Optimization of Pulmonary Rehabilitation Programmes: the OPTION Study

The study will be a prospective, randomized, multi-centre study with 3 parallel-intervention arms and blinding of outcome assessors. Subjects will be randomly allocated to 1 of 3 exercise-training arms: CT80, CTVT, or IT. The duration of the training phase will be adjusted for each arm such that the total amount of work will be equivalent to each arm. Training intensity will be ascertained with heart rate (HR) monitors. The frequency and length of training were chosen based on latest pulmonary rehabilitation (PR) guidelines. Upon completion of the 12-week intervention, subjects will be given standardized exercise recommendations to follow at home or in the community until their 1yr follow-up. Assessments will be made at baseline (week 0), program completion (week 12) and 1yr after program initiation (week 52). Randomization of groups of 6 subjects will occur in block, once all 6 subjects have completed baseline assessments. The randomisation process will consist of a computer-generated random listing of the 3 treatment allocations blocked by groups of 6 and stratified by site. The research assistant will be responsible for randomization and for coordinating assessment and intervention visits with subjects, assessors and exercise-training supervisors. Subjects will be instructed not to talk about their group assignment to the assessors. Eligibility criteria are meant to differentiate COPD from other respiratory diseases and to ensure clinical stability and patient safety. Subjects will be recruited by a research assistant from at Sacré-Coeur and at Mount Sinai Hospital Montreal. Assessments will take place at Sacré-Coeur. Subjects will complete the intervention phase at Sacré-Coeur or at Concordia University. To date, 36 subjects have completed the study. As such, in order to achieve the targeted sample size, an additional 84 subjects will have to be randomized. We expect a recruitment rate of 25-30%. Thus, approximately 300 patients will have to be screened to meet the recruitment goal. Given that we were able to screen 200 patients over 2.5yrs of active recruitment, we anticipate that an additional 2.5yrs of recruitment from 2 sites will enable us to reach our target sample size. Follow-up measurements will require an additional 12months of data collection from the time the last patient is randomized. Accordingly, we anticipate total data collection to last 3.5yrs. We estimate that 10% of subjects will be lost to the week 12, while the number is expected to double at the week 52.

A symptom-limited incremental cycling exercise test will be performed at week 0 to rule out the presence of cardiovascular co-morbidities. Subjects will be seated on an electromagnetically braked cycle ergometer (Ergoselect 200P, Ergoline, Germany) and connected to a cardio-respiratory circuit through a mouthpiece (Jaeger Oxycon Pro, CareFusion, Germany). After 3min of unloaded pedalling, the workload will be increased in a stepwise manner up to the individual's maximal capacity. Each step will last 1min and increments of 5-10W will be used. Gas exchange parameters and HR will be measured at rest and during exercise on a breath-by-breath basis. The ventilatory threshold will be determined using the V-slope method, a computerized approach to identify the breakpoint in the carbon dioxide output-oxygen uptake relationship. Peak work rate will be defined as the highest work rate maintained at a pedalling speed of at least 50rpm for a minimum of 30sec.

A constant-load cycling exercise test will be performed to measure exercise tolerance at week 0, week 12 and week 52. Subjects will be seated on the same ergometer and connected to the same cardiorespiratory circuit. After 3min of unloaded pedalling, the workload will be set at 80% of the peak work rate achieved on the incremental cycling test. Patients will be asked to pedal against that workload for as long as possible. Inspiratory capacities will be obtained every other minute during the test to evaluate the degree of dynamic lung hyperinflation. Dyspnea and leg fatigue will be evaluated at rest and every other minute during the test with the modified 10-point Borg scale. The final measure for this test will be endurance time, defined as the duration of pedaling at a minimum of 40rpm at 80% of peak work rate.

Field walking tests are felt to be more reflective of daily living activities than laboratory tests, thereby providing the means to objectively evaluate functional status. An incremental shuttle walking test (ISWT) will first be completed to determine peak walking capacity at week 0: walking speed will initially be set at 0.50m/sec and then increased by 0.17m/sec every minute until the patient reaches a symptom-limited maximum. The endurance shuttle walking test will then be performed at week 0, 12 and 52 to measure change in functional status: subjects will be asked to walk for as long as possible at a constant, predetermined speed corresponding to 85% of peak oxygen uptake, as predicted from the ISWT. The final measure will be the distance covered. Both tests will be completed following recommendations.

Subjects will be asked to complete psychosocial questionnaires that measure variables known to be associated with COPD and with positive responses to exercise interventions. Measures will include the 21-item BDI-II to assess depressive symptoms, the 16-item ASI to assess anxiety and the SES to assess patients' confidence in their ability to successfully perform certain behaviours (e.g., exercise). In addition, the Primary Care Evaluation of Mental Disorders will be administered to characterize subjects and to detect the most common Diagnostic and Statistical Manual-IV disorders in primary care settings. Dr. Lavoie will be in charge of administering psychological questionnaires and will ensure that patients who score in the clinical range are properly referred and followed.

Subjects will complete cognitive tests that measure variables known to be associated with COPD and with positive responses to exercise interventions. This will include measures of global cognitive functioning (MoCA), executive functions (TMT B and verbal fluency), attention and vigilance (CPT), working memory (WAIS-III), episodic verbal memory (RAVLT) and visuospatial abilities (ROCF). Alternative versions will be used to avoid test-retest effect. Dr. Gagnon will be in charge of administering these tests and will ensure that patients who score in the clinical range are properly referred and followed.

Health-Related Quality of Life (HRQL) will be measured using both generic and disease-specific questionnaires. The Medical Outcomes Study Short Form 36-Item Health Survey (SF-36) will be used to obtain a general HRQL score to characterize our study sample, generating summary scores that can easily be compared with existing population norms and across different disease populations. To overcome the less responsive nature of generic instruments, we will also use the CRQ, a widespread disease-specific HRQL questionnaire which measures patients' dyspnoea, fatigue, emotional function and mastery. Total score on the CRQ will be used to measure change in HRQL from week 0 to week 12 to week 52. HRQL questionnaires will be administered by Dr. Lavoie.

Compliance refers to the degree to which a patient's behaviour concurs with instructions from a health care practitioner. Patient compliance to the training regimen will be defined as the percent time spent at the prescribed training intensity during the 12-week program. This measure will be obtained through Technogym (Bike Med 700 CE-R LED; Polar HR monitor and computer; CardioMemory software package) and which allows second-by-second tracking of individual exercise sessions. Compliance to the target intensity will be determined by assessing the percent time spent within the target HR range (±5beats/min) during the training phase of each exercise session and then averaged for the 12-week program.

Long-term exercise adherence refers to an individual's free choice to continue to exercise at the recommended levels. Adherence to exercise recommendations will be measured as exercise levels achieved at home or in the community at week 52 (numerator) divided by recommended amounts (denominator). Exercise levels achieved will be estimated from a 7-day activity record completed by the patient and coded for intensity using the Compendium for Physical Activities. The energy expended in activities corresponding to (or exceeding) the patient's target intensity level will be computed and will form the numerator. The recommended amounts of exercise will be calculated for each patient as the energy expended during the last week of PR. Exercise adherence will be measured at week 52 only, but subjects will be asked to complete the 7-day activity record at week 0 and week 12 to familiarize themselves with the procedure.

Spirometry, lung volumes and lung diffusion capacity for CO will be obtained at week 0 and at follow-up evaluations according to recommended techniques. Values will be compared to predicted normal values from the Canadian prediction equations of spirometric lung function for Caucasian adults 20 to 90yrs of age.

The primary objective of this study is to compare the impact of CT80, CTVT and IT on exercise tolerance measured after the 12-week intervention phase. Preliminary data analysis of the first 18 subjects (6 in each arm) showed an average improvement in endurance time to constant-load cycling (ES=0.32). Due to this interim analysis, α level for sample size calculation should be adjusted from an α spending function. Using Pocock, α spending function at t=0.1765 (18 from 102 subjects projected), the loss of α calculated is 0.01324 resulting in a new α of 0.03676. For a power of 0.80 and an α of 0.03676, we would need 36 patients/group for a total sample size of 108 to see a statistical difference across the groups. Due to expected 10% loss of subjects during follow-up, 40 subjects/groups (N=120) will be recruited. To examine between-group differences in exercise tolerance from week 0 to week 12, 2-way repeated-measures ANOVA will be performed, with treatment group as the between-subjects factor with 3 levels and time as the within-subjects factor with 2 levels. If a significant interaction between treatment group and time is detected, simple main effect tests will be conducted to identify the time point for which differences between treatment groups are significant. If a significant simple main effect is obtained, pairwise comparisons with TukeyB corrections will be conducted to identify between which treatments groups the differences occurred. Lost cases will be managed using the "bringing the last value forward" method. Analyses will be conducted with SPSS 19.0 (Chicago, IL). Between groups differences in exercise tolerance from week 12 to week 52 will be compared with 2-way repeated-measures ANOVA. To compare the 3 arms on functional, psychological, cognition and HRQL variables from week 0 to week 12 to week 52, mixed model repeated measure analysis will be applied. If a significant interaction between treatment group and time is detected, simple main effect tests will be conducted to identify the time point(s) for which differences between treatment groups are significant. Mean compliance to the 12-week program and mean adherence at week 52 will be compared between treatment arms with 1-way ANOVA, followed up, if needed, by pairwise comparisons to locate the differences. Patient-related covariates and potential co-interventions (pulmonary function, medications and smoking status, exacerbations, hospitalisations) will be compared between treatment groups. If significant group differences are found on these variables, ANOVAs (or Mixed models) will be replaced by analyses of covariance (ANCOVAs or Mixed models) with the variable(s) in question as the covariate(s).