Using behavioural insights to reduce unnecessary antibiotic prescriptions by New Zealand doctors

How to cite

Gyani, Alex; Chappell, Nathan; Hayward, Sarah. 2020. "Using behavioural insights to reduce unnecessary antibiotic prescriptions by New Zealand doctors." AEA RCT Registry. AEARCTR-0005526. https://doi.org/10.1257/rct.5526-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Alex Gyani* , Nathan Chappell , Sarah Hayward

* Primary investigator

Registry ID
AEARCTR-0005526
DOI
https://doi.org/10.1257/rct.5526-1.1
First registered
Last updated
Status
Completed
Trial dates
2019-08-19 – 2020-04-30
Countries
New Zealand ()
Registry record
http://www.socialscienceregistry.org/trials/5526

Abstract

Antimicrobial resistance (AMR) is a global concern, which the World Health Organisation has named one of the biggest threats to global health, food security and development today. The over-prescription of antibiotics is a key driver of AMR. Unnecessary prescriptions are especially problematic in New Zealand, which had the 7th highest community prescription rate in the OECD in 2015.

The Behavioural Insights Team's previous work in the UK has shown that a letter with personalised feedback to practices can reduce prescriptions of unnecessary antibiotics, and a recent replication in Australia by the Behavioural Economics Team of the Australian Government was even more effective by targeting individual General Practitioners (GPs).

The current trial is being conducted by BIT, and draws on this prior work to tailor it to the New Zealand context. The trial aims to test whether personalised feedback to New Zealand GPs, in the form of a letter, can reduce prescriptions of unnecessary antibiotics. The trial will also test the impact on prescribing to Māori and Pacific patients, who tend to experience worse health outcomes.

Intervention

This trial contains two trial arms: the treatment group (who receive the letter intervention), and the control (who receive no letter intervention).

The trial launched on 19 August 2019, with the one-off letters sent out, and prescribing behaviour will be analysed from September - December 2019.

All high-prescribing GPs in the trial are stratified by area (District Health Board) before randomly allocating each GP to either the treatment group or the control group.

Experimental design

The intervention will be evaluated with a two-armed randomised controlled trial (RCT), with randomisation at the GP level and stratification by District Health Board (area).

Design details

Randomization method
Randomisation in this trial is done using the 'stratified' function within the 'splitstackshape' package in R, to randomise GPs into the treatment or control arm within each stratum,
Randomization unit
Randomisation is at the GP level.
Clusters
This trial is not clustered.
Observations
There are 1,234 high-prescribing GPs in our sample, of whom 617 are sent the intervention letter.
Arms
There are 617 GPs per trial arm.
Primary outcomes
There are two primary outcomes: 1. Do the intervention letters reduce the prescribing rate of high-prescribing GPs? The prescribing rate is measured as the number of patients prescribed antibiotics per 100 patients prescribed any medicine, and is taken as the monthly average from September - December 2019. 2. Do the intervention letters reduce the absolute number of antibiotics prescribed? This is measured as the number of antibiotic scripts dispensed from the doctors' prescribing, and is taken as a monthly average from September - December 2020.
Secondary outcomes
The secondary outcomes are listed below: 1. Do the intervention letters reduce prescribing to Māori patients, to Pacific patients, and to non-Māori and non-Pacific patients? This will be measured as the prescribing rate (as defined in the first primary outcome measure) to Māori patients, to Pacific patients, and to non-Māori and non-Pacific patients. 2. Do the intervention letters change prescribing of paracetamol? This will be measured as the number of people prescribed paracetamol scripts per 100 prescribed any medicine.

Cited by

No citation data available for this registration yet.

Keywords