By Sindhu Kutty for Forbes
As providers move toward increasing virtual care options across the care continuum, the use of AI in telehealth to enable physicians to make real-time, rich, data-driven decisions is a major factor in creating a better patient journey and better health outcomes. According to a study by MIT, 75% of healthcare institutions that implemented AI acknowledged an augmented ability to treat illnesses, and 4 in 5 said it proactively helped avert workplace burnout. With Covid-19 putting an increasing strain on both areas (volume of clinical information and associated patients and the increased workload for clinicians), AI in telehealth is a powerful approach for the future of care delivery.
I advise hospital executives on healthcare and health IT strategies, and the top two concerns I hear around AI are the perception that AI can replace physicians (it can’t) and the complexity of utilizing consumer-generated data in preparation of either virtual or in-person visits (for example, data from smartwatches that gather blood pressure and heart rate information). One of the primary goals for effective patient care is preventing hospitalization and the associated healthcare costs, so proactive implementation of treatment options is top of mind.
So, what are some ways in which AI and telehealth can partner to achieve these outcomes?
AI can augment physicians in making medical decisions (and I emphasize augment, not replace, here intentionally). For example, instead of a physician relying on two or three pieces of medical information (such as medical history, an exam and a lab test), AI can scour big datasets across thousands of patients with similar illness profiles from monitoring devices or medical telemetry products (including treatment protocols, side effects, etc.) to produce algorithmic patterns that can suggest potential next steps to the physician. The physician can then leverage this information to enhance their ability to treat illnesses.
Another way that AI can assist is with clinical information access. With the rise in synchronous (live, two-way audiovisual interactions) telehealth sessions due to Covid-19, physicians have faced several challenges. For example, accessing lab results and other EHR content while on a virtual call with patients was an administrative nightmare for providers both during the session (with the need to access different systems for clinical information) and after (recording the visit and treatment provided). So how can we equip the provider better while not compromising the patient’s experience? Several platforms can push real-time clinical information into virtual platforms using a configuration tool for seamless integration. These platforms typically have an AI aspect, which can deliver relevant data from disparate EHR systems across multiple care settings (for example, labs, X-rays) on a real-time basis. AI can also assist providers by making administrative tasks relating to medical records less onerous through voice-controlled tools based on natural language processing technology. The visit note can be automatically captured and transcribed into the electronic health record.
I mentioned consumer-generated data being a big area in terms of potential for proactive, early risk reduction. Telehealth applications can allow clinicians to monitor ECG, heartbeat, blood pressure, temperature and other vital signs remotely. This combined with AI can allow the patient and physician to be alerted about potential health conditions that are tailored to the unique patient’s needs through predictive analytics and setting monitoring thresholds. This combination can enhance the quality of care provided, the patient’s experiences and health outcomes, and the physicians’ experiences (less data fatigue and more proactive information).
Technology advances in AI, such as voice recognition, have led to the emergence of chatbots or conversational agents. Apple’s Siri, Microsoft’s Cortana, Amazon’s Alexa and Google Home have made this a household technology. Motivational messages, appointment or health reminders, health information and symptom checking, and assistance for the elderly in their homes are all healthcare use cases that are in play today. Conversational agents today offer asynchronous interactions that supplement traditional care delivery. As the comfort with this technology increases and patient safety concerns are addressed, the complexity of the use cases linked to medical history can be explored further to extend care delivery versus just supplementing it.
The implementation of AI in telehealth settings is gaining momentum. The NextDREAM Consortium Group conducted a study about the efficacy of AI in remote diabetes care. The primary finding was that consistent insulin adjustments transmitted remotely, which use the trialed automated AI system, can be as effective as expert physician dose adjustments. The automated AI-based system can be used by physicians and specialists for decision support. The University of San Francisco’s Center for Digital Health Innovation is testing AI that can read X-rays as an early warning detection system for pneumothorax.
The main AI use cases in telehealth include information analysis and collaboration, remote patient monitoring, and intelligent diagnostics and assistance. The power of AI can be leveraged to augment physicians in their abilities to diagnose and treat patients, mitigate their own burnout and enhance the entire spectrum of the patient journey. Due to the ongoing public health crisis, the focus on AI and telehealth continues to be a strong push for healthcare executives as they look to stay competitive by streamlining clinicians’ workflows and unlocking predictive potential through the analysis of patient data.