Optimising Management of Lumbar Spinal Stenosis

Low back pain is the leading cause of disease burden globally, impacting over half a billion people. Low back pain is a broad classification of a condition which includes multiple clinical presentations, such as spondylosis, discogenic pain, muscle and ligament injuries, and lumbar spinal stenosis (LSS). This thesis has focused on LSS, which itself is a leading cause of pain, disability, and reduced quality of life among older adults. Eight chapters are included in this thesis that aims to improve the management of LSS. Chapter One was an introduction to LSS, its diagnosis, prevalence, and current management. Chapter Two outlined the largest known cohort study of emergency department attendance and hospital admission for low back pain conditions (e.g., LSS, disc protrusion) in New South Wales, Australia, between 2005 and 2014. This cohort study found that attendance to emergency departments and hospital admissions increased over the 10-years of the study, possibly due to the ageing Australian population, given an increase in the mean age of the people who presented over the study period. The study also found that people who were from lower socioeconomic backgrounds or who were managed under a worker’s compensation scheme had lower odds of being admitted to hospital following their presentation to the emergency department with a low back pain condition. Having demonstrated that both the attendance rates to hospital emergency departments and the mean age of the population presenting for low back pain had increased in recent years, more information on currently endorsed treatments for LSS, the most common lumbar pain presentation among older people, was sought. Chapter Three was a systematic review of international LSS clinical practice guidelines that examined the quality of the guidelines themselves, and the evidence supporting their treatment recommendations. Chapter Three found that around three-quarters of the recommendations presented in international practice guidelines for LSS were based on poor-quality evidence. There were no recommendations in the guidelines based on high-quality evidence. This study also identified that overall, guidelines made more recommendations in favour of surgery and injections than for medications and other non-surgical treatments, despite the evidence for both interventions being comparable. The absence of good evidence to support treatments for LSS highlighted the need for more high-quality studies to be conducted on LSS. To assist with increasing the number of high-quality studies conducted on LSS, a series of studies were planned to help inform the design of future studies. Initially, two systematic reviews were planned on key aspects of study design. The first review in Chapter Four was on the measurement properties of outcome measures and the second, presented in Chapter Five, on placebo-controlled trials of surgery for musculoskeletal conditions, including spinal conditions. Although the choice of appropriate outcome measures should be at the centre of trial design, their measurement properties are often overlooked. The review on outcome measures outlined in Chapter Four aimed to identify the walking test(s) with the best measurement properties in people with LSS. Walking tests were selected as the outcome of priority, as people living with LSS often nominate walking ability as the most important part of their condition that they want improved. A systematic review and meta-analysis were therefore conducted on the measurement properties used to assess LSS. This study found that there was a limited number of studies assessing the key measurement properties of validity, reliability, and responsiveness. Of the data that were available, the self-paced walking test and walking item of the Oswestry Disability Index were recommended to be used together before and after any intervention for LSS. Next, the review on placebo-controlled trials of surgery for musculoskeletal conditions was completed as presented in Chapter Five. The rationale for this review was based upon the findings of Chapter Three confirming the lack of studies utilising high-quality methods including randomisation, placebo-controls, and adequate blinding. It is well known that randomised, placebo-controlled trials are the gold standard for determining the efficacy of interventions, because of their ability to account for placebo and other non-specific effects. A systematic review was conducted in Chapter Five to investigate the number and quality of existing randomised placebo-controlled trials of surgical management of musculoskeletal conditions. Whilst the study found that no randomised placebo-controlled trials had been completed on LSS, 18 trials had been completed on other musculoskeletal conditions (e.g., vertebroplasty for vertebral compression fractures). Chapter Five found a rapid increase in the number of placebo-controlled trials of musculoskeletal conditions, from 2 trials before the year 2000 to 20 trials between 2001 and 2020. The study also found that the majority of trials (59%) used a high-fidelity placebo control, which was the degree in which the placebo procedure mimicked all but the ‘active’ component of the surgical procedure being assessed. Minimal fidelity placebo controls were those that had only minimal resemblance to the surgical procedure, such as using a skin incision only, when assessing meniscal repair. The vast use of high-fidelity placebo surgery suggests an increased focus on accounting for non-specific effects and maximising blinding of patients in placebo trials of musculoskeletal surgery. Considering the level of fidelity and other methodological aspects of placebo-controlled trials of surgery can help to improve their feasibility and impact on policy and practice. Following decisions on level of treatment fidelity, another important aspect to trial design in placebo-trials included engaging key stakeholders in the design process, as discussed by Beard et al (2020) in their Lancet paper on consideration and methods of placebo studies. In Chapter Six, LSS patients considered eligible for spine surgery were surveyed to determine their opinions, including barriers to participation in a placebo-controlled surgical trial. The study suggested that a placebo-controlled trial was feasible, with just under 1 in 5 eligible patients with LSS stating that they would consider participation in a placebo-surgical trial of LSS. Following the engagement of LSS patients, spinal surgeons were then consulted in Chapter Seven, to better understand how they viewed the current evidence supporting LSS. Surgeons were also asked what change they expected following spine surgery, including on neurogenic claudication following lumbar decompression. The key findings in Chapter Seven were that spine surgeons expected a mean improvement in neurogenic claudication by 3 months post-surgery of 86% (median: 87%, interquartile range (IQR): 80%, 91%), using a +/- 100% change scale. The majority of surgeons also rated the quality of evidence for lumbar decompression procedures, including decompression alone, decompression with fusion, and interspinous spacer, as low to moderate, supporting the need for future studies. The combined results of Chapter Six and Seven then helped inform the first randomised placebo-controlled trial of decompression for LSS, which has its protocol outlined in Appendix One