A recent survey by the American Medical Informatics Association found that excessive clinical documentation is a common problem, causing clinician burnout and even affecting patient care. . It revealed doctors’ and nurses’ frustrations with the burden of charting electronic health records and the time and effort required to complete the necessary documentation.
There is technology that can help. Many health systems are using natural language processing tools, artificial general intelligence, and ambient scribes to help their clinicians with EHR documentation tasks.
But these tools are not plug-and-play. To operate effectively, they need to be customized to the specific needs of the clinical end user.
Dr. Dean Dalili, DeepScribe’s chief medical officer, said customization with note takers around is an essential feature, especially for medical specialties, as the ability to personalize note taking helps Reduces clinician’s automated note editing time.
“There is less time spent on the physical interface” to create formatting and other personal note options, and more time with the patient, he said.
In this Q&A, Dalili also discusses ambient intelligence and its benefits for businesses. This feature compares note taking against coding standards and creates a reporting structure to review the quality of clinical visit notes across points of care.
Ask. How can surrounding clinical documentation improve the care delivery experience and reduce burnout?
ONE. To understand the impact of the surrounding AI literature, consider previous standards. Clinical documentation is done on paper and then transferred to the EHR as part of the [federal] policy.
That transition is beneficial for quality and safety, especially regarding drug safety, and physicians will better understand their patients’ circumstances as they make decisions.
The problem, however, is that EHRs have reduced the ability to deliver care because physicians often interact visually with computers rather than their patients.
EHRs contribute to burnout – truly a phenomenon that causes high levels of mental stress. That’s part of being a clinician, but it’s worse when you’re constantly switching modes to different types of work and you’re translating an encounter into some kind of visual output in a different format. document format via keyboard.
AI clinical documentation allows providers to interact directly with patients and have casual conversations. That conversation becomes the source of information from which the software creates a comprehensive, structured document. In some ways, it’s more comprehensive than when relying on the supplier’s memory.
Technology is always listening and will sometimes catch details that the supplier forgets or doesn’t focus on. Those details are included in clinical documentation, resulting in a better quality patient experience and better quality notes.
Ask. Why are customizable AI scribing tools important for critical care facilities?
ONE. Customization is important for every physician but most important for specialists. As doctors, we fall into a rhythm.
A one-size-fits-all scribble solution that produces a standard structured-only note will not serve most providers. There are nuances to how providers want to capture information – it could be elements in the subjective history portion of the note, portions of the physical exam, or in the assessment and plan sections, where the provider groups parts of the treatment plan with each hospital. Evaluate.
There are also formatting elements of notes that each physician prefers. If you are a geriatrician, you may want to refer to patients as Mr. or Mrs.; If you’re a pediatrician, you won’t want to refer to a young patient that way and may just want to know the name.
Having customization options helps create notes that are more likely to mimic what the provider prefers, is familiar with, or has written in the past. That’s important because when the output matches the supplier’s documentation preferences, they have less editing to do.
Think about the value proposition of any AI note-taking: You don’t have to first spend time on a conversation and then spend more time recording the conversation. But if you still have to edit that note to make it look the way you want it, you still have a lot of work to do. That’s why customization is important for majors.
Specialty-specific workflows are documented differently than a typical primary care visit. As such, there are different areas that need emphasis and detail that vendors want to capture. The key is to establish a format that is not a generic, one-size-fits-all, applicable to some, but rather customizable output and structure so that the AI listens for components of the visit. access dedicated to that specialty.
For example, in oncology, there is often a very long data summary that will identify a patient’s diagnosis and all the data elements used to define their problem. Additionally, specific elements of the plan may be specific to cancer treatments — not just related to medical therapy but also related to social support, nutrition, and others. The note for orthopedics may look completely different and focus on musculoskeletal examination, imaging, etc.
Ask. How is this technology different from other AI note-taking tools available on the market?
ONE. First, we use a unique large language model that includes historical data from clinical encounters coded by live scribes, which helps create structured data elements . We train our LLM – unlike LLMs like ChatGPT4, which are trained on the internet.
If you use medical information to refine your LLM, you will be more likely to receive accurate medical-related output. If you train your LLM across the entire Internet, you’ll get additional noise that can echo into the content.
DeepScribe has the largest source of training data based on user notes used to create highly accurate documentation that helps build trust and adoption, while minimizing the time vendors spend for rework or editing.
The second difference is that the tool offers over 50 different customization elements, allowing vendors to create work that closely resembles what they would create from scratch, across multiple levels of expertise and users.
The third key difference is a new category called ambient intelligence, which functions beyond note-taking.
This is where patient conversations can be applied to any type of structured data, whether it’s coding standards or clinical quality standards. From there, the AI can determine whether the conversation meets that encryption standard.
This intelligence also allows us to create a reporting structure where, across multiple providers, the business can instantly see how physicians are performing. At the point of care, it is the ability to help identify high-value clinical content and also evaluate whether that content is being delivered.
Question: Does DeepScribe integrate into a physician’s existing workflow and technology systems?
ONE. With ambient AI, the workflow is fundamentally different and requires some adjustments – doctors must present findings in ways they may not have done before.
Suppliers can’t just say, “This doesn’t look right.” Instead they have to say ‘Your left knee looks swollen.’” so there is a degree of specificity in the language that providers have to adapt to. This level of detail makes AI listening more powerful.
The level of integration depends on the EHR, and DeepScribe has integrations with Epic, athenahealth, eClinicalWorks, and over a hundred other EHRs, but if a physician wants to connect a proprietary EHR, DeepScribe can also integrate via API.
Andrea Fox is a senior editor at Healthcare IT News.
Email: [email protected]
Healthcare IT News is a publication of HIMSS Media.
