How to Determine Effective CDS Tools for High-Risk Prescriptions

Clinical decision support (CDS) tools in EHRs have the potential to reduce high-risk prescribing for older adults, yet the results have not lived up to that potential, according to a study published in Implementation Science.

Furthermore, EHR tools have proven to have limited effectiveness at boosting prescribing methods.

Prescribing high-risk medications to older adults can have deadly consequences and unfortunately, it is far too common. Reducing these efforts has had limited success and EHR-based tools, such as CDS tools, aim to fix this issue. However, these tools have had limited adoption rates for prescribing high-risk medicine to older adults and most are poorly designed or are not accurate, according to the study authors.

The objective is to determine whether designing EHR tools using behavioral science principles reduces inappropriate prescribing and clinical outcomes in older adults.

Researchers at Atrius Health conducted a Novel Uses of adaptive Designs to Guide provider Engagement in Electronic Health Records (NUDGE-EHR) trial intending to answer these questions.

The NUDGE-EHR trial was a randomized pragmatic trial that aims to identify the best way to integrate EHR tools and it will provide guidance on which specific components of the tools could change prescribing and inform implementation. The authors suggested the design could be utilized as a blueprint to consider or overcome adaptive trials in EHR implementation in the future.  

The research team utilized this trial to identify the effectiveness of multiple EHR tools among primary care providers and patients over 65 who are using benzodiazepines, sedative hypnotics, or anticholinergics.

In the first stage, the research team randomized providers and their patients to one of 15 EHR tools designed using behavioral principles. After a six-month follow-up, researchers will identify and rank which EHR tools can effectively reduce inappropriate prescribing.

Looking forward to stage two, the research team will randomize the stage one providers to one of the up to five most promising interventions or continue usual care. The next outcomes will include the number of medications prescribed and utilized.

Additionally, the study authors conducted a systematic review that found 57 percent of studies reported CDS influenced provider behavior and most studies showed a high-risk bias. The authors said EHR tools could be ineffective due to a lack of focus on prescribing behavior, workflow, or clinical inertia.

“Once the trial is completed, these results will also need to be considered in relation to other efforts to reduce inappropriate prescribing in older adults,” the authors wrote. “Limited prior evidence suggests that providing information alone via decision support at the time of a patient encounter could be insufficient at reducing prescribing on its own.”

While EHR and health IT developers continue to optimize technology, there will be ample opportunities to boost prescribing efforts. Improving behavioral sciences to gain insight on prescriber motivations and leveraging workflow types could also aid this discovery.

“These observations and application of their principles has effectively changed behavior in other settings and are likely applicable to EHR systems to improve the uptake of evidence-based care, such as improving prescribing in older adults,” the study authors said. “The effectiveness of tools could be enhanced by leveraging principles of behavioral economics and related sciences, but they have had very limited application in EHRs and, more specifically, for prescribing in older adults.”

The researchers said NUDGE-EHR will determine which EHR tool components are most impactful at disrupting or improving provider behavior. This study design allows for the testing of separate implementation strategies and it will allow for added evidence that EHR developers and implementers can utilize in the future.

“NUDGE-EHR will advance our understanding about how behavioral science can optimize clinical decision support to reduce inappropriate prescribing and improve patient health outcomes as well as how to use adaptive trial designs in healthcare delivery and implementation science,” the study authors concluded.