Post-operative management
Function after spinal treatment, exercise and rehabilitation (FASTER): A factorial randomised trial to determine whether the functional outcome of spinal surgery can be improved.
Alison H McGregor1*, Caroline J Doré2, Tim P Morris3, Steve Morris4, Konrad Jamrozik5 on behalf of the FASTER team
Author for correspondence: Alison McGregor
1 Alison H McGregor, Surgery & Cancer, Faculty of Medicine, Imperial College London, Charing Cross Hospital Campus, London W6 8RP
Phone +44(0)203 313 8831
Fax +44(0)203 313 8835
2 Caroline J Doré, MRC Clinical Trials Unit, 222 Euston Road, London NW1 2DA
3 Tim P Morris, MRC Clinical Trials Unit, 222 Euston Road, London NW1 2DA
4 Steve Morris, Department of Epidemiology and Public Health, University College London, London WC1E 6BT
5 Konrad Jamrozik, School of Population Health and Clinical Practice Office, North Terrace, University of Adelaide, SA 5005, AUSTRALIA
Trial Registration
Current controlled trials ISRCTN46782945 / UK CRN ID: 2670
Acknowledgements
This study was supported by Arthritis Research UK
We would like to thank all the patients for making this study possible and their patience with the many questionnaires involved in this study and the FASTER team for making this study possible.
FASTER TEAM
Trial Steering Committee
Professor Jeremy Fairbank
Mrs Caroline Doré
Lord ER Oxburgh
Professor Jennifer Klaber Moffett
Mr Steve Eisenstein
Mr Tim Morris
Professor Alison McGregor
Trial Coordinator – see below
Trial Management Team
Principle investigator: Alison McGregor
Trial Coordinators: Ania Henley, Jack Kerr, Dr Julia Toschke
Physiotherapy Coordinators: Carla Ashford, Janet Deane, Carol Waugh
Data Monitoring & Ethics Committee
Professor Sarah Hewlett
Dr Simon Bowman
Dr Martyn Lewis
Participating Hospitals
Charing Cross Hospital
Guy’s and St Thomas’ Hospital
Heatherwood Hospital
Ravenscourt Park Hospital
Royal Free Hospital
St Mary’s Hospital
Wexham Park Hospital
Contributing Surgeons
Mr M Akmal
Mr N Anjarwalla
Mr R Bradford
Mr N Dorward
Mr T Ember
Mr D Fahy
Professor S Hughes
Mr J Johnson
Mr K Lam
Mr J Lucas
Mr N Mendoza
Ms M Murphy
Mr D Nandi
Mr J O’Dowd
Mr K O’Neill
Mr D Petersen
Mr C Shieff
Mr L Thorne
Mr C Ulbricht
Professor J van Dellen
Trial Physiotherapists
Katie De Albuquerque
Nathan Allwork
Hayley Boyle
Ann Bryan
Rachel Freeman
Jenny Heal
Annys Hawkins
Catherine Heathcote
Frances Honniball
Gary Jones
Biljana Kennaway
Gemma Kettle
Anna Kuppuswamy
Claire Marshall
Ann McCarthy
Hannah Mundy
Sandra Noonan
Sara Parker
Camilla Pleydell-Bouverie
Rebecca Portwood
Amy Powel
Adam Royffe
Victoria Salmon
Laura Segar
Alison Sentence
Tom Stenning
Hillary Thompson
Lorraine Tomeldan
Amanda Wall
Structured Abstract
Study Design
This was a multi-centre, factorial, randomised controlled trial on the post operative management of spinal surgery patients, with randomisation stratified by surgeon and operative procedure.
Objective
This study sought to determine whether the functional outcome of two common spinal operations could be improved by a programme of post-operative rehabilitation that combines professional support and advice with graded active exercise commencing 6 weeks after surgery and/or an educational booklet based on evidence-based messages and advice received at discharge from hospital, each compared with usual care.
Summary of background data
Surgical interventions on the spine are increasing, and whilst surgery for spinal stenosis and disc prolapse have been shown to be superior to conservative management, functional outcome and patient satisfaction are not optimal.
Methods
The study compared the effectiveness of a rehabilitation programme and an education booklet for the postoperative management of patients undergoing discectomy or lateral nerve root decompression surgery, each compared with “usual care” using a 2 x 2 factorial design, randomising patient to four groups; rehabilitation-only, booklet-only, rehabilitation-plus-booklet, and usual care only. The primary outcome measure was the Oswestry Disability Index (ODI) at 12 months, with secondary outcomes including visual analogue scale measures of back and leg pain.
Results
338 patients were recruited into the study and measurements were obtained pre-operatively and then repeated at 6 weeks, 3, 6, 9 and 12 months post-operatively. Twelve months post operatively the observed effect of rehabilitation on ODI was -2.7 (95% CI -6.8 to 1.5) and the effect of booklet was 2.7 (95% CI -1.5 to 6.9).
Conclusions
This study found that neither intervention had a significant impact on long term outcome.
Key words
Randomised controlled trial; post-operative management; rehabilitation; education; factorial
Mini abstract
This randomised controlled trial explored 4 different approaches to the post-operative management of spinal decompression and discectomy surgery. 338 patients were recruited and followed over a one year period using measures of function, pain, and health related quality of life.
Introduction
Spinal surgery rates have increased across the world 1;2, particularly surgery for spinal stenosis which in part reflects the growth in the elderly population3. This study focuses on spinal discectomy surgery for a herniated lumbar intervertebral disc and nerve root decompression surgery for lumbar spinal stenosis, two of the commonest spinal procedures performed1;4;5. Both procedures are deemed superior to conservative treatment although the benefits are known to diminish with time1. However, functional improvement and patient satisfaction are varied; in decompression surgery success rates range between 58-69%5-7 and satisfaction from 15-81%2;8;9; and for discectomy success rates ranged between 65-90%2;10;11, with less known regarding patient satisfaction.
Consequently, this study sought to improve the functional outcome and patient satisfaction following spinal surgery by addressing post-operative recovery. Further rationale for study has been previously published12. Briefly, it builds upon survey findings and literature reviews of post-operative management which identified wide variations in practice amongst surgeons, indicative of uncertainty which was further supported by a lack of evidence to justify many of these practices 13-17, and a growing literature base highlighting muscle dysfunction in spinal patients18;19 and subsequent damage to the muscles with surgery20;21. Two approaches to addressing this were identified; formal rehabilitation and educational material to encourage self-management. A systematic review of rehabilitation following lumbar disc surgery22 provided evidence that rehabilitation could enhance outcome, and subsequently a number of small studies23-28 reported mixed findings. Similarly educational material has been found to have an impact on patient outcome24;29;30. However, many of these studies were underpowered, or of poor design with few comparing the interventions to usual care.
The objective of this factorial randomised controlled trial (FASTER ~Function after spinal treatment, exercise and rehabilitation) was to evaluate the benefits of a rehabilitation programme and an education booklet for the postoperative management of patients undergoing discectomy or lateral nerve root decompression, each compared with “usual care”. Our hypothesis is that a programme of post-operative rehabilitation that combines professional support and advice with graded exercise will improve the long-term outcome of surgery, and that appropriate educational information will also improve outcome but to a lesser degree than rehabilitation. We assume that the effect of the combination of the two interventions will be additive i.e., there will be no interaction.
Methods
FASTER is a multi-centre, parallel group, factorial, randomised controlled trial which aims to determine the optimal post-operative management strategy following spinal surgery12. The study was approved by Hammersmith and Queen Charlotte’s & Chelsea Hospitals Research Ethics Committee, with site approval for 7 hospitals in the West London Region. Patients presenting with symptoms, signs, and radiological findings of either lateral nerve root compression or disc prolapse scheduled for surgery were recruited and consented into this study and randomised using a 2 x 2 factorial design to receive:
· Factor 1 – either a six-week programme of post-operative rehabilitation or the relevant surgeon’s usual postoperative care (according to the surgeon’s discretion and routine practice):
· Factor 2 – either an educational booklet (“Your Back operation” 31), or the surgeon’s usual advice.
This created four study groups; rehabilitation-only, booklet-only, rehabilitation-plus-booklet, and usual care only. Allocation to a study group was by central telephone randomisation stratified by surgeon and surgical procedure using random permuted blocks to ensure that each participating surgeon and each surgical procedure had approximately equal numbers of patients allocated to each group. Treatment allocation was concealed prior to surgery to avoid selection bias during recruitment. Patients were notified of their randomisation following their surgery and those patients allocated to either the booklet-only group or the rehabilitation-plus-booklet group received the booklet entitled “Your Back operation” 31 on discharge. Patients were assessed using a range of validated outcome measures pre-operatively and then at 6 weeks, 3, 6, 9 and 12 months post-operatively.
Study population
Patients were recruited by the trial coordinator from the surgical waiting lists of the contributing surgeons (20 in total; 8 orthopaedic and 12 neurosurgical) at the different hospital sites and written informed consent obtained. Eligible patients included those awaiting spinal surgery with either (a) signs, symptoms and radiological evidence of lateral nerve root compression, that is, patients presenting with radicular pain with an associated neurological deficit or with neurogenic claudication (pain in the buttock, thigh or leg that improves with rest), or (b) lumbar disc prolapse, that is, patients with root symptoms and signs and MRI confirmation of lumbar disc herniation. Patients presenting with any of the following were excluded from participation; any condition where either the intervention or the rehabilitation may have an adverse effect on the individual; previous spinal surgery; spinal surgery where a fusion procedure was planned due to the unknown hazards of the activity programme for this type of surgery; pregnant women; inadequate ability to complete the trial assessment forms; unable to attend or unsuitable for rehabilitation classes. Following poor compliance with return of baseline forms at the start of the trial, patients were only allocated a trial number and treatment group after successful completion of these forms; those not complying were removed from the trial.
Trial Interventions
Surgical Intervention: Participants underwent spinal surgery according to their surgeon’s routine practice for that condition and the surgical details were recorded.
Rehabilitation Programme: Patients in the rehabilitation arms were invited to commence the intervention 6 to 8 weeks following their surgery. The programme comprised 12 standardised one hour classes run twice weekly by an experienced physiotherapist. They included general aerobic fitness work; stretching; stability exercises; strengthening and endurance training for the back, abdominal and leg muscles; ergonomic training; advice on lifting and setting targets; and self-motivation along with an open group discussion at the end of each class. Compliance was based on attendance records, and the quality and content of the classes were reviewed regularly. Any clustering effects due to therapist or treatment centre were considered negligible due to the large number of physiotherapists running classes, thus patients were routinely crossing between therapists.
Educational Booklet: Patients in the booklet arms received a copy of “Your Back operation” 31 on discharge from hospital. The design and content of which have been previously reported32.
Usual care: Patients receiving usual care were managed according to the relevant surgeon’s usual practice. Previous work has shown this to be varied and limited13.
Outcome Measures
The Oswestry Disability Index (ODI, version 2.1a), a disease-specific validated patient-completed questionnaire documenting function33;34 was the primary outcome measure. It is scored from 0-100 and interpreted as a disability percentage with lower scores representing lower levels of disability. Secondary outcome measures included 10cm visual analogue scales (VAS) which recorded average back and leg pain35, which were expressed as a percentage with a lower score indicative of less pain; the hospital anxiety and depression (HADS) questionnaire recorded anxiety and depression with scores between 0-21 - where higher values represent a greater level of anxiety /depression36; the Fear Avoidance Beliefs Questionnaire (FABQ) was used to assess pain behaviours giving scores between 0-24 with higher scores representing stronger fear avoidance beliefs 37; the EQ-5D was used to determine Health Related Quality of life, the thermometer of the score was used to assess health which was rated between 0-100 with 100 representing a perfect health state38;39; and return to work. Following recent findings the relative severity of low back pain to leg pain was also determined (Average leg pain minus average back pain)40. Baseline information was captured on the participant’s occupation, working and marital status, ethnicity, age, body mass index, and smoking habits. Measures of patient expectations and satisfaction were also obtained, and a full economic analysis performed the results of which will be reported separately. It was not possible to assess outcome measures blind to the randomised intervention since all outcome measures are patient assessments.
Sample size
The trial aimed to recruit 344 patients with 86 patients in each of the 4 interventions12. This calculation was based on a 90% chance of detecting a 20% relative improvement in the ODI corresponding to a between-group difference of 8 points in the ODI and declaring it statistically significant using a two-sided alpha= 0.05. This calculation assumed a standard deviation of 20 in each group and allowed for loss to follow up of 23%, as seen in earlier descriptive work41.
Statistical Analysis
This was performed according to our protocol12. Initially a descriptive comparison of the trial groups before surgery was done to confirm that randomisation had produced balanced groups with respect to known predictors of outcome such as age, sex, type of surgery, ethnic background, marital status, body mass index, occupation type, work status and smoking status. Baseline characteristics are summarised across the four groups as no. (%) for categorical variables or mean (sd) for continuous variables. Baseline values of outcome scores – including ODI, average back pain, average leg pain, FABQ, HADS and VAS health summary – were summarised as mean (sd) if approximately normal.
The primary outcome was the between-group difference in score on the ODI at one-year follow up, based on intention-to-treat. Secondary outcomes included average back and leg pain, FABQ, HADS anxiety & depression scores and VAS for overall health (all measured at one year follow up). Groups were compared using analysis of covariance adjusting for baseline value of outcome and stratifying factors: surgery type as a fixed effect and surgeons as random effects, to increase efficiency in estimating the effect of intervention. Analyses were performed for booklet vs. no-booklet and rehabilitation vs. no-rehabilitation simultaneously. Comparisons were followed by a test for interaction of the two interventions.