Supervised Physical Therapy Versus Unsupervised Home Exercise After Surgery for Lumbar Disc Herniation

Objectives:

The primary aim of this thesis is to determine whether unsupervised home exercises are not inferior to supervised physical rehabilitation in patients after surgery for lumbar disc herniation.

Ho: There is a difference in improvement in outcomes, as measured by the change in Oswestry Disability Index at six months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

H1: There is no difference in improvement in outcomes, as measured by the change in Oswestry Disability Index at six months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

Secondary aims of this study are to determine if there is a difference in cost-effectiveness in patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises. Differences in effectiveness will be determined using Quality Adjusted Life Years (QALYs) calculated from the EQ-5D and the SF-6D. Both these health state utility values will be used, as studies in low back patients have shown a marked difference in the calculation of QALYs when one or the other is used (9-11).

Ho: There is a difference in gain in health state utility, as measured by the EuroQOL-5D at 12 months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

H1: There is no difference in gain in health state utility, as measured by the EuroQOL-5D at 12 months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

Ho: There is a difference in gain in health state utility, as measured by the SF-6D (transformed from the SF-36) (12) at 12 months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

H1: There is no difference in gain in health state utility, as measured by the SF-6D (transformed from the SF-36)(12) at 12 months post-operative compared to baseline, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

In order to calculate costs, both direct (cost of rehabilitation) and indirect costs (lost wages) need to be determined.

Ho: There is a difference in number of days between surgery and return to work, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

H1: There is no difference in number of days between surgery and return to work, between patients undergoing supervised physical rehabilitation in patients after surgery for lumbar disc herniation compared to unsupervised home exercises.

Methods:

Design:

Randomized clinical trial, 1:1 parallel group, non-inferiority, single blinded trial comparing (Arm1) supervised physical rehabilitation (SPT) in patients after surgery for lumbar disc herniation to (Arm 2) unsupervised home exercises (UHE). Performed in accordance with the CONSORT guidelines (13).

A computer generated randomization random variable block of 4-6 at 1:1 allocation will be created by a researcher that will not be involved in recruitment, screening or identification of potential subjects (LYC). The assigned treatment arm will be kept in sequentially numbered, opaque, sealed envelopes, which will be opened after the informed consent process is complete. The randomization envelope and sheet will be kept with the subject's research record.

Stratification factors, such as age, gender, race and approach (microdiscectomy versus open) are not included for balancing the randomization but will be included in the analyses. Since the primary outcome measure is a difference within same subject (subject is their own control), some variability due to these factors will be reduced. In additional, due to block randomization and assuming random patient arrival (such as not all patients scheduled for microdiscectomy will be coming to clinic in the beginning of the study and then followed by patients scheduled for open discectomy), there will be enough representation based on these factors. In addition, at the Center for Spine Surgery microdiscectomy can denote the use of a microscope during an open discectomy or the use of tubular, limited dissection without a microscope.

Study population:

Study subjects will be recruited from patients assessed and scheduled for primary discectomy for a lumbar disc herniation from L1 to S1 at the Center for Spine Surgery and Research, Sygehus Lillebælt.

Patient course:

During the initial consultation at Sector for Spine Surgery and Research, Middelfart, study subjects will undergo a neurological exam including range of motion, sensibility, reflexes and strength in the lower extremities and back. Patients are also asked to complete multiple questionnaires (basic personal information, occupational status, DaneSpine questionnaire, ODI, SF-36, and EQ-5D) as mentioned in section 5.1 Data collection. These procedures are all standard of care.

Patients identified by the attending surgeon as eligible for inclusion are invited to participate in the study by the Principal Investigator (RTP). Patients who express an interest in participating will be given written and oral information on the background, purpose and method of the study. Information and inclusion of participants will be conducted in accordance with the guidelines of The Health Research Ethics Committee System in Denmark, from which approval will be sought.

Before surgery when a subject has been screened and included in the study, and the informed consent process is completed and the informed consent document is signed, the randomization envelope will be opened. The subject will be randomize to either the supervised physical therapy or the unsupervised home exercise arm using a variable blocks of 4-6 with 1:1 allocation.

The supervised physical therapy group will be rehabilitated 4 weeks postoperative at the municipality rehabilitation unit with training sessions 2 times a week in 8 weeks. The home exercise group will be discharged with an informational booklet and advice to begin normal daily activities as soon as possible. All patients will be asked to keep a diary of contact to external health care professionals, amount of trainings and painkillers.

All patients participating in the study will undergo standard operative treatment, either open or microscopically, at Center for Spine Surgery and Research, Middelfart, and scheduled for an outpatient clinical visits at 1 and 3 months post-operative. Control at 1 month is conducted by a physiotherapist and as part of standard care. All patients must keep a diary of trainings, external medical- and physiotherapist

Statistics:

The primary analysis will be an intent-to-treat analysis with the change in ODI scores between six months and baseline as the primary outcome measures. Repeated measures ANOVA with the baseline ODI score as a covariate will be used to compare the two arms. In addition Fisher's exact test will be used to determine differences in categorical variables will between groups.

Sample size rationale:

Patients will be randomized in two groups, supervised physical therapy (SPT) and unsupervised home exercises (UHE). The main outcome variable is the Oswestry Disability Index, which is a disease specific outcome measure, and will be used for sample size justification. Based on the published literature, the baseline ODI for patients with disc herniation is 48 points with a standard deviation of 19 units (15)This is similar to the data from N2QOD. Change in ODI score from the baseline to one year was 24 points (95% CI: 21-27) with estimated standard deviation of 16 points.

Using two-sided, two-sample independent t-test, we can detect a mean difference of 0.50 SD units at alpha=5% and power=80 with n1=64 and n2=64. This translates into the effect size of 10 points. Since final follow-up is at one year an interim analysis is not planned. We plan to randomize subjects in 1:1 ratio in supervised physical therapy and unsupervised home exercise arms in randomized block sizes of 4 and 6. Stratification factors, such as age, gender and race are not included for balancing the randomization but will be included in the analyses. As the primary outcome measure is a difference within same subject (subject is their own control), some variability due to these factors may be reduced. In addition, due to block randomization and assuming random patient arrival, there will be enough demographic representation based on these factors. We will increase sample size by 15% to account for drop out or death before the completion of the study. Thus we need to enroll 74 subjects in each arm. Using the method of Diggle et al. for sample size calculation in longitudinal studies and assuming a positive correlation around 0.4, we will have at least 85% to detect the same effect size.

Quality control and quality assurance:

The study will be registered at the Ethical Committee of Southern Denmark and the Danish Data Protection Agency.

All patient data, including information on private matters or other confidential information, will be strictly confidential and stored according to the Danish Open Administration Act, the Danish Act on Processing of Personal Data and the Health Act.

The study will be performed in accordance with the guidelines recommended by STROBE (STrenghtening the Reporting of OBservational studies in Epidemiology).