Facet-joint injections for non-specific low back pain: a feasibility RCT

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 11/31/02. The contractual start date was in July 2015. The draft report began editorial review in May 2017 and was accepted for publication in September 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Rod S Taylor is the chairperson of the National Institute for Health Research (NIHR) Health Services and Delivery Research programme researcher-led panel and is a current member of the NIHR Priority Research Advisory Methodology Group (PRAMG). He was a member of the NIHR Health Technology Assessment General Board between 2014 and June 2017. Funding from BioQ Pharma has been awarded to the Swansea Centre for Health Economics to conduct a pilot study of the use of OneDose ReadyfusOR in the postoperative delivery of pain relief. Deborah Fitzsimmons is not involved in this study but is the Academic Director of the research centre that has received the grant. Stephanie Poulton reports personal fees from the Neuro Orthopaedic Institute (NOI) and personal fees from Pain and Performance, outside the submitted work. Pain and Performance is a private organisation that teaches and runs seminars about pain for professionals and treats people in pain (see www.painandperformance.com)

Copyright statement

© Queen’s Printer and Controller of HMSO 2017. This work was produced by Snidvongs et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2017 Queen’s Printer and Controller of HMSO

Funding history

The FACET (Feasibility of Assessing the Clinical- and cost-Effectiveness of Therapeutic lumbar facet-joint injections) feasibility study was a commissioned proposal funded by the National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme (reference number HTA 11/31/02); the funding contract was agreed in July 2013. A favourable ethical opinion was given on 21 May 2015 and NHS permission was granted on 20 January 2016, with the first participant recruited to the study on 22 January 2016. Following an extension agreement with the funder, recruitment ended on 30 September 2016 and the study closed on 31 March 2017.

Structure of this report

In this chapter, the existing evidence for therapeutic intra-articular lumbar facet-joint injections (FJIs) for non-specific low back pain (LBP) is reviewed and the need for a pilot trial is presented. In Chapter 2 the feasibility study procedure and associated methodological work are described, with the results of this work being presented in Chapter 3. The implications of our findings are discussed in Chapter 4 and the conclusions drawn with regard to a full trial are presented in Chapter 5.

Background

Low back pain causes more global disability than any other condition, and has a lifetime prevalence of up to 85%. 1 Non-specific LBP, in which symptoms are experienced without any recognisable pathology,2 is thought to affect around 90% of all LBP sufferers, with between 1% and 5% of patients presenting with LBP having a serious spinal pathology such as vertebral fracture, malignancy, infection and inflammatory disease. 3

The economic costs of LBP have been reported to be £12.3B per annum in the UK alone. 4 Chronic LBP, with a prevalence of 3–10%,5 is associated with depression, anxiety, deactivation, inability to work and substantial societal costs. 1,6

Common contributors to LBP in adults are thought to include lumbar facet-joints. 7 These are paired synovial joints between the superior and inferior articular processes of consecutive lumbar vertebrae and between the fifth lumbar vertebra and the sacrum. Encapsulated nerve endings have been demonstrated in these facet-joints, supplied by medial branches of the dorsal rami nerves (‘medial branch nerves’). Facet-joint contributions to LBP may arise from any structure that is part of the facet-joints, including the fibrous capsule, synovial membrane, hyaline cartilage and bone.

Low back pain with a likely facet-joint component can be treated with interventions targeting the facet-joints, including intra-articular (within the joint itself) FJIs, periarticular medial branch nerve blocks or radiofrequency denervation of the medial branch nerves innervating the joints. The technique of FJI is not standardised and may be carried out with or without radiological guidance to confirm needle placement. 8 Lumbar FJIs involve the injection of an active substance, typically steroids with or without a local anaesthetic, intra-articularly or next to the joint (periarticular injections). They are commonly carried out under radiological or fluoroscopic guidance, although they can be performed under ultrasound or computerised tomography scanning guidance. 9

The National Institute for Health and Care Excellence (NICE) guidelines for managing LBP were recently updated [NICE guideline (NG59)10]. These guidelines propose a care pathway in which all those aged ≥ 16 years with LBP with or without sciatica should be provided with advice and information, tailored to their needs and capabilities, to help them self-manage at all steps of the treatment pathway, including education on the nature of LBP and sciatica and encouragement to return to work and pursue normal activities of daily living. 11 Those with a specific episode or flare-up of LBP should consider a group exercise programme, manual treatment or a psychological therapy package. If these therapies fail, pharmacological options and combined physical and psychological (CPP) programmes should be offered, followed by radiofrequency denervation or surgical approaches such as fusion. The guidelines make specific ‘do not use’ recommendations for a range of groups of treatments including acupuncture and electrotherapy; traction, braces and corsets; disc replacement; and spinal injections (including FJIs). The NICE guidelines omit intra-articular FJIs on the grounds of there being insufficient high-quality evidence to support their use and recommend instead targeting the facet-joints’ nerve supply (medial branch nerves) as the predominant pain generator source. However, intra-articular FJIs remain in common use. 12

Review of the literature: a review of systematic reviews and meta-analyses

We undertook a literature search to identify systematic reviews and meta-analyses of randomised controlled trials of intra-articular lumbar FJIs for chronic LBP. We searched MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials from 1966 to February 2017; the search strategies used are detailed in Appendix 1. Two reviewers (Saowarat Snidvongs and Fausto Morell-Ducos) independently screened and assessed full-text articles for eligibility. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram is illustrated in Figure 1. Eleven systematic reviews14–24 met the inclusion criteria; the sources of the reviews are summarised in Appendix 2. The Assessment of Multiple Systematic Reviews (AMSTAR) checklist,25 applied by two independent reviewers to assess the methodological quality of the systematic reviews, demonstrated significant variations between the reviews, with AMSTAR scores ranging between 2 and 10 out of a maximum of 11 points, as shown in Appendix 3 (a low score is associated with a higher risk of bias).

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram for the systematic review search. Adapted from Moher et al. 13 © 2009 Moher et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Adapted from Moher et al. 13 © 2009 Moher et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited

In total, 14 randomised controlled trials were identified across these 11 reviews; 13 were obtained as full-text articles and their findings are summarised in Table 1. Appendix 4 illustrates that, given the variation in dates when the reviews were undertaken and their precise inclusion/exclusion criteria, no one review included all of these trials. The authors of these systematic reviews concluded that the level of clinical heterogeneity across the included randomised controlled trials – different injection procedures, substances and comparators – precluded any meta-analyses.

The conclusions drawn from the systematic reviews were generally equivocal. Although there was some trial evidence demonstrating the effectiveness of therapeutic lumbar FJIs in the management of chronic LBP, the limited to moderate quality of this evidence was insufficient to support their use in practice. The current evidence base for FJI is neatly summed up by a recent and high-quality systematic review by Vekaria et al. ,23 which concluded that ‘The studies found here were clinically diverse and precluded any meta-analysis. A number of methodological issues were identified. The positive results, whilst interpreted with caution, do suggest that there is a need for further high-quality work in this area.’23

Rationale for a feasibility study

To provide further high-quality research in this area, the NIHR HTA programme issued a commissioning brief in 2011 to answer the research question, ‘Is a definitive study to assess the effectiveness and cost-effectiveness of facet-joint injections compared with best non-invasive care for people with persistent non-specific low back pain feasible?’ [see www.journalslibrary.nihr.ac.uk/programmes/hta/113102/#/ (accessed 31 October 2017)].

Two research teams were funded by the NIHR in response to this commissioned call: (1) the Facet Feasibility study (the addition of intra-articular facet-joint injections to best usual non-invasive care) (reference HTA 11/31/01) led by Professor Martin Underwood, University of Warwick [see www.journalslibrary.nihr.ac.uk/programmes/hta/113101/#/ (accessed 31 October 2017)] and (2) this project, the FACET feasibility study (a multicentre double-blind randomised controlled trial comparing intra-articular lumbar facet-joint injections with a sham procedure, followed by a combined physical and psychological programme) (reference HTA 11/31/02) [see www.journalslibrary.nihr.ac.uk/programmes/hta/113102/#/ (accessed 31 October 2017)] led by Professor Richard Langford, Barts Health NHS Trust.

Study aims and research questions

The aim of the FACET feasibility study was to assess the feasibility of conducting a definitive study to evaluate the clinical effectiveness and cost-effectiveness of FJIs compared with a sham procedure in patients with non-specific LBP of > 3 months’ duration.

To inform a full trial, a number of questions first need to be assessed by a feasibility study:

1. Given the multiple sites with the potential to generate back pain, can patient selection criteria be optimised, using clinical and investigative diagnostic methods?

2. Can the method of injection be standardised and an appropriate sham procedure be established?

3. Can justification for further studies to evaluate treatment methods to target and attenuate the source of chronic LBP of facet-joint origin be delivered?

4. Is a sham-controlled trial design acceptable to patients and clinicians?

5. Can a sufficient number of patients be recruited and retained?

Study objectives

• To assess the eligibility criteria and recruitment and retention of patients in the two treatment arms (FJIs vs. sham procedure) by assessing the feasibility of recruitment in the three centres, reviewing the number of completed patient data sets, auditing the quality of data entry at the centres and assessing and analysing any protocol violations (such as failure to deliver the CPP programme – recommended therapy at the time of the establishing this study39), side effects and other adverse outcomes.

• To assess the feasibility and acceptability of the two treatment arms from the point of view of patients and their pain teams.

• To assess the feasibility of the proposed definitive study design, including:

  • testing of the randomisation and blinding procedures

  • development of appropriate active and sham procedures for FJIs

  • assessment of the consistency of the trial sites in terms of delivering the CPP programme

  • assessment of the ability to collect the outcomes proposed for the main trial (pain, functioning, health-related quality of life, anxiety and depression, health-care resource utilisation, complications and adverse events).

• To estimate outcome standard deviations (SDs) to inform the power calculation for a definitive study.

• To finalise the protocol design, statistical plan, number of centres required and study duration for the definitive study.

Study design

This feasibility study utilised a blinded parallel two-arm pilot randomised controlled trial design. Following a positive diagnostic medial branch nerve block, participants with non-specific LBP were individually randomised in a 1 : 1 ratio to receive either the FJI (intervention group) or a sham (placebo injection) procedure (control group). Both the intervention group and the control group received a CPP programme after their active or sham injections.

Participants

The study sought patients with non-specific LBP based on the following inclusion and exclusion criteria.

Recruitment procedures

It was originally planned to recruit patients from pain clinics at the three participating NHS centres and their associated community based pain clinics. However, because of the early termination of the study by the funder, recruitment was undertaken at only one centre, Barts Health NHS Trust. Recruitment took place over 9 months. The first participant was recruited in January 2016 and the last participant was recruited in September 2016.

Patients were referred by their general practitioner (GP) as a standard clinical referral with LBP requiring further specialist assessment, for reasons such as uncertain diagnosis, failure of conservative treatment and expectation of therapeutic interventions. Potentially eligible patients were identified by a pain clinician based on the inclusion and exclusion criteria.

Participants were free to withdraw from the study at any time without giving a reason; the participant information sheet stated that ‘a decision to withdraw from the study at any time will not affect the standard of care that you receive now or in the future’ (see Appendix 5). Participants who withdrew from the study received a routine follow-up appointment in their pain clinic for continuing assessment and management by their pain physician.

Informed consent

Potential participants were given a copy of the information sheet (see Appendix 5) and a verbal explanation of its contents, including information on the nature of the study, the implications and constraints of the study protocol and any known side effects and risks involved in taking part in the study. A medically qualified investigator on the delegation log obtained written informed consent.

Sample size calculation

At the outset of the study it was expected that a total of 60 patients would be recruited, to be able to estimate the precision of an assumed 20% attrition rate with an error of ±5% at the 95% confidence level. Assuming that 24 full data sets per arm were completed at the end of the study, this would give a reasonable estimate of variance of the outcomes. 41

Diagnostic test

Following screening and consent, all participants underwent a diagnostic medial branch nerve block. Those who achieved a positive response were randomised to the intervention group or the control (sham) group.

The diagnostic medial branch nerve injections were carried out at each painful lumbar level under radiological guidance. With the patient lying in the prone position on a radiolucent table, the investigator examined the patient’s back to elicit paraspinal tenderness and confirmed appropriate landmarks and facet-joints to be injected using radiological image intensification. The C-arm of the image intensifier was obliquely rotated as required to facilitate visualisation of the target for injection. The spinal needle was used to inject 0.5 ml of 1% lidocaine per level, with six levels injected. A positive response was defined as a ≥ 50% pain reduction lasting for > 30 minutes, that is, the duration of action of lidocaine, measured using a pain intensity numerical rating scale (NRS) and assessed in the standing position.

Interventions

The FJIs, sham procedure and diagnostic tests were carried out by the Principal Investigator at Barts Health NHS Trust, a Fellow of the Faculty of Pain Medicine of the Royal College of Anaesthetists in the UK. During all of the interventional procedures strict aseptic conditions were adhered to and local theatre protocols were followed with regard to admission and discharge criteria, including the use of the World Health Organization (WHO) Surgical Safety Checklist42 to identify the correct patient prior to starting the procedure. The investigator carrying out the injections was not blinded to randomisation group.

Regulatory approvals

The study was conducted in compliance with the principles of the Declaration of Helsinki (1996)45 and the principles of Good Clinical Practice46 and in accord with all applicable regulatory requirements including but not limited to the Research Governance Framework47 and the Medicines for Human Use (Clinical Trials) Regulations 2004,48 as amended in 2006 and 2008, the sponsor’s policies and procedures and any subsequent amendments.

The required regulatory approvals were obtained in the UK. The study received ethics approval from the National Research Ethics Service (NRES) Committee London – City & East (reference 15/LO/0500) and clinical trial authorisation from the Medicines and Healthcare products Regulatory Agency (MHRA) (reference 14620/0046/001–0001). The trial protocol was reviewed by the MHRA’s clinical trials team and the trial was considered to be a type A Clinical Trial of an Investigational Medicinal Product (CTIMP), that is, the risks are no higher than those of standard medical care. The Summary of Product Characteristics (SmPCs) for each investigational medical product (bupivacaine and methylprednisolone acetate) are available to view on the Electronic Medicines Compendium. 49,50 Health Research Authority (HRA) approval was obtained and the study was given permission by the sponsor’s Joint Research Management Office (JRMO) to recruit patients at Barts Health NHS Trust.

Imaging authorisation was given by the sponsor’s clinical radiation expert and medical physics expert, as all of the participants would receive ionising radiation in the form of X-rays for the diagnostic injections, FJIs and sham procedure.

Randomisation and blinding

Participants were allocated to either the intervention group or the control group in a 1 : 1 ratio, with stratification by centre and minimisation on baseline pain scores (categories). To ensure concealment, the allocation sequence was computer generated and provided through a password-protected web-based portal developed and maintained by the Peninsula Clinical Trials Unit (PenCTU).

It was not possible to blind the operator (Principal Investigator) as the injections were intentionally given at different sites (intra-articular vs. periarticular) and the injections looked different (methylprednisolone is provided as a cloudy suspension, whereas the sham injection was clear). However, study participants and the remainder of the research team, including the Chief Investigator, research nurses conducting the outcome assessments and data analysts, were blinded for the duration of the study. Unblinding took place at the end of the study once data analysis had been completed. A standard operating procedure was in place for emergency unblinding, in accordance with the sponsor guidelines.

Outcomes

The outcome questionnaire visits took place in research nurse-led clinics at baseline (pre randomisation) and at 6 weeks, 3 months and 6 months post randomisation. A sample case report form is shown in Appendix 7 and the schedule of outcome assessment is shown in Table 2. The outcome questionnaire covered a range of pain- and disability-related issues and was in accord with the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT)-recommended core outcome measures for chronic pain trials. 51 The following assessment tools were used in the study.

• Pain intensity and characteristics – Brief Pain Inventory (BPI) (Short Form) Modified,52 with its 11-point NRS, and short-form McGill Pain Questionnaire version 2 (SF-MPQ-2). 53 As movement could potentially influence the intervention (lumbar FJIs or sham procedure), all numerical rating scores were assessed in the standing position.

• Use of co-analgesics in the previous week – participant self-report.

• Lack of efficacy of pain relief, or, for side effects, early withdrawal from the study.

• Expectation of benefit (asked at baseline only) – measured on a scale from 0 to 6, ranging from ‘expect no improvement’ to ‘expect total improvement’.

• Health-related quality of life – EuroQol-5 Dimensions five-level version (EQ-5D-5L)54 and Short Form questionnaire-12 items (SF-12). 55

• Functional impairment – Oswestry Low Back Pain Disability Questionnaire56 and Pain Self-Efficacy Questionnaire (PSEQ). 57

• Satisfaction with treatment (after treatment given) – NRS from 0 to 10, with 0 = ‘extremely dissatisfied’ and 10 = ‘extremely satisfied’.

• Complications and adverse events – these were the subject of enquiry at visits and following procedures, as well as being spontaneously reported at any time. They were acted on as necessary and for the patients’ benefit and were fully documented on case report forms and medical notes.

• Co-psychological well-being – Hospital Anxiety and Depression Scale (HADS),58 Pain Catastrophizing Scale (PCS),59 SF-12 and BPI.

• Health-care utilisation and costs and impact on productivity –Stanford Presenteeism Scale (SPS) 6,60 self-reported measures of sickness absence over the previous 3 months and health-care utilisation in the form of hospital visits, treatments and medications. These data were collected at each outcome visit on the case report form.

Adverse events

Adverse events were assessed by a blinded subinvestigator at each visit and an adverse event form was completed as necessary (see Appendix 8). Adverse events are defined as any untoward medical occurrence in a subject to whom a medicinal product has been administered; an adverse reaction is an untoward and unintended response in a subject to an investigational medicinal product (IMP) that is related to any dose administered to that subject. A serious adverse event or reaction results in death, is life-threatening, requires inpatient hospitalisation or prolongation of an existing hospitalisation, results in persistent or significant disability or incapacity or is a congenital anomaly or birth defect. A suspected unexpected serious adverse reaction is any serious adverse event that is both suspected to be related to the IMP and unexpected.

Adverse events not already identified locally were recorded at each trial visit and managed in accordance with the sponsor’s requirements. Serious adverse events were reported to the JRMO by the investigators within 24 hours of the research team becoming aware of them and causality and expectedness were confirmed by the Chief Investigator, as the sponsor’s medical representative.

Study management and committees

The Trial Management Group (TMG) was responsible for the overall management of the project and included all co-applicants and members of the study research team. A Trial Steering Committee (TSC) provided independent advice and support to the study and aimed to report to the funder on study progress. It was chaired by an independent clinician with experience of pain trials. A Data Monitoring Committee (DMC) had access to unblinded data and made recommendations to the TSC on whether there were any ethical or safety reasons why the trial should not continue. It included independent members who were all experts in pain medicine.

Patient and public involvement

Patients with personal experience of LBP collaborated in the early stages of study design, for example advising on the acceptability of study visits and the outcome questionnaires. The questionnaires were tested on patients presenting to the multidisciplinary pain clinics and were deemed to be acceptable. Patient representatives were invited to attend the TSC meetings; however, there was no patient or public involvement in the management or running of the trial beyond the initial set-up stage.

Statistical analysis

Analyses were conducted in accordance with the International Conference on Harmonisation (ICH) statistical guidelines for clinical trials61 and the Consolidated Standards for Reporting Trials (CONSORT) reporting checklist for trials. 62 As this was a feasibility study, it was not planned to formally inferentially test differences in outcomes or costs between or within the groups. Mean recruitment and attrition rates were calculated with 95% confidence intervals (CIs). Means and SDs for all outcomes for the two groups at baseline and at the follow-up visits were reported. A detailed statistical analysis plan was prepared by the study statistician (RST) prior to any data analysis. Analyses were performed blinded to group allocation using Stata^® 14.2 (StataCorp LP, College Station, TX, USA).

Health economics analysis

A health economics analysis plan was developed in collaboration with the study’s health economist and statistician. A formal economic analysis was not proposed (as this was a feasibility study). Any outcomes from this feasibility study would be used in the design of the definitive study. In particular, the health economics analysis looked at the ability to collect the outcomes proposed for the main trial and to inform a robust framework to assess the cost-effectiveness of lumbar FJIs for persistent non-specific LBP for a future definitive randomised controlled trial.

The resources used in the delivery of the intervention and sham procedure were calculated from the case report forms and in consultation with the trial team. Health-care resource use was captured through administration of specific questions to trial participants at each assessment, with responses recorded on case report forms, with a focus on collecting data on the most relevant and important drivers of health resource use and costs. Published national costs were used to calculate the costs of delivering each treatment arm. A summary of the costing methods employed are presented in Appendix 9. In addition, a literature review was conducted to review the scope and quality of the current economic evidence base for the use of FJIs in patients with non-specific LBP (see Appendix 9).

Descriptive analyses of the outcomes were used to report utilities based on the relevant tariff for each of the health-related quality-of-life outcomes. Economic outcomes were measured using the EQ-5D-5L (the preferred approach of the study team and for NICE decision-making) and the Short Form questionnaire-6 Dimensions (SF-6D63), used as a means of calculating utilities from the SF-12. A descriptive analysis of the health-related quality of life outcomes was summarised and was to provide estimates of quality-adjusted life-years (QALYs) gained as a result of receiving the intervention within the study period.

Summary of changes to the study protocol

The minor and major amendments made to the protocol over the duration of the study are detailed in Appendix 10.

Screening and recruitment

It was originally planned to screen and recruit participants across three centres; however, given the delays in study set-up, the funder directed that screening and recruitment take place at one centre only. The timelines of the study and the reasons for the delays are presented at the end of this chapter.

Recruitment began at Barts Health NHS Trust on 20 January 2016 and was terminated on 30 September 2016. As shown in Figure 2, during this time 628 patients referred to the recruiting clinics by their GP with non-specific LBP were screened for eligibility to enter the study. Of the 50 patients who met the inclusion criteria, 16 agreed to take part in the study and 11 received the diagnostic test for facet-joint disease. Nine participants had a positive response and were randomised to receive either lumbar FJIs with steroid or a sham procedure. The target sample size in each centre was 20 participants. The actual participant screening-to-recruitment ratio was 70 : 1 (628 : 9), which contrasts with an expected prestudy ratio of 17 : 1 (1000 : 60). The recruitment rate varied between zero and four patients per month (median two participants per month (Table 3).

FIGURE 2.

Consolidated Standards for Reporting Trials (CONSORT) flow diagram showing the flow of participants through the trial.

The reasons for screening failure are detailed in Table 4. The greatest proportion of patients was screened from the pain clinics at St Bartholomew’s Hospital, with a screened/recruited fraction of 1.69% (Table 5). Although only 16 patients were screened from the community pain clinic at the Essex Lodge GP surgery in Plaistow, East London, the screened/recruited fraction was higher at 12.5%. No participants were recruited from the spinal orthopaedic clinic at The Royal London Hospital or from the pain clinics at Whipps Cross University Hospital and Mile End Hospital.

Nine out of the 11 enrolled participants had a positive response to the diagnostic lumbar facet medial branch nerve block (82%, 95% CI 48% to 98%).

Adherence to allocated treatment

Study dropout and attrition

Of the 16 patients recruited and who consented to take part in the study, five withdrew before they received their diagnostic injections (although all completed the baseline assessment questionnaires). One patient already had lumbar FJIs booked for a future date and two changed their mind about taking part (one for personal reasons and one for family reasons). Two were unable to attend for the injections as the study dates were not suitable for them.

Of the nine participants who were randomised, eight completed the study (defined as having completed the final set of questionnaires 6 months after the randomised procedure), resulting in an 11% (95% CI 0.2% to 48%) attrition rate. The expected attrition rate was 20%. Six study deviations were recorded: three participants did not complete the CPP programme and three participants did not complete all of the questionnaire sets (Figure 3). A ‘study deviation’ was defined as a participant who did not attend a study visit or CPP programme session but who did not drop out of the study completely.

FIGURE 3.

Flow diagram showing the study deviations and dropouts during the study. (a) Combined physical and psychological programme and facet-joint injections; and (b) combined physical and psychological programme and sham procedure.

There were low levels of missingness of within-questionnaire data, as shown in Appendix 11. This is summarised in Table 6.

Baseline characteristics and outcomes

The mean age of eligible participants was 45 years, with a similar proportion of males and females. Six out of 14 participants (43%) were not working at baseline (Table 7).

Baseline patient-reported outcomes indicated a population with substantial levels of pain [mean 8.5 on a 0–10 visual analogue scale (VAS)] that was predominantly bilateral (12/16, 75%) and which had a mean duration of 72 months (see Table 7). In terms of the primary and secondary outcomes, high baseline levels of disability and mental ill health and poor overall health-related quality of life were seen (Tables 7 and 8).

Given the small number of participants randomised, not unexpectedly there was evidence of imbalance in participant characteristics and patient-reported outcomes between the two groups at baseline (see Table 8).

Primary and secondary outcomes at follow-up

Given the feasibility nature of the study and the small number of participants randomised, the primary and secondary outcomes at 6 weeks’ and 3 and 6 months’ follow-up are presented descriptively, with no inferential between or within comparisons undertaken or reported (Table 9).

Adverse events

Three study participants reported adverse events, with two serious events reported by one participant. All three participants reported a flare-up of their LBP, which resolved (Table 10).

Blinding to treatment allocation

To test the fidelity of blinding, participants were asked to guess which allocation group they had been randomised to, prior to being unblinded at the end of the study. Only one out of eight participants who completed the study correctly guessed their allocation group. The blinded outcome assessor correctly guessed the allocation group for four of the nine participants.

Health economics analysis

Given the importance of health economics evidence in informing decision-making, the incorporation of a health economics work package as part of a feasibility study can inform the delivery of a robust trial in the future, evaluating clinical effectiveness and cost-effectiveness. The main objective of the health economics analysis was to assess the feasibility of collecting data and produce an appropriate framework for a future full economic evaluation. This included a description of the main resource and cost drivers during the patient pathway in delivering the intervention and an assessment of collecting resource use information based on a resource use questionnaire devised for the feasibility study. An assessment of the performance of two preference-based health-related quality of life measures – the SF-12 (SF-6D)55 and EQ-5D-5L54 – within the feasibility study was also proposed in the original health economics analysis plan.

We intended to present both of the scores from the full EQ-5D (EuroQol-5 Dimensions) questionnaire, that is, the EQ-5D-5L descriptive system and the EQ-5D VAS, to present a comprehensive picture of patient-reported outcomes. Patients are asked to assess their health state on five dimensions – mobility, self-care, usual activities, pain/discomfort and anxiety/depression – using five levels ranging from ‘no problems’ to ‘severe problems’. A single summary index is obtained by applying a formula that attaches weights to each of the levels in each dimension. This formula is based on the valuation of EQ-5D health states from a general population sample in the UK. 64 The EQ-5D VAS is a thermometer-type vertical 20-cm scale with the end points of best imaginable health state (at the top) and worst imaginable health state (at the bottom) having numerical values of 100 and 0 respectively. This is a self-rated valuation and represents the respondents’ views of their health state and how it has affected their life on the day.

The utilities derived from the EQ-5D-5L and SF-6D were converted into QALYs to present a profile of QALY gains/losses over time.

As this was not a full trial, no inferential statistical comparisons of health-related quality of life or service use between the intervention group and the control group were planned. Instead, the focus was on reporting descriptive findings for the two groups at baseline and at the follow-up points to inform a suitable framework for a future economic evaluation if the findings from the feasibility study could be used to inform a future definitive study.

For the feasibility study, the perspective adopted was that of the UK NHS, focusing on primary and secondary resource use. However, given the impact of back pain on the patient and wider society (e.g. because of reduced/lost work productivity), a preliminary assessment of employment-related outcomes using the SPS 660 was reported as part of the trial outcomes. The economic analysis was conducted using IBM SPSS Statistics version 22 (IBM Corporation, Armonk, NY, USA).

Study timelines

The start date of the study was delayed by 18 months (546 days), with 244 days taken to obtain NHS permission to recruit after obtaining a favourable research ethics opinion (Figures 4 and 5). A no-cost negotiated 1-month extension period was granted by the funder to allow for collection of follow-up data and delivery of the final report 45 days after collection of the final outcome data.

FIGURE 4.

Study timelines and milestones. REC, Research Ethics Committee.

FIGURE 5.

Study Gantt chart.

Summary of findings

The overarching aim of the FACET feasibility study was to assess the feasibility of conducting a definitive study to evaluate the clinical effectiveness and cost-effectiveness of FJIs compared with a sham procedure in patients with non-specific LBP of > 3 months’ duration.

We were unable to meet our feasibility objective of a recruitment target of 60 randomised LBP patients across three investigative sites. Instead, we were able to randomise only nine patients in one investigative site. However, we believe that we met our other pre-stated feasibility objectives within the constraints of this being a single-centre study with a small sample size: (1) we were able to select patients into the study using a joint clinical and diagnostic approach; (2) we have demonstrated successful standardisation of the methods of FJI and the sham procedure, using the Delphi method to generate consensus among interventional pain specialists in the UK; (3) we were able to carry out the sham procedure (periarticular injections with saline) in participants randomised to the control group; (4) the sham control study design was accepted by the clinicians involved in the study and was also deemed acceptable by patients; and (5) we were able to retain patients over the 6 months of follow-up.

The small number of participants recruited to the study and the feasibility design preclude us from drawing any conclusions on the clinical effectiveness or cost-effectiveness of intra-articular lumbar FJIs in the management of non-specific LBP. Additionally, although eight out of the nine patients randomised completed the study, the small sample size precludes any definitive conclusions being made about patient retention. However, the clinical procedures appeared to be safe and well tolerated, with no significant adverse events related to the steroid injection. Furthermore, we were able to successfully collect clinical and economic outcomes from the majority of the patients over the duration of the study.

Numerical differences between the groups were identified in the feasibility study (particularly with regard to increased medication costs in the intervention group) that warrant further attention, particularly given the inherent skewness in the data and the very small sample size.

The remainder of this chapter considers in more detail the findings of the feasibility study in terms of the detailed objectives, reviews the practical problems that were experienced and identifies the key learning outcomes for any definitive future trial in this area.

Issues encountered and how they were resolved

This feasibility study encountered system-level barriers but no significant delays in obtaining research ethics or MHRA approval.

Strengths and limitations

Many of the key uncertainties involved in undertaking a definitive study were addressed in this feasibility study.

Major strengths of the study were the Delphi exercise, which was undertaken to obtain a consensus on the procedural techniques, and that the interventional procedures themselves were shown to be safe and reproducible. Furthermore, the positive response in nine out of 11 participants (82%, 95% CI 48% to 98%) following the single diagnostic lumbar facet medial branch nerve block indicates some accuracy of the assessment performed by the clinical team in determining pain of lumbar facet-joint origin based on history and clinical examination alone. However, the results are imprecise because of the small sample size. We have shown that the CPP programme is deliverable at a single recruiting centre and that it was possible to train the study physiotherapists to deliver the programme at each of the three planned recruitment centres. We were able to retain most of the recruited participants for the duration of the study.

Embedding a health economics assessment into the feasibility study allowed us to obtain important insights that can be used to develop the framework for an economic analysis within a future trial. Although the feasibility study was not primarily designed to act as a formal microcosting exercise, we have endeavoured to capture all salient aspects of the patient pathway relevant to the intervention using a triangulation of data sources, including trial documentation and structured questions to the clinical experts and discussion with the trial team, with validation of any assumptions made throughout the costing procedure. We have also been transparent in the costing methods employed, as detailed in Appendix 9. A recognised limitation of our analysis is that the resource use questionnaire was not sufficiently comprehensive to capture all health-care resource use nor did we investigate the potential for other approaches (including capturing routine data in a systematic manner). In future feasibility studies we suggest that a worthwhile endeavour would be to undertake a more ‘bottom-up’ exercise of costing, subject to trial resources being available to do so. We therefore recommend that, in a future trial, the current measure is tested and challenged against existing measures [e.g. search of the DIRUM (Database of Instruments for Resource Use Measurement);82 see Table 29 in Appendix 9], including investigating ways that data can be collated [such as using routine sources such as Hospital Episode Statistics (HES) and/or utilising patient diaries to collect data, e.g. related to primary care contacts and medication use].

The main weakness of the study was the failure to achieve our expected recruitment targets and the consequent early closure by the funder. There were substantial system-level barriers that the study team were unable to control (such as delays in obtaining HRA approval and changes in key study personnel), which led to long delays in obtaining research governance approvals, and permission to recruit was granted at only a single site.

It became clear early in the recruitment phase that many patients presenting to the pain clinics with LBP of > 3 months’ duration were not suitable for the study and that patients screened in the spinal orthopaedic clinics also did not meet the eligibility criteria. Despite employing additional strategies in the final stages of recruitment, we can conclude that the patient population presenting to these hospital-based specialist clinics was generally not suitable because of the complexity of their pain problems and that we may have had better success in recruiting from primary care-based services.

We would advise caution in extrapolating the results of this study as the study took place at a single centre in a secondary care setting. Patients presenting to primary care may additionally have lower levels of functional disability, despite high levels of pain intensity.

The feasibility study resulted in a number of suggestions for refinements to data collection, including resource use. One of the key balances has to be the development of a measure (method) of capturing resource use that is sufficiently comprehensive and accurate enough to capture important and relevant changes while not overburdening the trial in capturing extensive and irrelevant drivers of costs. The derived measure was driven primarily by clinical expertise; however, this must of course be balanced with the needs of the health economist in terms of having data of sufficient quality to undertake a comprehensive and transparent costing exercise. Appendix 14 presents recommendations from reflecting on the current measure of resource capture.

Because of the trial limitations, we were able to present only a descriptive profile using the EQ-5D-5L and SF-6D to determine utilities; thus, any interpretation of our findings should be appropriately tentative. The choice of which measure to use should also be considered further, given the potential for different numbers of QALYs to be generated by different measures and the subsequent impact on calculations of cost-effectiveness. 73

Implications for a future definitive study

Neither of the research teams funded by the HTA programme to investigate the feasibility of carrying out a definitive study to assess the use of lumbar FJIs for persistent non-specific LBP met the target recruitment rate. Professor Underwood’s team concluded that a definitive study is indeed feasible, but that recruitment from the pain clinics alone was insufficient. 83 Both teams experienced significant delays in study set-up.

Any future definitive trial must consider patient and public involvement at all stages of the research cycle, including in the design and running of the trial and representation on the TMG.

Low back pain and sciatica were selected by the Trauma Programme of Care Board as their Pathfinder Project; these projects were established by NHS England in 2013 to set up clear ‘end-to-end’ generic pathways from primary care into specialised services as required, enabling collaborative commissioning and incorporation of the latest evidence into the pathways. The National Low Back and Radicular Pain Pathway 201784 was updated following the publication of NG5910 and includes all of NICE’s recommendations. Improved co-ordination and communication with GP practices is necessary for any future collaborative work and we envisage working alongside the Pathfinder Project (in which spinal injections are not recommended) to recruit patients from primary care with high levels of functional disability who may benefit from such an intervention and rehabilitation at an earlier stage. One finding from the feasibility study was the long duration of pain awareness among the study participants, which warrants further investigation as this may reflect the time taken for patients to be referred to specialist clinics.

Based on our findings, we propose three research recommendations. First, we would agree with the Underwood team that a definitive study is potentially feasible, with adjustments made to the target population and increased primary care involvement, to enable patients to be screened earlier in their pain trajectory. Second, a definitive study will need to draw on lessons learned from both research teams (such as the recruitment enhancement strategies) and could involve a future collaboration between the two research groups, combining the successful elements of the feasibility studies. The updated NICE guidelines10 recommended radiofrequency denervation of the lumbar facet-joint medial branch nerves in patients with LBP that have not responded to less invasive treatments, following a positive diagnostic block; this is in contrast to a recent Cochrane review that has not recommended its use. 85 There is, however, new evidence from a large multicentre Dutch trial that radiofrequency denervation offers no clinical benefit compared with a sham procedure. 86 Third, as both intra-articular injections and radiofrequency denervation are widely used in the UK, we contend that there remains a need for future definitive studies to assess the clinical effectiveness and cost-effectiveness of both of these procedures in the management of persistent LBP.

Original co-applicant team

Professor Richard Langford (former Chief Investigator).

Dr Vivek Mehta (Chief Investigator, former Principal Investigator at Barts Health NHS Trust).

Professor Rod S Taylor (Study Statistician).

Dr Simon Thomson (Principal Investigator at Basildon and Thurrock University Hospitals NHS Foundation Trust).

Dr Manohar Sharma (Principal Investigator at the Walton Centre NHS Foundation Trust).

Professor Ceri Phillips (former Study Health Economist).

Professor Andrew Moore (Study Methodologist).

Professor Paul Watson (former Lead Physiotherapist).

Dr Amanda Williams (Study Health Psychologist).

Mr Neil Buxton (Consultant Neurosurgeon at the Walton Centre NHS Foundation Trust).

Dr Stephanie Stokes (Patient Liaison Representative).

Trial Management Group

Dr Vivek Mehta (Chief Investigator).

Professor Rod S Taylor (Study Statistician and Trial Mentor).

Dr Simon Thomson (Principal Investigator at Basildon and Thurrock University Hospitals NHS Foundation Trust).

Dr Manohar Sharma (Principal Investigator at the Walton Centre NHS Foundation Trust).

Professor Deborah Fitzsimmons (Study Health Economist).

Dr Saowarat Snidvongs (Principal Investigator at Barts Health NHS Trust).

Ms Alia Ahmad (Trial Co-ordinator).

Trial Steering Committee

Dr Sanjeeva Gupta (Bradford Teaching Hospitals NHS Foundation Trust) (chairperson).

Dr Julius Bourke (Queen Mary University of London).

Dr Theresa Wodehouse (Queen Mary University of London).

Dr Thomas Gilkes (Homerton University Hospital NHS Foundation Trust).

Dr Vivek Mehta (Barts Health NHS Trust).

Data Monitoring Committee

Dr Ajoy Pandit (West Hertfordshire Hospitals NHS Trust) (chairperson).

Dr David Pang (Guy’s and St Thomas’ NHS Foundation Trust).

Dr Michael Husband (West Sussex Hospitals NHS Foundation Trust).

Trial management team (Pain and Anaesthesia Research Centre at Barts Health NHS Trust)

Ms Alia Ahmad (Trial Co-ordinator).

Ms Athina Karavasopoulou (Research Nurse).

Ms Georgina Wilson (Research Assistant).

Pain consultants at Barts Health NHS Trust

Dr Jayne Gallagher (Lead Clinician).

Dr Vivek Mehta.

Dr Serge Nikolic.

Dr Kristin Ullrich.

Dr Shankar Ramaswamy.

Dr Saowarat Snidvongs.

Spinal orthopaedic research team at Barts Health NHS Trust

Mr Alexander Montgomery (Consultant Spinal Surgeon).

Ms Jamila Kassam (Lead Research Physiotherapist).

Ms Catherine Hilton (Orthopaedic Research Physiotherapist).

Ms Shahnaz Ahmad (Orthopaedic Research Physiotherapist).

Study physiotherapists

Ms Stephanie Poulton (Lead Study Physiotherapist).

Ms Kasia Gabriel.

Ms Roisin Ormond.

Peninsula Clinical Trials Unit

Laura Cocking.

Brian Wainman.

Elliot Carter.

Dr Hardip Nandra (GP at Essex Lodge GP surgery, London).

Dr Fausto Morell-Ducos (Specialty Registrar in Anaesthesia and Pain Medicine) – second reviewer for the systematic reviews.

Ms Isatou N’jie (Clinical Support Librarian at Barts Health NHS Trust) – provided assistance in developing search terms for the literature searches.

Contributions of authors

Saowarat Snidvongs (Consultant in Anaesthesia and Pain Medicine) was the Principal Investigator at Barts Health NHS Trust, drafted the manuscript and carried out the diagnostic injections, facet-joint injections and sham procedure.

Rod S Taylor (Professor of Health Services Research, Director of the Exeter Clinical Trials Unit and NIHR Senior Investigator) was the study statistician and a co-applicant, contributed to the protocol design and development, drafted the statistical analysis plan, undertook statistical analyses for the study and contributed to writing of the manuscript.

Alia Ahmad (Research Assistant) was the trial co-ordinator and was responsible for facilitating and co-ordinating the study by maintaining trial activities and paperwork according to good clinical practice.

Simon Thomson (Consultant in Anaesthesia and Pain Medicine) was the Principal Investigator at Basildon and Thurrock University Hospitals NHS Foundation Trust and a co-applicant, contributed to the protocol design and development and reviewed the final manuscript.

Manohar Sharma (Consultant in Pain Medicine) was the Principal Investigator at the Walton Centre NHS Foundation Trust and a co-applicant, contributed to the protocol design and development and reviewed the final manuscript.

Angela Farr (Health Economist and Data Analyst) contributed to the writing of the health economics analysis plan, undertook the health economics analysis, drafted the health economics results and detailed costing methodology and provided feedback on all aspects of the health economics sections and the final draft of the manuscript.

Deborah Fitzsimmons (Professor of Health Outcomes Research and Academic Director of the Swansea Centre for Health Economics) was the lead study health economist, wrote the health economics analysis plan, had senior oversight of the health economics analysis, drafted the health economics sections of the manuscript and provided feedback on the final draft of the manuscript.

Stephanie Poulton (Clinical Specialist Physiotherapist in Pain Management) was the lead study physiotherapist co-ordinating the training of the study physiotherapists for delivery of the CPP programme.

Vivek Mehta (Consultant in Pain Medicine and Director of the Pain and Anaesthesia Research Centre at Barts Health NHS Trust) was the Chief Investigator and former Principal Investigator and was a grant holder and co-applicant, contributed to the protocol design and development, carried out the final outcome assessments and reviewed the final manuscript.

Richard Langford (Consultant in Anaesthesia and Pain Medicine and former Director of the Pain and Anaesthesia Research Centre at Barts Health NHS Trust) was the former Chief Investigator and was a co-applicant, contributed to the protocol design and development and reviewed the final manuscript.

Conferences

The study protocol was presented as a poster at the International Association for the Study of Pain’s 16th World Congress on Pain in Yokohama, Japan, in September 2016. The preliminary results were presented as a poster at the British Pain Society’s 50th Anniversary Annual Scientific Meeting in Birmingham, UK, in May 2017. The full study was presented as a poster at the European Pain Federation’s 10th Congress in Copenhagen, Denmark, in September 2017; the critical appraisal of systematic reviews was also presented as a poster at the same congress. The study has been accepted as a poster presentation at the Society for Back Pain Research in Northampton, UK, in November 2017.

Thesis in preparation

The FACET feasibility study will form the basis of Dr Snidvongs’s Doctor of Medicine thesis entitled ‘Lumbar facet-joint injections for the management of chronic low back pain’. This will be submitted to the University of Exeter in 2017.

Data sharing statement

We shall make data available to the scientific community with as few restrictions as feasible while retaining exclusive use until the publication of major outputs. Requests for the data should be made to the corresponding author (Dr Saowarat Snidvongs).

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health.

MEDLINE

Date range searched: 1966 to 6 February 2017.

EMBASE

Date range searched: 1966 to 6 February 2017.

Cochrane Central Register of Controlled Trials

Date range searched: 1966 to 6 February 2017.

Literature review

Resource use associated with facet-joint injections

Calculating quality-adjusted life-years

For both groups in the study the number of QALYs gained was calculated for the complete cases for the 6 months between baseline (before treatment) and the final 6-month follow-up visit, as shown in Table 28.

Existing measures used to collect resource use data in lower back pain studies

A search of the DIRUM82 to compare existing resource use measures specifically related to LBP was conducted. The results of this search are shown in Table 29.