How one healthcare startup is making data…July 22, 2022 2022-07-22 2:29
How one healthcare startup is making data…
How one healthcare startup is making data…
Healthcare is deeply human, and each successful doctor-patient interaction is unique. Some patients feel the most reassured by a straight-forward physician in a white coat; others respond better to a warm and humorous bedside manner. But if there’s one thing nobody wants their doctor to be, it’s overwhelmed.
In addition to the detrimental effects on the doctors themselves, rampant burnout among healthcare providers puts patients at risk. “It’s almost inevitable that doctors will miss important things,” says Anne Amario, VP of Marketing at Navina. “That includes inevitably missing data that’s hiding in those overstuffed patient records, and missing significant diagnoses, too. That’s where Navina comes in; only AI can make this data work for physicians and not against them.” Founded in 2018, Navina is leveraging the full AWS toolkit to improve the human-to-human interactions at the heart of healthcare. “[The result is] a better physician experience,” says Amario, as well as “better diagnosis and care.”
A 2018 survey found 78 percent of doctors reported feeling burned out. If the workings of the human body weren’t complicated enough, healthcare providers today must navigate a complex terrain of patient data and clinical research while carrying ever-growing patient loads. The same survey found that doctors saw more than 20 patients each day on average. As the pressures of the COVID-19 pandemic have driven healthcare providers across all levels of care out of the industry, the trend toward burnout has only gotten worse.
In an effort to cut administrative costs and make important patient data more accessible to distributed networks of providers, the passage of the Affordable Care Act in 2009 mandated the digitization of patient records. “But having all the information doesn’t mean the information is readily available for physicians to use,” Amario says. Patient records remain complex, unstructured, and unprocessed. Parsing them all is more than most overwhelmed doctors can manage in the scant minutes before each patient meeting.
Navina changes that by providing physicians with the tools to control their data in every interaction with their patient. Using proprietary transformers-based Natural Language Processing (NLP) models, Navina’s AI can analyze and filter this unprecedented volume of available information. Navina integrates the data from electronic medical records (EMRs) and other sources, surfaces what’s important, identifies potential quality and risk gaps requiring attention – and empowers physicians to take action and to document on the spot. It uses the Amazon Simple Storage Service (Amazon S3) to manage millions of patient records. Meanwhile, Amazon Textract and Amazon Comprehend scan clinical documents including hospital and consultation notes, cutting through medical jargon to identify issues that require physicians’ attention, like missing labs, abnormal results, and potentially missed diagnoses.
And, as Amario explains, “Navina is shifting the patient-doctor interaction from reactive to proactive, from “What’s wrong with you today?’ to ‘How can we optimize your health?’” so users feel safer when they have Navina’s support. “It’s like having another set of eyes, or another virtual doctor that has all the time in the world to read through every single document and alert them to what could have clinical significance.” And in a value-based care environment, where physicians are financially reimbursed for their performance and missed diagnoses mean lost revenue as well as poor patient outcomes, Navina helps providers avoid millions of dollars in losses. One physician group that has integrated Navina into their practice has pulled in $45,000 in additional revenue per provider per year, adding up to more than $1 million in revenue they would have otherwise foregone.
Navina’s user base is growing fast, having started off with having started off with hundreds of physicians and now serving over 3,000 users in dozens of clinics throughout the United States eager to put the tool to work. The growth of the value-based care model throughout the healthcare industry means the most exciting developments are likely still to come. The machine learning (ML) infrastructure of AWS allows Navina to accelerate the training process for its proprietary ML and NLP models, so the company can scale quickly and streamline experiments as needed.
Looking towards the future, Amario says, “The irony is that the entire ecosystem is focused on generating more and more data about patients, but in this way everyone is adding to the workload and nobody is focused on how the data will be digested. Navina is staying ahead of the curve, with an eye toward how physicians can make the most of all this data, and creating space for more meaningful patient interactions.”