How AI Can Augment Rural & Community ER Teams

Overcoming Resource Challenges: AI’s Role in Improving ER Efficiency in Rural & Small Hospitals

How AI is Transforming Emergency Care in Rural and Community Hospitals given the constrained resources many rural and small hospitals face in terms of both capital and clinicians, trying to implement and maintain new technological solutions to improve care and efficiency may seem out of reach. Moreover, after many of the efficiencies promised by the shift to electronic medical records failed to materialize, many physicians and administrators may be skeptical of the next generation of technologies designed to enhance both the patient and provider care experience. 

Of course, artificial intelligence-powered tools are that next generation. Though AI isn’t a cure-all, when implemented correctly, it can play a positive role in the emergency room, especially in ERs in resource-limited areas.

“It looks like we’ve reached an inflection point with AI. The amount of processing ability AI systems have are doubling at a pretty high rate,” says

TECHealth Chief Operations Officer Eric Wilke, MD.  “The AI we have today isn’t the same AI we’ll have six months or a year from now.”

Enhancing ER Performance: How AI Tools are Making a Difference

AI can significantly improve patient outcomes and lessen risk, while also addressing two of the largest physician pain points—the electronic medical record and administrative burden. Here are a few ways it’s already proving useful in the ER environment:

Avoiding misses: From a clinical standpoint, adding an AI tool for chart review to find gaps in care, red flags or suggest diagnoses that aren’t readily available or frequently seen in smaller ERs can support physician decision-making. 

Removing cognitive burden on physicians: Among the most helpful AI tools is ambient listening technology such as TECdoc Scribe. Ambient listening allows ER physicians to focus on the patient during an encounter rather than on documenting. 

“With TECdoc Scribe, for example, you can go into the room, hit the button, start the recording and talk to the patient and it builds out the text for the chart for you,” says Dr. Wilke. This allows physicians to concentrate more on what they need to do for the patient, rather than trying to remember everything to dictate later.

Reducing admin time: Ambient listening tools can populate some of the chart while also offering a recording a physician can refer back to later. This helps speed physician charting, which in an ER, may not occur until the end of a shift. “If it’s busy in the ER, a physician may have to remember the details of an encounter with a patient they saw at 9 a.m. when they’re doing charting at 7 p.m.” 

Some AI tools can further charting efficiency by more quickly (and perhaps accurately) entering lab results into the EMR. “With TECdoc Scribe, you can take a picture of lab results, and it will read those characters and insert them into the text of the chart for you.” 

Improving performance: With EMRs and administrative burden being major drivers of physician burnout, AI tools can be a win for physician satisfaction. By identifying potential gaps in care, they can also serve as an important asset for risk reduction and better outcomes. 

Additionally, AI ambient listening technologies like TECdoc Scribe will soon be able to analyze the sentiment of conversations between physicians and patients. Is the physician being compassionate and reassuring during an encounter or short and hurried? This analysis can help medical directors address issues with a physician’s communications skills and improve the patient experience. 

ISupporting Better Decision-Making In ERs With AI

Two of the biggest concerns about AI are data security and accuracy. Despite AI sometimes being described in the broader culture as ‘the wild west,’ many companies developing AI solutions for healthcare have already thought through these issues according to Dr. Wilke.

“You don’t want a physician using their personal ChatGPT and entering patient information, but if you’re wanting to use AI as an organization, many of the AI companies have business associate agreements with HIPAA protections for protected health information that meet federal requirements.” 

In other words, working directly with the companies to establish an account with PHI guardrails in place is often doable. Other options are to host AI through internal IT capabilities or contracting with a third-party to do so. The best approach depends on the AI functionality desired and resources available.  

“Of course, you’ll want to do a risk assessment prior to implementing any AI and review your systems in place for data protection,” Dr. Wilke advises. 

As far as accuracy is concerned, AI is certainly not perfect, but neither are people. In some areas, AI may already outperform physicians in the aggregate (e.g., AI being 97% accurate vs. a person being 91% accurate in certain scenarios). How an AI tool is set-up also is key. 

“Different AIs, depending on how they’re set, can introduce more randomness or less randomness,” explains Dr. Wilke. “In medicine you want less randomness, so the same input will give the same output every time for reliable diagnostic help.”

ERs should pilot the AI solution with a range of test cases first, where the correct decision is known to ensure the AI reaches the same conclusion. 

Affordable & Efficient: Getting Started With AI In Your ER