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
ACTRN12618000769280
Ethics application status
Approved
Date submitted
3/05/2018
Date registered
8/05/2018
Date last updated
9/07/2018
Type of registration
Prospectively registered

Titles & IDs
Public title
Choice of management options for low back pain: A vignette-based study in general practitioners
Scientific title
Effect of two behavioural 'nudging' interventions on choice of management options for low back pain: A randomised vignette-based study in general practitioners
Secondary ID [1] 294785 0
None
Universal Trial Number (UTN)
Trial acronym
Linked study record

Health condition
Health condition(s) or problem(s) studied:
Low back pain 307708 0
Condition category
Condition code
Musculoskeletal 306765 306765 0 0
Other muscular and skeletal disorders

Intervention/exposure
Study type
Interventional
Description of intervention(s) / exposure
The interventions are known as behavioural 'nudging' interventions. In this trial the behavioural 'nudge' refers to subtle changes to the way management options are presented on a computer screen. Each participant is exposed to 4 vignettes by following a link to an online survey. The vignettes present a short case study of a person with low back pain. After reading the vignette participants will be taken to a screen where they are asked to choose from one of six management options by clicking a box next to the option. All participants receive the same six management options for a given vignette. Management options are a combination of guideline concordant (e.g. simple medicines) and guideline discordant options (e.g. opioid medicines) and vary according to the vignette. The only difference between trial arms is how the options are presented on the computer screen. Each trial arm has the menu of management options displayed differently. There are four trial arms a participant could be randomised to. The menu for Intervention Group 1 will include 3 guideline concordant options listed vertically (these are 'default' options) and discordant options which are only displayed after clicking an additional button on screen. The menu for Intervention Group 2 will include guideline discordant options listed horizontally rather than vertically. The menu for Intervention Group 3 will include guideline concordant options presented as default options and guideline discordant options listed horizontally. The survey is self-adminstered online, takes approximately 15 minutes, and is completed once only. The online survey software (Qualtrics) remind participants who fail to complete survey questions. A recruitment company will monitor those who enrol in the study but do not complete the survey.
Intervention code [1] 301096 0
Behaviour
Comparator / control treatment
The menu for the Control Group will include all six management options listed vertically, in random order
Control group
Active

Outcomes
Primary outcome [1] 305758 0
The primary outcome will be the proportion of scenarios where clinicians choose any of the guideline discordant (low-value) clinical management choices. Information on management choices will be collected from responses to online survey questions that the participant completes immediately after reading the clinical vignettes on their computer screen. All responses for the primary outcome will be collected using the online survey software Qualtrics. We will classify management options as "guideline discordant (low-value)" if they conflict with the 2016 NICE clinical guideline titled 'Low back pain and sciatica in over 16s: assessment and management' or the 2017 American College of Physicians clinical guideline for the management of low back pain.
Timepoint [1] 305758 0
Immediately after survey is completed
Secondary outcome [1] 346381 0
We will assess the likelihood a participant would choose each of the options presented using a Likert scale: ‘How likely would you be to recommend each of the below options for this case? (1 = very unlikely and 5 = very likely)’.’

Timepoint [1] 346381 0
Immediately after survey completion
Secondary outcome [2] 347025 0
Drop out rate
Timepoint [2] 347025 0
Immediately after survey

Eligibility
Key inclusion criteria
General practitioner in Australia
Minimum age
18 Years
Maximum age
No limit
Gender
Both males and females
Can healthy volunteers participate?
Yes
Key exclusion criteria
None

Study design
Purpose of the study
Educational / counselling / training
Allocation to intervention
Randomised controlled trial
Procedure for enrolling a subject and allocating the treatment (allocation concealment procedures)
Methods used to generate the sequence in which subjects will be randomised (sequence generation)
Masking / blinding
Blinded (masking used)
Who is / are masked / blinded?
The people receiving the treatment/s
The people administering the treatment/s
The people assessing the outcomes
The people analysing the results/data
Intervention assignment
Factorial
Other design features
Phase
Not Applicable
Type of endpoint(s)
Statistical methods / analysis
2.4 Measurements and statistical analysis plan

The primary outcome will be the proportion of scenarios where clinicians choose any of the low-value clinical management choices. We will assess the likelihood they would choose each of the options presented as a secondary (continuous) outcome using a Likert scale by asking participants to answer the following question after they have made their choice: ‘How likely would you be to recommend each of the below options for this case? (1 = very unlikely and 5 = very likely)’.’

Regression analyses will investigate whether nudges can reduce the proportion of low value care choices based on the following primary hypotheses:
H1: Default settings reduce low value care choice (when high value options shown as a default)
H2: Partitioning (where low-value options are displayed horizontally) reduces low value care choice

Sample size estimate and analysis plan
Previous research has found nudging interventions can have small to moderate effects on dichotomous measures of decision making. For example, Tannenbaum found that although the average effect of partitioning was a reduction of around 12% in those choosing low value options, in some groups of scenarios partitioning alone could reduce the number of aggressive antibiotics chosen by up to 31.2%.

A logistic regression of the binary outcome variable low value choice (selected vs. not selected) on two binary independent variables (default setting vs. non-default setting; partitioning vs. non-partitioning) will be used for the primary analyses, with clustering by GP to control for repeated observations across the vignettes We estimate a conservative baseline rate of 0.45 (i.e., low value choices chosen in 45% of observations) in the no nudge group. With an allocation proportion of 0.5 to each of the nudges (0.25 to each of the four experimental groups), a sample of 163 observations will provide 80% power at the 0.05 significance level to detect a rate ratio of at least 0.44 due to the presence of either nudge. This corresponds to a difference between no nudge and nudge groups of approximately 25%. We adjusted this sample size since a regression of the variable of interest on the other covariates in the logistic regression model is expected to explain 25% of the variance. Thus, to detect a between-group difference in the proportion of vignettes where clinicians choose a low value of 25% (i.e., anticipated rate in the intervention group of 20%) with at least 80% power, and assuming alpha of 0.05, a total sample size of 120 is required to test the main effects of single nudges

Blinded interim analysis
We will perform a blinded interim analysis after 20 participants complete the study (80 completed scenarios). We will evaluate the following: 1) proportion of whole sample making low value choices, 2) comments in response to vignettes and 3) completion rates. Because we anticipate the control rate of low value care choices to be 45%, we will edit the vignettes if participants make low value care choice in less than 20% of the 80 completed scenarios. We will in no way alter the experimental or control intervention, only the vignettes. We will also consider adjusting the vignettes if participants suggest this is necessary in their free-text responses. If completion rates are low we will consider adjust the survey flow and use of reminders.

Exploratory analyses
Secondary hypotheses will include:
H3: Combining default settings plus partitioning reduces low value care choices more than either single nudge
H4: Participants with low CRT scores will be more influenced by implicit nudges than participants with high CRT scores
We will also compare drop-out rate (number of participants who are randomised but do not complete the survey) between experimental groups.

Recruitment
Recruitment status
Completed
Date of first participant enrolment
Anticipated
Actual
Date of last participant enrolment
Anticipated
Actual
Date of last data collection
Anticipated
Actual
Sample size
Target
Accrual to date
Final
Recruitment in Australia
Recruitment state(s)
ACT,NSW,NT,QLD,SA,TAS,WA,VIC

Funding & Sponsors
Funding source category [1] 299392 0
University
Name [1] 299392 0
University of Sydney
Address [1] 299392 0
Camperdown NSW 2006
Country [1] 299392 0
Australia
Primary sponsor type
Individual
Name
Dr Adrian C Traeger
Address
University of Sydney
Level 10 KGV Building | Missenden Road, Camperdown | NSW | 2050
Country
Australia
Secondary sponsor category [1] 298671 0
Individual
Name [1] 298671 0
Dr Carissa Bonner
Address [1] 298671 0
University of Sydney
Rm 226A, Edward Ford Building A27 | The University of Sydney | NSW | 2006
Country [1] 298671 0
Australia

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 300296 0
Research Ethics and Governance Committee, RPA Hospital
Ethics committee address [1] 300296 0
Royal Prince Alfred Hospital
Camperdown NSW 2050
Ethics committee country [1] 300296 0
Australia
Date submitted for ethics approval [1] 300296 0
15/03/2018
Approval date [1] 300296 0
01/05/2018
Ethics approval number [1] 300296 0
HREC/18/RPAH/138

Summary
Brief summary
The objective of this study is to collect data on current general practitioner preferences for management of low back pain. This includes both preferences for imaging of low back pain and preferences for prescription of medication for low back pain.
Trial website
Trial related presentations / publications
Public notes

Contacts
Principal investigator
Name 83190 0
Mr Tet Far Soon
Address 83190 0
Rm 226A, Edward Ford Building A27 | The University of Sydney | NSW | 2006
Country 83190 0
Australia
Phone 83190 0
+61 2 9351 7125
Fax 83190 0
Email 83190 0
tsoo7799@uni.sydney.edu.au
Contact person for public queries
Name 83191 0
Dr Carissa Bonner
Address 83191 0
Rm 226A, Edward Ford Building A27 | The University of Sydney | NSW | 2006
Country 83191 0
Australia
Phone 83191 0
+61 2 9351 7125
Fax 83191 0
Email 83191 0
carissa.bonner@sydney.edu.au
Contact person for scientific queries
Name 83192 0
Dr Adrian Traeger
Address 83192 0
University of Sydney
Level 10 KGV Building | Missenden Road, Camperdown | NSW | 2050
Country 83192 0
Australia
Phone 83192 0
+61416122784
Fax 83192 0
Email 83192 0
adrian.traeger@sydney.edu.au

No information has been provided regarding IPD availability
Summary results
Have study results been published in a peer-reviewed journal?
Other publications
Have study results been made publicly available in another format?
Results – basic reporting
Results – plain English summary