Better clinical decision support improves CPOE effectiveness

Article

An effective CPOE system needs clinical decision support to help prescribers, pharmacists, nurses, and others use the system effectively.

Computerized provider order entry (CPOE) is one of the most effective steps a healthcare system can take to reduce medication order errors, but CPOE alone is not enough. An effective CPOE system needs clinical decision support (CDS) to help prescribers, pharmacists, nurses, and others use the system effectively.

The goal of CPOE is to promote safer prescribing and better patient support," said Lolita White, PharmD, lead clinical analyst at National Children's Medical Center in Washington, D.C. "The timing of information that is presented by the system and what is presented are important factors in how effective CPOE is in any institution."

White discussed her experiences with CDS as National Children's moved from paper orders to CPOE during a webinar in March. The presentation was sponsored by Pharmacy OneSource.

Success lies in the CDS

But while CPOE helps reduce medication errors, it is not omnipotent. During 8 years of transition from paper orders to CPOE, White found that CDS is a key factor in the success or failure of CPOE. The more effective the CDS, the more effective electronic ordering can be at reducing medication errors and improving outcomes.

Automation can create a false sense of security, White cautioned. CPOE eliminates many errors but can introduce new types of error. Every time a prescriber sees a drop-down list of medications or doses, it is possible to click on the wrong entry.

Neither CPOE nor CDS are plug and play, she cautioned. Systems must be customized for each institution based on formulary needs and local prescribing practices. And institutions need to constantly test, refine, update, and customize systems to keep up with changes in prescribing needs, trends, and practices.

A big problem

Alert fatigue has emerged as a key problem area. CDS comes into play alerting prescribers to drug interactions, allergies, diagnostic, and therapeutic advice; routine tasks, such as updated lab values; and important information that can affect orders. If alerts don't come at the right time in the order entry process or are less than vital, prescribers begin to tune them out.

CDS must organize and prioritize alerts, White said. The most important alerts should generate a hard stop, halting the order until some vital error is addressed.

"You should not be able to order a fatal drug combination," she said. "That's a hard stop situation."

Important but non-fatal alerts can generate a soft stop that the prescriber can override. Less important alerts should be filtered out so the prescriber seldom or never sees them.

"If an alert fires, the best time is real time when an order is being written," she explained. "If it fires later or is too wordy to take in easily, it comes off as very annoying."

CDS is based on standard databases, she said, but customized content is more effective. Drug-drug interactions are the single most important alert category. Because potentially severe interactions are the most likely to affect patient care and outcomes, most prescribers only want to see severe alerts.

"Drug interactions are a significant cause for prolonged hospitalization, morbidity, mortality, and increased costs," White explained. "They are predictable and, therefore, preventable, which makes them ideal for clinical decision support."

Limited choices are key

The goal, she said, is to support clinical workflow, not to interrupt it. That means preselected order sentences and order sets that suggest specific and preferred dosing regimens for each medication.

Order sets should automatically eliminate irrelevant choices. A fentanyl patch order should not offer oral or IV dosing choices. If an erroneous administration route is offered, someone will accidentally chose it and generate a new error.

"Too many choices, especially wrong choices, leads to chaos," White said.

Related Videos
fake news misinformation | Image Credit: Bits and Splits - stock.adobe.com
Dr. Charles Lee
© 2024 MJH Life Sciences

All rights reserved.