Case Study: Improving Medical Message Intake with Augmented Intelligence for Your Operators


By Ravi K. Raheja, MD

How accurate are the medical messages that your nonclinical operators send to your providers? According to Gilman and Bedigian, LLS, a law firm that specializes in medical malpractice, one of the biggest factors that leads to a lawsuit is when operators don’t send critical messages to doctors immediately. In fact, we have found that out of all emergency messages, about 30 percent do not accurately relay the situation for the provider to react. (Also see this white paper on the accuracy of medical messages.)

So how do you ensure that nurses and providers are receiving appropriate information? Below we discuss an easy-to-use tool that helps train operators so that all medical message intake verifies potential emergencies related to each patient complaint. Best of all, this solution only increases call time by an average of 30 seconds. 

Consider this example of a patient message: a patient calls and reports they are feeling dizzy, experiencing both a headache and what they think is a sinus infection. At first glance, this may not be serious. To an uncertified operator, this wouldn’t sound like an emergency. However, after using our augmented intelligence tool, MedMessage Assist, the message adds that the patient is feeling weakness in their arm, leg, or face. 

This information, coupled with the reported headache, automatically reprioritizes the patient’s request. The triage nurse calls the patient back within a minute and ends up advising them through triage protocols to go to the emergency department immediately.

The bottom line: the difference between a nonurgent and emergent call can be a single overlooked symptom.

When we refer to accuracy, we aren’t just talking about how well an operator records what a patient tells them, we’re talking about whether the operator knows to ask additional questions based on those symptoms. This is because patients and operators aren’t medically certified, so it’s easy for both of them to overlook potentially harmful health conditions that their initial symptoms represent. 

This also means that if a message doesn’t accurately reflect the urgency of the patient’s symptom, the patient might end up waiting too long to receive a callback. Unnecessarily long wait times for patient care then leads to poor health outcomes.

So how can augmented intelligence improve medical message intake?

Analyze Data Recorded by Your Operators

When you implement MedMessage Assist (MMA), its software will analyze the text that your operators are typing into their online forms while they’re listening to patient callers. Then, when it detects symptoms that might indicate more severe medical conditions, it prompts operators to ask qualifying follow-up questions. Those answers will improve the specificity of each message and allow providers to establish a proper sense of urgency for processing patients’ requests.

How big of a difference that could really mean?

By the Numbers

When reviewing 2,661 report tickets for patient complaints submitted to triage nurses, MedMessage Assist generated additional information for 726 of them, or roughly 27 percent.

We also learned that about a third of all emergency messages submitted were initially insufficient. MMA helped clarify the messages that were emergencies, and in about 2 percent of the cases, the symptoms were serious enough to require 911 intervention.

Using MMA adds little time to any patient call: roughly 24 seconds when the original complaint isn’t changed, and 40 seconds when it is. That means each call requires less than a minute to ensure the accuracy of patient symptoms so that they receive the appropriate care advice.

As operators gain more experience with MMA, we also expect to see them anticipate which questions to ask and use MMA to compare their work. Implementing MMA has led to accuracy scores of over 99 percent, in some cases vastly improving the message intake that some providers previously reported.

Additional Benefits

Message accuracy has a positive domino effect on several additional aspects of the patient experience.

Improving the patient intake process leads to faster dispositions and greater safety for those who may need assistance sooner than scheduled appointments. In turn, operators become more efficient, which shields providers from malpractice liability.

Beyond these benefits, MMA can work with existing software, so you don’t have to worry about implementing an entirely new system or synchronizing your data with it. There’s also a minimal learning curve for MMA, as the system is designed to be self-teaching. Finally, it meets all requirements necessary to be HIPAA compliant.

Improve Medical Message Intake with a Free Trial of MMA

We’re confident that MedMessage Assist and its augmented intelligence will transform the way that offices manage its patients’ calls. 


Ravi K. Raheja, MD is the CTO and Medical Director of the TriageLogic Group. Founded in 2007, TriageLogic is a URAC accredited, physician-lead provider of high-quality telehealth services, remote patient monitoring, nurse triage, triage education, and software for telephone medicine. Their comprehensive solutions include integrated mobile access and two-way video capability. The TriageLogic group serves over 9,000 physicians and covers over twenty-five million lives nationwide.

Are you ready to improve medical message intake and your patient health outcomes? Email us at to schedule a 15-minute demo. You may also qualify for a free trial of MedMessage Assist.