New study suggests that simple nudges could improve clinical trial registration and reporting
Simple ‘nudges’ could motivate more researchers to register their studies and make results public, a new study applying behaviour science approaches to a clinical trial registry suggests.
The study found that changing the wording of reminder emails sent out by the German trial registry DRKS in some cases had a significant effect on researchers’ subsequent actions.
Currently, many German trials are registered retrospectively or not at all, so the study’s findings may inform efforts to reduce research waste and improve the robustness of medical research.
What is nudging?
Nudging approaches have been widely applied in the public sector as a low-cost way to achieve desirable outcomes, but this is the first time they have been deployed by a trial registry.
The basic premise of nudging is that slight changes to processes or messaging can substantially modify people’s behaviours.
What is the DRKS registry?
The DRKS registry is widely used in Germany to register non-drug clinical trials and observational medical studies.
Currently, its database contains several hundred incomplete entries where researchers started to register a study but did not complete the process. These preliminary entries are not publicly visible.
DRKS’ approach to date has been to email automated reminders to those researchers after 2, 4 und 8 weeks, and thereafter every six months. The emails ask researchers to complete the process (or delete the entry if the study was cancelled.)
Study design
Using a classic nudging approach, the study stratified 756 incomplete entries and randomised them into three groups:
Group A received a one-off email explaining the possible negative consequences of not registering a study
Group B received a one-off email combining a variety of positive encouragements to register
Group C received no email
Five weeks later, the study assessed whether researchers had taken one of several actions, including logging into the system, contacting DRKS for support, etc.
Key findings
Both interventions were successful at nudging researchers to complete their registrations.
While positively framed email A elicited more registrations than negative email B, the difference was not statistically significant (p=0.317).
However, the positively framed email A was far more successful than email B at motivating researchers to delete entries for studies that had never started (p=0.006).
In addition to clinical trials, DRKS also accepts registrations for observational studies. The study found no difference in the impact of its nudges on interventional clinical trials versus observational studies.
Why all registries should routinely nudge
The study breaks new ground in regulatory science by demonstrating how clinical trial registries can use low-cost nudging of investigators to improve data quality.
It opens the path for other registries to also run tests systematically comparing the impact of differently worded emails on investigator behaviour. Trial registries are well positioned to do this because they hold large datasets including investigators’ email addresses. Hopefully, other registries will soon apply this approach to improve results reporting.
As the chart below shows, researchers were most likely to convert their incomplete entries into full study registrations if the original entry had been made less than 200 days earlier. In future, registries should use nudging from the very first reminder message onwards.
Commercial marketers routinely mainstream A/B testing across all their communications on an ongoing basis. There is no reason why registries – including study registries for other scientific disciplines – should not do the same.
No substitute for effective regulation
At the same time, the study argues that nudging is no substitute for effective regulation. Nudging can help to reduce non-registration and (probably) non-reporting, but cannot resolve the problem entirely.
The only way to completely fix the problem is for regulators to directly register clinical trials as soon as they receive ethics approval. The UK has formally decided to take this approach, but the UK Health Research Authority has yet to deliver this key component of the country’s national transparency strategy.
A coalition of health groups led by Cochrane Germany is currently pushing for a similar solution in Germany.
The nudging study notes that problems in Germany run even deeper:
“The waiting time for ethics approval decisions was given [by researchers] as a reason for the sometimes considerable delays [in completing study registrations]. One principal investigator even asked for the deletion of his preliminary entry, which had been made over 800 days earlier, as he had still not heard back from the ethics committee.”
Study strengths and weaknesses
The key strength of the study is methodological, introducing behavioural science approaches into regulatory science. The study design and reporting are very rigorous.
DRKS is a comparatively small registry, so the study has a quite small sample size, and includes entries submitted to DRKS a long time ago. Nudging to improve results reporting was not tested. The five week follow up period is very short, but the author plans to follow up again after 6 and 12 months.
The study was run by Lisa Bieselt in the framework of a masters dissertation. She can be contacted on LinkedIn. The full study is available online in German language (click here for PDF download).