If you’re a healthcare provider using an ambient AI medical scribe, you’ve experienced a transformation in the way you work. Your encounters have returned to focused conversations with the patient, one-to-one interactions centered on care - not note-taking.
With clear benefits – time savings, more comprehensive notes – the adoption of ambient AI has been relatively rapid. Industry analysts predict that AI medical documentation technology will be in wide use within two to five years.
At DeepScribe, we’re thrilled to be at the forefront of this care delivery evolution. But as we continually build upon the intelligence that produces the industry’s most complete clinical note, we are thinking and developing tools to serve clinicians beyond documentation.
To truly benefit clinicians and healthcare organizations, ambient AI technology should do more than document care. It can – and should – deliver insights. There’s enormous value in tapping into structured data to bring clinicians essential information during the patient conversation.
That’s why we’re proud to introduce DeepScribe Assist, ambient intelligence that brings high-value information to clinicians when it really counts - at the point of care.
AI that keeps clinicians informed
To understand the impact ambient intelligence can have, let’s take a step back to the “old” ways. Healthcare providers have spent years of patient visits with a keyboard and screen in the way. If you use an AI medical scribe for documentation, you’ve effectively removed that distraction from the moment of care. Of course, that’s one of the most powerful values of ambient AI technology, the reason for a more enriching and connected experience between clinicians and patients.
But while providers have relatively unencumbered conversations with patients, there’s a wealth of clinical information out there that can be enormously helpful in prompting the right actions and decisions. How can it be extracted – and how can it show up for the clinician? That's where ambient intelligence comes in – and moves AI technology forward.
Ambient intelligence produces short, relevant prompts for a current patient visit based on key data from past conversations. These recommendations occur in real time – a clinician can see the information before meeting with the patient, during the visit, or post-visit when his or her actions can be assessed and shared as feedback.
For DeepScribe Assist, clinicians see this information on the same DeepScribe app they use to document their visits.
Using DeepScribe Assist for HCC coding intelligence
After speaking with health system leaders, we quickly identified the first use case for DeepScribe Assist: helping organizations succeed with value-based care programs.
Achieving top performance within value-based models is challenging, if not time consuming. To help health systems optimize results, DeepScribe Assist focuses on a crucial aspect of value-based programs, the capture of Hierarchical Condition Categories (HCC) coding.
HCCs reflect a patient’s illness burden and are directly related to the reimbursement an organization receives to care for that patient. For full reimbursement, clinical documentation must detail all of the patient’s conditions via HCC codes. An incomplete record of a patient’s conditions can be a financial loss - each HCC is worth an estimated $2,500 - but can also result in insufficient care.
Before each exam, DeepScribe Assist displays a list of relevant HCC codes for the patient – not just those that need recapturing, but potential suspect codes to be added. This gives the clinician a full picture of the patient's needs before the visit starts. The clinician now has information for which they can take action, addressing the necessary conditions during the conversation.
Of course, our ambient AI takes care of scribing that conversation accurately and in a manner required by billers. DeepScribe Assist can assess whether HCC coding compliance requirements are being met and can offer a prompt for the clinician to adjust the discussion to capture necessary details.
After the exam, DeepScribe Assist provides a summary of how each HCC condition was addressed, using MEAT criteria (Monitor, Evaluate, Assess, Treat). This quick overview helps clinicians identify any areas that may require follow-up, further ensuring comprehensive care and accurate documentation.
We see this as a groundbreaking advance, a way ambient AI can be used in exam rooms and clinics to help healthcare organizations achieve their goals – clinically, operationally, and financially.
Future uses for DeepScribe Assist and ambient intelligence
There are a few reasons we are the only ambient AI company positioned to develop, apply, and refine an ambient intelligence solution. A couple of reasons rest with our technical foundation: our AI platform, Heal AI, is built on the insights of more than five million patient conversations. It also features custom AI models that have been trained to understand the specific clinical needs and complexities of specialty care. This means you're not getting generic AI assistance, but rather insights tailored to your specific field of practice.
Perhaps most important is the partnership approach we take with DeepScribe Assist. Put simply, we develop what groups need to achieve their clinical and financial goals. As Matt Ko, DeepScribe CEO and co-founder says, “We’re working with customers to develop other ambient intelligence uses that serve their most important organizational needs. We can’t wait for what’s next.”
The introduction of ambient intelligence into the exam room marks a significant step forward in the use of AI in healthcare. It's not about replacing your expertise, but about augmenting it with the right information at the right time to make the best decisions for your patients.
DeepScribe Assist is the first step. Our focus on HCC coding demonstrates the potential of ambient intelligence to improve both the quality of care and the efficiency of healthcare operations. As this technology continues to evolve, it promises to transform not just how we help document care, but how you deliver it.
