Q&A: Informatics expert sets off alarms on alert fatigue

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - Shobha Phansalkar, RPh, PhD

Mobile devices send alerts throughout the day, whether calendar reminders or notes about a sale at a nearby store. For clinicians, such near-constant alerts can seem impossible to fully address, leading to anxiety and stress. In an exclusive interview with Clinical Innovation & Technology, Shobha Phansalkar, RPh, PhD, director of informatics and clinical innovation at Wolters Kluwer Clinical Drug Information, explained the impact of alert fatigue on clinicians and the challenges it presents to electronic medical records and interoperability.

Clinical Innovation & Technology: How has the growing reliance on digital devices increased alert fatigue?

Shobha Phansalkar, RPh, PhD: Healthcare professionals have benefited greatly from new and increasingly sophisticated electronic tools, including electronic health record (EHR) systems with clinical decision support (CDS) functionality. At the same time, these systems have presented users with new challenges. For example, while CDS tools can streamline the workflow by providing a clinician with fast and efficient access to useful information at the point-of-care, the additional information and alerts can impede the workflow if alerts do not facilitate the decision-making process. A systematic review found that users chose to override between 49 and 96 percent of medication decision support alerts, and another study found that a large number of overridden alerts were clinically significant. In order for CDS to continue to evolve, developers must not only direct more and better content into the hands of providers, they must also get a deeper understanding of how to deliver content that is not so voluminous or irrelevant that it overwhelms the user, disrupts the workflow and creates a risk of alert fatigue where valuable information is lost in the background noise of unhelpful information.

What led to alert fatigue becoming a larger consideration in EHR development? How can EHRs assist in decreasing alert fatigue?

With the widespread adoption of EHRs, content and systems developers have learned from user experiences about how the additional availability of information has benefited and created new challenges for users. The next phase of the EHR development process is to use context-specific information about the clinical care setting and patients to refine functionality and improve the relevance of reference content and alerts that are delivered to the patient at the point of care. In particular, when EHRs can harness greater patient context from the medical records stored in the patient’s profile, that information can influence the alerts provided to clinicians in a way that makes them less voluminous and more relevant. Clinicians receiving more useful alerts are less likely to suffer from alert fatigue.

With the movement toward interoperability will alert fatigue increase?

We do not know whether interoperability, which allows CDS to draw data from disparate systems, will reduce or increase alert fatigue. On the one hand, more data may lead to more alerts. On the other hand, a broader data pool, if used effectively, has the real potential to improve the specificity of alerts and reduce alert fatigue. Standards, such as Fast Healthcare Interoperable Resources (FHIR), may offer the opportunity to draw and deliver patient data to disparate systems and bring greater context to a patient’s profile. The key challenge for CDS is how to harness this additional data and use it to deliver better, but not necessarily more, alerts. For example, if a system could take patient characteristics like race, age, obesity or co-morbidities into account when providing information to a clinician treating a hypertensive patient, it could provide the clinician with more effective information about potential choices for first- and second-line antihypertensive agents. Similarly, if systems could take patient laboratory parameters into account, they would be better equipped to determine whether a clinician would benefit from an alert about a possible risk of hyperkalemia due to a drug-drug interaction.

What is the best approach to addressing physician alert fatigue? Are there any examples of success in limiting alert fatigue?

Perhaps the ideal approach to address alert fatigue is to focus on CDS development that includes greater clinical care setting and patient context in the system logic in a way that better appreciates the specific-use cases in which decision support alerts are being