Study design and setting

This is a 24-month, multicentre, prospective cohort study conducted at Zealand University Hospital Køge, Denmark and the Spine Centre of Southern Denmark, Kolding. Participants in this study are patients with symptomatic MRI-verified central canal stenosis undergoing in-patient surgical decompression at either centre. Measurements of postural control and activity will be undertaken up to 3 months before decompressive surgery, as well as 3, 6, 12 and 24 months postoperatively, as presented in figure 1. A recruitment phase of 12 months is planned, and study end is defined as ‘last patient out’. Study inclusion began in September 2023, and the study is expected to be completed by December 2026.

Figure 1

Flow chart of participant timeline and patient recruitment flow during the study period. LSS, lumbar spinal stenosis.

Recruitment

Patients referred for surgical decompression at either the Spine Section at Zealand University Hospital, Køge or the Spine Center of Southern Denmark, Kolding, will be invited to participate. The full scope of the study, as well as any associated risks and benefits, will be explained prior to the collection of informed consent. Participants will be assessed for eligibility (see below), and eligible participants will undergo preoperative data collection procedures prior to their scheduled surgery.

Eligibility criteria

Inclusion criteria: (1) age ≥65 years at time of evaluation, (2) planned for decompressive spinal surgery due to symptomatic LSS at ≥1 level, (3) central canal LSS grade C or D by Schizas’ classification51 at ≥1 level, verified by MRI and (4) a minimum of 3 months of unsuccessful non-operative treatment.

Exclusion criteria: (1) signs of malignancy or infection in the spinal column, (2) severe comorbidities defined as American Society of Anesthesiologists’ (ASA) physical status classification score ≥3, (3) revision surgery defined as previous decompressive surgery at the same vertebral level, (4) any spinal surgery up to 1 year prior to the date of evaluation, (5) Mini Mental State Examination score ≤27 or (6) degenerative spondylolisthesis ≥3 mm on preoperative imaging.

Participants may withdraw from the study at any time. Participants will be inquired as to their reason for withdrawal, though providing a reason is not mandatory. Reasons for withdrawal, as well as causes of loss to follow-up, will be recorded. To limit the prevalence of competing comorbidities that may influence the physical activity levels and postural control ability of our cohort, it was decided to exclude patients with ASA score ≥3.

Intervention

All patients will undergo intervention in the form of open decompressive surgery at ≥1 vertebral level due to central canal stenosis. Treatment before and after the surgical procedure will remain as standard-of-care, including rehabilitative and physiotherapeutic efforts.

Outcome measuresPrimary outcome measure of postural control

The primary outcome measure will be the mean change in sway area of COP (cm2) from baseline (preoperative) to 24 months postoperative. Sway area denotes the area in which COP oscillates during quiet standing and will be defined as the area of the ellipse which contains the true mean of (Yn,Xn)1≤n≤N with 95% probability, as derived by Schubert and Kirchner.52

Measures of postural control will be calculated from the vertical forces recorded from the WBB using the BrainBLoX interface developed by the neuromechanics laboratory at University of Colorado Boulder.53 The WBB will be calibrated according to guidelines set forth by Clark et al.34 As data from the WBB are sampled at a variable rate, a Sliding Window Average with Relevance Interval Interpolation method will be used to resample the signal at 100 Hz. The resulting data will be processed using a fourth-order, low-pass Butterworth filter with a 10 Hz cut-off frequency and centred using the arithmetic mean. Postural control measures will be calculated using custom software based on standardised methods set forth by Quijoux et al.33

Consultations will be conducted in an unperturbed room. Measurements are taken while patients stand on the WBB in a standardised and consistent location, with their arms hanging by their sides and being told to focus on a visible point ahead of them at head height. Patients will be told to stand as still as possible.54

Measurements are recorded through two sets of five trials of 90 s each, with eyes open and eyes closed, respectively. Rests of 15 s intervals are given between each trial, and a rest of 60 s will be given between each set. The number of trials and the lengths of these are based on recommendations from reliability studies by Doyle et al and Ruhe et al.54 55 A Visual Analogue Scale concerning the current degree of leg pain will be recorded before each set.

The need to take a corrective step during the trial to avoid falling will be noted if necessary.

Secondary outcome measures of postural control

Secondary outcome measures will be changed in mean COP velocity in the anteroposterior and mediolateral plane (mm) from baseline to 24 months postoperative, and change in the Danish version of the Mini Balance Evaluation Systems Test (Mini-BESTest) and Tandem test score from baseline to 24 months postoperative. The Mini-BESTest was initially developed to aid in the analysis of various postural control systems while being brief and simple to administer; it has since been validated for use in elderly populations and has previously been applied in studies concerning patients with LSS, to identify various balance limitations.56–59

Primary outcome measure of activity level

The primary outcome measure will be the mean change in Total Activity Counts per day from baseline to 24 months postoperative, as measured by the ActiGraph wGT3X-BT accelerometer.

At each data collection visit, participants will be instructed to wear the accelerometer for seven consecutive days and instructed to wear the accelerometer on their hip during all waking hours except when bathing or swimming. A wear-time diary will be issued, and participants will be instructed in the use hereof.

The accelerometer will be set to sample at a rate of 30 Hz, and an epoch length of 60 s will be implemented. As the study population concerns elderly patients, the low-frequency extension filter will be applied, to better help capture slower movements that may be prevalent among this group. Non-wear time will be defined using the Choi et al algorithm, consisting of 90 min of zero counts per minute (cpm) with an allowance of 2 min of activity when it is placed between two 30 min windows of zero cpm.60 Sampling rate, epoch length, filter choice and non-wear-time definition are chosen in accordance with recommendations set forth by Migueles et al for older adults, while activity intensity levels (sedentary, light, moderate and vigorous intensity) will be defined using cut-points generated and validated by Bammann et al.31 38 47

To achieve a suitable degree of intraclass correlation between measurement intervals, a valid day of measurement will be defined as awake wear-time ≥10 hours per day, and a valid week of measurement will be defined as ≥4 valid days of wear-time during the 7-day interval.31

Secondary outcome measures of physical activity

Secondary outcome measures will be change in time spent in MVPA per day (min/day) and sedentary time (hours/day) from baseline to 24 months postoperative, as calculated from accelerometer data, and change in self-reported activity level as measured by International Physical Activity Questionnaire short form (IPAQ-SF) score. The IPAQ-SF will be issued to all patients upon completion of a week of activity measurements using the wGT3X-BT accelerometer. While the validity of the IPAQ-SF concerning absolute measurements of activity is dubious, for the purpose of estimating self-reported physical activity, the IPAQ-SF has proven sufficient.61

Patient-reported outcome measures

Primary outcomes of postural control and activity levels will be compared with patient-reported outcome measures of quality-of-life and physical function as additional secondary outcome measures. Change in quality-of-life and physical function will be collected through surveys using the Danish versions of the European Quality of Life-5 Dimensions (EQ-5D) questionnaire, the Short Form Health Survey (SF-36) and the Zurich Claudication Questionnaire (ZCQ). Both the EQ-5D and SF-36 have been translated and validated for use in Danish and are used extensively in clinical settings in Denmark.62 63 The ZCQ, also known as the Swiss Spinal Stenosis Measure, was initially developed for the purpose of assessing symptom severity for LSS, physical function and surgical management satisfaction.64 65

Preoperative data collection will be performed by the principal investigator (PI) at both centres, while postoperative data collection will be performed by licensed physiotherapists employed at either study centre, trained in the use of the data collection tools by the PI. As such, assessors of postoperative physical activity and postural control are blinded to the participants’ preoperative status. All data collection materiels will remain identical for each study site, and all data collectors are instructed to contact the PI immediately should any deviation from the protocol occur. Minor deviations, defined as those with no discernible impact on the collected data, will be resolved by the PI. Major deviations, defined as those with the potential to impact the collected data, will be noted and resolved by consensus from the steering committee, consisting of all authors of the study.

For an overview of data collection tools administered during the study period, see table 1. The order of tests during each data collection visit will remain consistent to account for subject fatigue. For an overview of the order of tests, see figure 2.

Table 1

Overview of the data collections tools administered during the course of the study

Figure 2

Overview of the order of tests administered during each data collection visit. Total time needed for all tests at each visit is estimated to be 65 min. Mini-BESTest, Mini Balance Evaluation Systems Test, VAS, Visual Analogue Scale; WBB, Wii Balance Board.

Sample size

Previous studies on the effects of decompressive surgery on postural control and physical activity are sparse and show varying effect sizes depending on methodology and choice of outcome variable. For the calculation of sample size, a Cohen’s d of 0.40 and 0.35 was calculated for the postural control and physical activity measures respectively, using results from previously published literature with approximate parallel methodology.21 25 28 66 The lower value was decided on to ensure the necessary power for both outcomes, though due to the scarcity of literature on the subject, a degree of error must be attributed to the calculated effect sizes. Assuming 80% power and α=0.05 for a two-sided test in a paired setup, and with an assumption of 10% loss to follow-up, a population of 80 participants will be necessary. An interim analysis of the standard deviation (SD) of the preoperative data is planned after the first 30 patients have been included, which may result in an amendment to the target sample size if the calculated SD varies greatly from the literature.

Data monitoring, collection and analysis

A data monitoring committee (DMC) has been assembled consisting of the PI and the study guarantor, with responsibility for maintaining the data management plan and monitoring data collection and entry. All data pertaining to the study will be entered into a Research Electronic Data Capture (REDCap) database. REDCap is a secure, web-based software platform designed to support data capture for research studies, which enables audit trails for tracking data manipulation and export procedures.67 68 Data from separate study sites will be gathered at the Spine Section, Zealand University Hospital; physical data will be collected and transported by the PI while digital data will be transferred using the Secure Information Facility maintained by Copenhagen University. Data will be double-entered into REDCap by two independent data entry operators, and any conflict will be resolved by consensus from the DMC. Full access to the dataset will be limited to the members of the DMC while a pseudonymised version will be made available to all members of the study group.

Statistical analysis will be conducted using R V.4.3.2 (R Core Team, 2021). Patient characteristics and variables of postural control and activity levels at baseline and at each data collection visit will be detailed. Continuous variables will be presented as means with SD or medians with IQR, depending on normality. Categorical variables will be presented as frequencies and percentages. The Kolmogorov-Smirnov test will be used to test for normality where appropriate.

Differences between means from baseline to 24 months postoperative will be reported with 95% CIs and will be assessed using paired t-tests for continuous variables, Wilcoxon signed-rank test for ordered categorical variables, and McNemar’s test for binary categorical variables. Multivariable linear regression analysis will be used to adjust for potential confounders and linear regression will be used to test for trends across the data collection visits. Throughout, a p value of 0.05 will be considered statistically significant.

Missing data

We estimate loss to follow-up at 10%. Additionally, data may be missing due to participants missing one or more data collection visits during the study period. Missing data during the postoperative data collection period will be handled using a linear mixed effects model in the main analysis. Additional sensitivity analyses will be conducted using multiple imputation models to assess the robustness of the results to different missing data assumptions. Patients who do not participate in any postoperative data collection visits will be excluded from the analysis.

Patient and public involvement

A patient committee (PC) has been assembled during the preliminary study preparations, with the stated purpose of providing patient-directed feedback during the study period. The PC members consist of volunteers from the list of included participants and rotate with respect to availability. Meetings are planned to occur before, during and after the inclusion period, and feedback is gathered concerning, but not limited to, study design, burden of participation, patient recruitment and dissemination of results.