A literature review by Vanderbilt University Medical School researchers has identified barriers that make clinical decision support (CDS) systems less effective for some patients, especially those without the necessary technological skills and equipment.
CDS systems involve technology that can assist in diagnoses or be used to alert providers, for example to a patient’s negative drug interactions, through automated alerts and reminders.
“Scientists now recognize the importance of bringing health data directly to patients to guide their health decisions.”
The literature review by Vanderbilt researchers, reported in Yearbook of Medical Informatics, looked at 45 relevant studies from 2020 to 2022. They found the most frequently sent alerts, 33.3 percent, occurred as point-of-care reminders to providers.
More Patient Engagement
The study offers good news about rising patient engagement in CDS systems, identifying an even split of systems designed for patients (patient-facing) and those designed for clinicians.
“Scientists now recognize the importance of bringing health data directly to patients to guide their health decisions,” said lead author Brian Douthit, Ph.D., R.N., a U.S. Office of Veterans Affairs postdoctoral fellow in medical informatics and quality improvement at Vanderbilt University Medical Center.
“CDS is uniquely poised to deliver personalized interventions, yet we continue to see CDS widen disparities.”
Patient-facing CDS systems use digital technologies to offer health services and promote specific types of patient engagement. By using sensors to monitor a diabetes patient’s blood glucose levels or a heart-disease patient’s activity level, for example, they can help facilitate self-management of chronic conditions. The systems also may help patients meet more easily with their health care team, involving less or no travel.
Information technology is an important part of quality health care, but can create a “digital divide,” Douthit said.
“Health information technology developers must ensure that CDS does not widen health disparity gaps, and that systems are designed to carefully maximize the impact on all patients, he said. “CDS is uniquely poised to deliver personalized interventions, yet we continue to see CDS widen disparities.”
Four Major Themes
The literature review found four major areas where there was a potential for disparities:
Inaccessibility of technology. Technologies may not be accessible to the socioeconomically disadvantaged or for the seven percent of Americans who do not have access to the internet. A patient-facing CDS tool typically requires the patient to have a smart phone. “Can we convert CDS to a phone call instead?” Douthit asked. Language and visual impairment are two other common accessibility barriers.
Access to care. A mobile app alert does not guarantee access. If CDS recommends that a patient at high risk for lung cancer receive an MRI, lack of transportation can derail the requested response, especially in rural areas where advanced medical resources could be hours away, Douthit said.
Trust of technology. Most advanced CDS tools are driven by artificial intelligence (AI), but Douthit says many patients and clinicians mistrust AI algorithms. “If an AI algorithm suggests radiological screening for lung cancer, the patient may say, ‘Why should I trust the algorithm and expose myself to radiation? I have never smoked cigarettes,’” he said.
Technology literacy. CDS should be as straightforward as possible, Douthit advises, not overengineered with technological requirements. Douthit said a classic example is the “elderly patient with a smart phone who has never been taught to use its major functions, such as accessing the internet.” Doing something in a text or phone call may be preferable to the portal.
Links to Health Disparities
The next step in this research, Douthit said, is to understand the link between the digital divide and health disparities by looking at subgroups and subsets of disease states.
Without attention to these issues, CDS may perpetuate disparities by excluding known risk factors for specific subgroups, inappropriately applying subgroup risk, or underrepresenting some subgroups.
“Once we become more acutely aware how CDS affects health disparities, we can focus efforts on those who may not be able to access modern technology,” Douthit said.
