Managing the EHR inbox takes a system. Here’s how to set one up.

AMA News Wire

Managing the EHR inbox takes a system. Here’s how to set one up.

Aug 1, 2024

The EHR inbox has become a primary means of communication in health care, but the tedious task of routing these messages disproportionally falls on physicians who too often become the primary triage point in inbox routing—and this tedious work quickly becomes a major contributor to physician burnout.

Major problems include too many messages that are clogging up the EHR inbox that do not need to be there. And, of those messages that should be in the inbox, the majority are ones that do not need to be seen by a physician.

Coordinated, systematic efforts that include continuous measurement and auditing of inbox volume can help physicians reduce the EHR inbox burden for both physicians and care team members, according to an AMA STEPS Forward^® toolkit that describes a systematic approach to reducing your EHR inbox burden.

Making technology work for physicians

The AMA is working to make sure technology is an asset to physicians—not a burden.

Learn More

From AI implementation to EHR adoption and usability, the AMA is making technology work for physicians, ensuring that it is an asset to physicians—not a burden.

The toolkit recommends taking the steps below to reengineer the inbox and reduce low-value and unnecessary work, and then using team-based care to boost efficiency and reduce a primary source of physician burnout.

Develop an EHR inbox task force. This should include a C-suite champion and a physician leader. The panel could also include representatives from stakeholder groups such as practicing physicians, IT and clinical leaders, EHR vendors and a process-improvement specialist. Financial investment may be required to ensure that the task force has adequate time and resources.

Measure using audit-log data. This will help you understand the greatest opportunities for improvement and assess the impact of interventions. Messages can be classified as results, prescription authorizations, care-team messages, shared charts, patient calls and other categories.

“Assessing time in conjunction with volume is essential to get the most accurate picture of your current EHR inbox state,” the toolkit says, adding that “low-volume, high-effort messages should be prioritized during improvement efforts.”

Content variation among individual messages is so great that “no simple or singular tactics could address all message types,” the toolkit says. For example, previous research has found that found results folders often contain a mix of routine preventive normal lab results and critically abnormal diagnostic results.

Adopt a strategic framework to “eliminate, automate, delegate, collaborate.” The toolkit recommends setting goals for each of these strategies and then developing specific actions designed to achieve them.

Outlined is the experience at one health system that established a “reinventing the inbox” initiative complete with vision and mission statements for each of the four steps.

Members save on travel & entertainment

AMA members save up to 25% on car rental base rates at participating Hertz locations; get discounts on travel and cruises.

Learn More

Start with a “great purge.” It is often helpful to begin efforts with a “grand gesture” to convince people change is possible. At one health system, the toolkit says, older EHR inboxes never expired, leaving some physician inboxes with 15,000 messages—many of those dating back several years. Over a three-week period, the health system deleted 12 million messages, clearing them at night to avoid EHR performance slowdowns.

Eliminate low-value and preventable messages. The toolkit says that one health system discovered that CC’d charts from consults, urgent care and notes from colleagues accounted for 13% of total inbox volume, and it was difficult to determine which ones had clinical importance.

After consulting with physicians, the governance committee shut off automatic sharing of charts and created practice agreements on what to send and other details. This resulted in a 40% drop in monthly CC’d charts in primary care physicians’ inboxes.

Automate protocols and pathways for routine tasks. The toolkit recommends removing “human involvement whenever possible.” This includes automatic release of normal lab and imaging results straight to the patient portal and bypassing the physician inbox. Abnormal results or those with a high dependency on clinical context should still be routed to the ordering care team.

The toolkit recommends using self-sorting options such as letting patients check if their message pertains to medical advice, test results, prescription refills, making or canceling an appointment or other common requests.

Delegate message handling. Team-based EHR-inbox management shifts inbox-management responsibilities away from the physician.

Messages can be grouped, with different teams responsible for different buckets and a primary manager assigned to each one. This approach helps avoid duplicative work while ensuring that messages don’t go unanswered because staff is unsure who addresses concerns. Only 5% to 10% of messages involve complex matters that require the physicians’ attention, according to the toolkit.

Standardizing tasks such as prior authorizations and short-term disability form requests and using designated teams is recommended.

Cover the inbox during physicians’ time off. Checking the inbox during time off contributes to physician burnout. Practices need to develop a systematic approach to EHR inbox coverage during physician time off, with the goal that physicians leave with an empty inbox and return to one that is mostly empty—without logging in during their time off.

Having the physician serve as gatekeeper of the inbox results in more burnout and ambiguous responsibility. Care team members should come to view the EHR inbox as “a collection of clinical responsibilities that should be addressed by the entire clinical team for the sake of efficiency as well as good patient care,” the toolkit says.