Trial Review

The ANZCTR website will be unavailable from 1pm until 3pm (AEDT) on Wednesday the 30th of October for website maintenance. Please be sure to log out of the system in order to avoid any loss of data.

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers
Trial registered on ANZCTR


Registration number
ACTRN12616000888460
Ethics application status
Approved
Date submitted
24/06/2016
Date registered
6/07/2016
Date last updated
21/09/2021
Date data sharing statement initially provided
7/05/2019
Date results provided
21/09/2021
Type of registration
Prospectively registered

Titles & IDs
Public title
Can a new approach to managing low back pain in general practice improve patient recovery?
Scientific title
The Fear Reduction Exercised Early (FREE) approach to low back pain: a cluster randomised, parallel-group, superiority trial of patient impairment 6-months post general practice consultation
Secondary ID [1] 289428 0
Nil known
Universal Trial Number (UTN)
U1111-1178-7448
Trial acronym
LBPinGP
Linked study record

Health condition
Health condition(s) or problem(s) studied:
Low back pain 299107 0
Condition category
Condition code
Musculoskeletal 299133 299133 0 0
Other muscular and skeletal disorders

Intervention/exposure
Study type
Interventional
Description of intervention(s) / exposure
The Fear Reduction Exercised Early (FREE) approach to low back pain (LBP). The FREE approach aims to assist GPs to identify and address attitudes and beliefs which may be barriers to recovery from LBP, explain back pain to their patients, and to provide patients with a framework which enables them to resume normal employment and activities of daily living.

General practitioner (GP) participants will be instructed in the FREE approach through an initial four-hour workshop facilitated by an experienced physiotherapist. The workshop will be supported by a booklet, a website, and a module within the practice's electronic patient management system. GPs will then practise using the approach for approximately one-month with any patients who present with low back pain before attending a one-hour refresher workshop facilitated by a physiotherapist. The refresher workshop will discuss experiences of using the approach and address any barriers identified . GPs will also be provided with a novel information booklet to provide to their patients. This booklet outlines current best practice for acute LBP, supports recommendations from the GP booklet, and addresses commonly held beliefs. It encourages an active approach to managing LBP. The booklet will also be supported by a website. Patients and GPs will have access to the website throughout the trial and on an ongoing basis following the trial. GPs will be able to continue using the FREE approach following the six to twelve-week patient recruitment period in each practice.

Patients with a primary complaint of low back pain who consult GPs who have completed FREE training will receive care based on the FREE approach. Care will be provided within general practices. GPs will be able to personalise the care they provide to any individual patient based on their own clinical reasoning. All intervention group patients will receive one consultation based on the FREE approach; some patients may return to their GP for additional consultations related to their back pain and these will also be based on the FREE approach. The FREE approach will be delivered within a standard fifteen minute GP consultation. Fidelity to the approach will be assessed through i) consultation note audit; ii) patient report and; iii) analysis of audio-recordings of a random sample of 15% of all consultations, Additional care received by patients will not be controlled but it will be assessed through patient report.
Intervention code [1] 295017 0
Behaviour
Intervention code [2] 295018 0
Treatment: Other
Comparator / control treatment
Usual care for low back pain received from a general practitioner in a general practice clinic. This care will be delivered within a standard fifteen minute GP consultation. The content of this care will be at the discretion of the general practitioner and will not be controlled but it will be assessed through i) consultation note audit; ii) patient report and; iii) analysis of audio-recordings of a random sample of 15% of all consultations, Additional care received by patients will not be controlled but it will be assessed through patient report.

At the end of the trial after all follow-up data has been collected from patients recruited in a given practice, GPs from this practice will be offered an opportunity to attend the FREE workshop. This will provide comparable benefits for being part of a research trial.
Control group
Active

Outcomes
Primary outcome [1] 298606 0
Patient back pain related impairment measured with the Roland Morris Disability Questionnaire
Timepoint [1] 298606 0
6-months post initial GP consultation (also measured at 2 weeks, 6 weeks, 3 months, 2 years)
Secondary outcome [1] 324704 0
Back pain intensity measured with the numeric pain rating scale
Timepoint [1] 324704 0
2 weeks, 6 weeks, 3 months, 6 months, 2 years post initial GP consultation
Secondary outcome [2] 324705 0
Activity limitation measured with an eleven point numeric rating scale
Timepoint [2] 324705 0
2 weeks, 6 weeks, 3 months, 6 months post initial GP consultation
Secondary outcome [3] 324706 0
Health related quality of life measure with the EuroQoL-5D
Timepoint [3] 324706 0
3 months, 6 months post initial GP consultation
Secondary outcome [4] 324707 0
Satisfaction with information received, care received, and the outcome of care measured with a seven point likert scale
Timepoint [4] 324707 0
Immediate post-consultation, 2 weeks, 6 weeks, 6 months post initial GP consultation
Secondary outcome [5] 324708 0
Economic consequences of back pain measured with a modified Otago Costs and Consequences Questionnaire
Timepoint [5] 324708 0
2 weeks, 3 months, 6 months, 2 years post initial GP consultation
Secondary outcome [6] 324709 0
GP beliefs about pain and impairment measured with the Health Care Providers Pain and Impairment Relationship Scale
Timepoint [6] 324709 0
4 weeks, 4 months post initial GP training workshop
Secondary outcome [7] 324710 0
GP confidence to manage LBP measured with the Provider Self Confidence Tool
Timepoint [7] 324710 0
4 weeks, 4 months post initial GP training workshop
Secondary outcome [8] 324711 0
GP reported clinical behaviour related to a patient vignette measured using responses to three items rated on a five-point likert scale
Timepoint [8] 324711 0
4 weeks, 4 months post initial GP training workshop
Secondary outcome [9] 324712 0
Objective assessment of GP behaviour: GP consultation content recorded from Electronic Medical Record consultation notes and patient report of GP: i) advice to work normally/stay at work on reduced hours or restricted duties/take time off work; ii) advice to continue with normal activity/limit or reduce normal activity/avoid activity; iii) advice to take medication; iv) referral or recommendation for physiotherapy, osteopathy, chiropractic, or acupuncture; v) referral or recommendation to see a specialist; vi) referral for an investigation or a test.
Timepoint [9] 324712 0
Post-initial consult

Eligibility
Key inclusion criteria
GP participants: New Zealand registered GP; working within the Hutt Valley region; working in a practice which has consented to participate in the study.

Patient participants: LBP of any duration as their primary complaint; over 18 years of age
Minimum age
18 Years
Maximum age
No limit
Sex
Both males and females
Can healthy volunteers participate?
Yes
Key exclusion criteria
GP participants: participation in pilot testing of the FREE approach.

Patient participants: have had back surgery in the last six months; have been unable to do their normal work for more than three of the last six months (normal work includes unpaid work or normal daily activities for those who are retired or work at home); are unable to read and write in English; LBP is due to a potentially serious condition (e.g. cauda equina syndrome, inflammatory arthritis, malignancy, infection, aortic stenosis or aneurysm); LBP is referred from another body region or visceral organ; have a significant concomitant health condition affecting participation (e.g. pregnancy, major psychological disturbance if the patient’s GP considers it unwise for the person to participate).

Study design
Purpose of the study
Treatment
Allocation to intervention
Randomised controlled trial
Procedure for enrolling a subject and allocating the treatment (allocation concealment procedures)
All general practices which will participate in the trial will be recruited before group allocation occurs. Practices will be randomised to intervention or control arms (i.e. all GPs in a practice will be randomised to the same arm). GPs currently working in participating practices will be recruited before practice randomisation occurs. Randomisation of all practices will occur at one time point. Randomisation will be performed by an independent statistician at a central administration site who will communicate this to the primary investigator so that training may be planned in intervention practices and data collection dates agreed with all practices.

Patient recruitment will occur post-randomisation. All patients who meet the inclusion criteria and have an appointment with a participating GP will be invited to participate. These patients will be unaware of the trial’s existence and goals prior to practice presentation. Both patient participants and the research assistants recruiting these patients will be unaware that two different treatment approaches are being compared, as such patient participants and the person screening patients for inclusion will be masked to cluster allocation.
Methods used to generate the sequence in which subjects will be randomised (sequence generation)
General practices (and GPs) will be randomly assigned to either the intervention or control group with a 1:1 allocation using a computer generated randomisation schedule, stratified by the number of participating GPs within the practice (less than or equal to 8 FTE GPs versus greater than 8 FTE GPs). An independent statistician will conduct the randomisation process.
Masking / blinding
Blinded (masking used)
Who is / are masked / blinded?
The people receiving the treatment/s

The people assessing the outcomes
The people analysing the results/data
Intervention assignment
Parallel
Other design features
GP participants will be cluster randomised by practice for the following two reasons: 1) patient randomisation is not possible as GPs may find it difficult to provide ‘usual care’ after learning about the new approach, resulting in contamination at the level of the patient; 2) randomisation at the level of the individual GP may result in contamination effects (i.e. altered behaviour) if GPs discuss the FREE approach with other GPs within their practice, i.e. contamination at the level of the provider (GP).

Patients with acute low back pain who present to intervention group GPs will receive the intervention, patients acute low back pain who present to control group GPs will receive usual care.

The results dataset will be stored with the study arm identity blinded by way of a unique code for each study participant. The key to this code will be held (independent of the dataset for analysis) by the independent statistician, the chair of the Data Monitoring Committee and the Primary Investigator.

Participants will remain blind to the presence of two study arms throughout the trial. Consultation appearance will be similar from a patient perspective except for the provision of an information booklet and referral to the website for intervention arm participants. Control arm participants may be given other information materials (e.g. ACC pamphlet) or be referred to other information sites (e.g. patient.org). Even if a patient perceives they are receiving a different approach to previous GP LBP consultations, they will remain blind to the study hypothesis.

The Trial Administrator will remove all reference to group allocation and study ID number when a participant is discussed by the Trial Management Committee (e.g. exclusion, withdrawal). If there is a risk that these discussions will result in the Trial Statistician being exposed to group allocation information, he will be excused before the discussion occurs. The independent statistician will take his place if necessary.


Phase
Not Applicable
Type of endpoint/s
Efficacy
Statistical methods / analysis
Statistical significance will be judged with an a of 0.05. All results will be reported as estimates of effect size (e.g. mean difference, risk ratio, number needed to treat [NNT]) with 95% confidence intervals (95% CI). All analyses will be conducted on an intention to treat basis.

Analyses will also account for clustering of data. For the patient-level outcomes, this clustering will be specified at the GP level – while randomisation is at the practice level. For the GP level outcomes, clustering will be specified at the practice level. The stratified nature of the randomisation schedule will be handled by including the practice size stratum identifier in analytical models (less than or equal to 8 FTE GPs, greater than 8 FTE GPs).

Primary patient outcome

The primary analysis will follow an intention to treat approach and compare mean RMDQ scores at 6 months between the intervention and control arms using a linear mixed model (adjusted for baseline RMDQ as a second fixed effect). This model will also be adjusted for other factors known to affect the outcome: age, gender, socioeconomic status, current back pain duration and nature (constant or episodic), receipt of recent or ongoing non-GP health care for back pain, previous history of back pain, baseline pain intensity, and baseline beliefs about back pain (self-efficacy and expectation, fear avoidance). The analysis will also adjust for baseline HC-PAIRS scores as measured at the level of the GP (as a measure of baseline treatment competence). This adjustment will improve the precision of the intervention effect estimate, and will also serve to improve handling of missing follow-up data.

A random effect for GP will be used to account for the clustered nature of the data. Linear mixed models will be used to reduce bias due to incomplete follow-up: in this case, data from the 2w, 6w, and 3m follow-up times will be included in the model when estimating differences at 6m to allow for missing data at this final end point. The relative risk of a 30% improvement in RMDQ scores will be reported, along with the numbers needed to treat (NNT) to achieve this target. As a sensitivity analysis, the NNT and RR of a 2.5 point absolute reduction in RMDQ score will also be reported (in line with the absolute effect size stipulated for the primary outcome).

Subgroup analysis will compare mean RMDQ scores at 6 months using the same model as the primary outcome for those who presented with a new episode of acute back pain of less than 6 weeks duration who had not received health care for back pain in the previous three months.

Secondary patient outcomes

Mean RMDQ scores at each of 2w, 6w, and 3m will be compared using the same analytical framework as the RMDQ primary outcome (using each of these interim time-points as the end-point in analysis, with RMDQ scores at all time-points included in analysis).

Pain intensity and impairment data (11 point scales) will also be compared between arms at 6 months. These will follow the same analytical framework as the RMDQ primary outcome: use a linear mixed model to compare means at 6 months adjusted for baseline scores; include interim scores (2w, 6w, 3m) to account for missing final-endpoint outcome data; and include random effect for GP to handle the clustered nature of the data.

Patient satisfaction data (7 point scale) will be compared at 0w, 2w, 6w, and 6m (outcomes reported separately, from a single analysis model) using a generalised linear mixed model with a cumulative logit link function. In the absence of clustering, the hypothesis test component of this model is equivalent to the non-parametric Mann-Whitney-Wilcoxon test: in this study the model will be run including a random effect for GP, again to account for the clustering of responses by GP.

Patient cost-utility analysis

A Cost Utility Analysis will estimate mean incremental cost per quality-adjusted life year (QALY) gained (QALYs will be calculated with the NZ EQ-5D Tariff as recommended by PHARMAC) from health care system, Accident Compensation Corporation, and societal perspectives. Reference costs will be assigned for i) all health care items, to allow direct comparison and decrease patient recall requirements, ii) paid work (based upon gender and age mean income) to avoid data being skewed by participants with high income and make results more generalisable to the general population, and iii) unpaid/voluntary work (based upon the minimum wage). The human capital approach will be used for work loss. As each patient will be followed for 6-months, discounting will not be necessary.

Confidence intervals will be calculated for the Incremental Cost-Effectiveness Ratio (ICER). Bootstrapped data will be displayed on a cost-effectiveness plane, and cost-effectiveness acceptability curves (CEAC) will be calculated to determine the likelihood that the FREE approach will be considered cost-effective using one, two, and three times GDP per capita as policy-relevant Willingness To Pay (WTP) thresholds. Costs or effects of important variables that have, or are likely to have, wide variability (e.g. between small towns and large cities, or chance variability introduced by large costs in a single individual in one group probably due mostly to chance) will be subjected to sensitivity analyses by varying the input value over a range of feasible estimates. We will report the results across that range.

GP outcomes

As per the patient outcomes, all analyses here will be conducted on an intention-to-treat basis. Clustering of responses for the GP-level outcomes will be handled by clustering according to GP practice.

Mean GP HC-PAIRS (potential score range: 13 to 91 points) and confidence scores (potential range 4-20) at 4 months will be compared between the intervention and control arms using a linear mixed model (adjusted for baseline confidence as a second fixed effect; including 4week measure to handle missing data at 4 months; and including GP practice as a random effect in the model).

Reported clinical behaviour relating to the case vignette (categorical outcome from five options) will be compared in two ways:
i. changes in response proportions of reported behaviour across the entire range of responses will be compared between arms using a generalised linear mixed model with a cumulative logit link function (to allow for ordinal nature of outcome), adjusted for baseline values and including a random effect for GP practice (to handle clustering)
ii. changes in the proportions of consultations following 'guideline-consistent' reported behaviour (in contrast to 'guideline-inconsistent' behaviour) will be compared between arms using a generalised linear mixed model with a logit link (i.e. akin to a logistic regression specification), with responses categorised into two levels (“guideline consistent” vs “guideline inconsistent”). This analysis will be adjusted for baseline values and include a random effect for GP practice to handle clustering.

GP behaviour differences will be compared for the LBP patient consultation between intervention and control group GPs using data extracted from patient management systems and patient report.

Each of the six indicators (recommended/referred to: Work recommendation, Activity recommendation, Medication, Physiotherapy/Osteopathy/Chiropractic, specialist, X-rays or scans) will be handled separately. Summary statistics will be calculated for each GP (e.g. proportion of all patients seen by that GP who reported being advised to take time off work); a linear mixed model will then compare the two study arms on the mean score for each GP for work/activity recommendations or the average proportion of patients (across GPs) meeting that indicator for medication, referral, or investigations (separate analyses for each indicator; accounting for clustering by GP practice).

Sample size

For 80% power to detect a between-group difference of 2.5 RMDQ points (the minimal clinically important difference) at 6 months assuming SD = 6.0, at a p < 0.05 significance level, an individually randomised trial would require 91 patients per group. This has been inflated to 110 patients per group after adjustment for GP cluster effects (assuming a GP intracluster correlation coefficient at a conservative 0.05 and on average 5 patients per GP completing the trial). We aim to recruit 275 patient participants to allow for 20% loss to follow-up.

This represents 22 GPs per study arm. To account for potential zero-recruitment by some GPs, we will recruit 30 GPs per study arm. To further account for patient loss to follow-up (assumed ~20% of patients per GP) we will ask GPs to recruit 7 patients. This gives an expected maximal sample size of 210 patients per study arm (total sample size = 60 GPs, 420 patients). Sample size calculations were performed in Stata 12 (Statacorp, Texas).

Post-trial follow-up

Post-trial patient participant follow-up includes collection of information on (RMDQ, pain intensity, fear avoidance, anxiety, and cost data) at two years after patient entry into the study; these results will be analysed following the methods described above for these outcomes, accounting for non-retention of participants at the two year follow-up.

Recruitment
Recruitment status
Completed
Date of first participant enrolment
Anticipated
1/08/2016
Actual
6/07/2016
Date of last participant enrolment
Anticipated
31/07/2017
Actual
31/07/2017
Date of last data collection
Anticipated
30/08/2019
Actual
8/08/2019
Sample size
Target
275
Accrual to date
Final
226
Recruitment outside Australia
Country [1] 7951 0
New Zealand
State/province [1] 7951 0
Hutt Valley

Funding & Sponsors
Funding source category [1] 293803 0
Other
Name [1] 293803 0
Accident Compensation Corporation
Country [1] 293803 0
New Zealand
Primary sponsor type
Individual
Name
Dr Ben Darlow
Address
Department of Primary Health Care and General Practice
University of Otago, Wellington
23 Mein Street
Newtown
Wellington 6021
Country
New Zealand
Secondary sponsor category [1] 292635 0
None
Name [1] 292635 0
Address [1] 292635 0
Country [1] 292635 0

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 295235 0
Central Health and Disability Ethics Committee
Ethics committee address [1] 295235 0
Ethics committee country [1] 295235 0
New Zealand
Date submitted for ethics approval [1] 295235 0
14/03/2016
Approval date [1] 295235 0
05/05/2016
Ethics approval number [1] 295235 0
16/CEN/43

Summary
Brief summary
Trial website
Trial related presentations / publications
Public notes

Contacts
Principal investigator
Name 66642 0
Dr Ben Darlow
Address 66642 0
Department of Primary Health Care and General Practice
University of Otago, Wellington
23 Mein Street
Newtown
Wellington 6021
Country 66642 0
New Zealand
Phone 66642 0
+6449186051
Fax 66642 0
+6443855539
Email 66642 0
[email protected]
Contact person for public queries
Name 66643 0
Ben Darlow
Address 66643 0
Department of Primary Health Care and General Practice
University of Otago, Wellington
23 Mein Street
Newtown
Wellington 6021
Country 66643 0
New Zealand
Phone 66643 0
+6449186051
Fax 66643 0
+6443855539
Email 66643 0
[email protected]
Contact person for scientific queries
Name 66644 0
Ben Darlow
Address 66644 0
Department of Primary Health Care and General Practice
University of Otago, Wellington
23 Mein Street
Newtown
Wellington 6021
Country 66644 0
New Zealand
Phone 66644 0
+6449186051
Fax 66644 0
+6443855539
Email 66644 0
[email protected]

Data sharing statement
Will individual participant data (IPD) for this trial be available (including data dictionaries)?
Yes
What data in particular will be shared?
All of the individual participant data collected during the trial, after de-identification
When will data be available (start and end dates)?
10 September 2019 onwards (there is no end date for data availability).
Available to whom?
Anyone who wishes to access it
Available for what types of analyses?
Any purpose
How or where can data be obtained?
Hosted on Dryad: Darlow, Ben et al. (2019), Data from: The Fear Reduction Exercised Early (FREE) approach to management of low back pain in general practice: a pragmatic cluster-randomised controlled trial, Dryad, Dataset, https://doi.org/10.5061/dryad.4t375b2


What supporting documents are/will be available?

Doc. No.TypeCitationLinkEmailOther DetailsAttachment
1933Study protocol   



Results publications and other study-related documents

Documents added manually
No documents have been uploaded by study researchers.

Documents added automatically
SourceTitleYear of PublicationDOI
EmbaseThe Fear Reduction Exercised Early (FREE) approach to low back pain: Study protocol for a randomised controlled trial.2017https://dx.doi.org/10.1186/s13063-017-2225-8
EmbaseThe Fear Reduction Exercised Early (FREE) approach to management of low back pain in general practice: A pragmatic cluster-randomised controlled trial.2019https://dx.doi.org/10.1371/journal.pmed.1002897
EmbaseA Guideline-Implementation Intervention to Improve the Management of Low Back Pain in Primary Care: A Difference-in-Difference-in-Differences Analysis.2023https://dx.doi.org/10.1007/s40258-022-00776-3
N.B. These documents automatically identified may not have been verified by the study sponsor.