a Likert-type scale. The symptom severity domain has seven questions with five response options, the functional disability domain contains five questions scored on a 4-point scale, and the treatment satisfaction domain consists of six questions that are also rated on a 4-point scale. In all domains, greater numerical values correspond to worse disease status.

The reliability of K-FLS and K-ZCQ questionnaires was determined using test–retest comparisons between initial and subsequent evaluations. Intraclass correlation coefficients (ICC, 2, 1) and agreement statistics were used to measure reliability for each item. Internal consistency was evaluated using Cronbach’s α. As the K-FLS and K-ZCQ utilize ordinal Likert scales, Spearman’s rank correlation coefficients were used to determine concurrent and construct validity. Convergent validity (criterion validity) was evaluated to determine how well the questionnaire correlates with another valid instrument measuring similar constructs. Accordingly, the correlations among K-FLS, K-ZCQ, and K-ODI were calculated. Spearman’s rank correlation coefficient values of ≥0.40 were considered satisfactory (r≥0.81–1.0 as excellent, 0.6–0.80 as very good, 0.41–0.60 as good, 0.21–0.40 as fair, and 0.0–0.20 as poor) [16]. Finally, responsiveness to change as a psychometric property of the questionnaire was also evaluated; as such, patients’ preoperative and postoperative scores were compared using the paired t-test to determine whether the scale could capture changes after surgery. Responsiveness was further evaluated using effect size (ES). This index was derived by dividing the mean difference between baseline and 3-month postoperative scores by the standard deviation of baseline values. According to conventional criteria, an ES of 0.8 indicates a large magnitude of change [17]. Larger ES values indicate greater sensitivity of the instrument in capturing clinical improvement. Data analysis was performed using SPSS ver. 16.0 (SPSS Inc., Chicago, IL, USA).

The adequate sample size was determined based on the following parameters: n=Z² P(1–P)/d², where n is the sample size, Z is the level of confidence (for the 95% confidence level, which is conventional, the Z-value is 1.96), P is the expected response rate of 90%, and d is the precision of 5% [18]. Accordingly, the sample size calculation using power analysis yielded 138. Hence, we aimed to enroll approximately 140 participants. This study was conducted according to the Declaration of Helsinki, and the study protocol was reviewed and approved by the Institutional Review Board of Pusan National University Hospital (No. D-2109-003-184). Written informed consent was obtained from all participants.