Can melding the best of electronic health records (EHRs) and artificial intelligence foster greater empathy among healthcare providers? Basil Hayek, Director of Business Consulting, Sapient Health, and Paul Penta, Manager of Business Consulting, Sapient Health, examine that possibility in this guest post.
EHRs, artificial intelligence, and empathy: If this were a Sesame Street segment, it would be easy to pick the one that is not like the others. But could combining the first two actually enable greater empathy?
First, let's set some context. Industry-wide, there is a high degree of dissatisfaction amongst physicians with EHRs. This can result from the increased clerical burden, poor user interfaces, and feeling of cookbook medicine. These challenges add to the burden already faced by physicians, and contribute to burnout, sub-optimal prescribing and referral behaviors, and erosion of clinician empathy.
As empathy diminishes, so do outcomes. Studies across multiple conditions, including diabetes, cancer, and the common cold, found evidence supporting this hypothesis. In addition, a broad evidence review saw a consistent positive association between patient experience, patient safety and clinical effectiveness. As clinician empathy can dramatically influence a patient’s experience, it can be surmised that empathy is a contributing factor.
There are various approaches to addressing EHR challenges. Healthcare organizations are deploying scribes, tablets and optimization programs. Legislative and policy changes are in the works. Although not enacted before the end of the congressional session, the Senate Committee on Health, Education, Labor, and Pensions Committee introduced a bill last year to help improve EHR usability. In addition, the Agency for Healthcare Research and Quality within the Department of Health & Human Services has called for certification requirements on EHR usability.
Ultimately, healthcare provider satisfaction with EHRs will improve through these tactics. That's not to say that EHRs cannot move from simply meeting a HCP’s expectations to exceeding them. This is where artificial intelligence (AI) will play a role.
When AI is mentioned, the first thing that comes to mind is probably IBM's Watson. Watson, which gained fame for winning Jeopardy against two former champions in 2011, has dramatically evolved its cognitive capability and reach to make an impact in various industries, including healthcare. Healthcare providers can use Watson to analyze medical records, assist in diagnosis, and help find evidence-based treatments, and its capabilities continue to grow. These developments are exciting, but only hint at what is achievable, which includes helping to achieve the Quadruple Aim of an improved patient experience, improved population health, reduced costs, and an improved clinician experience.
AI can contribute in realizing the Quadruple Aim due to its ability to efficiently analyze large volumes of data, discover patterns, and make logical inferences. The potential population health and cost implications resulting from AI are fairly self-evident. What is intriguing is how AI can play a role in helping improve the provider and patient experience. How this could come together is better told through two scenarios.
Consider 47-year-old Gary, recently diagnosed with type 2 diabetes. He is due for a follow-up with his doctor to review his treatment after completing blood work. In the first scenario, Gary is trying to navigate his diagnosis in the current state environment.
Because of a lab location that requires him to drive instead of taking public transit, Gary misses three scheduled lab appointments and is forced to reschedule his follow-up. When he finally completes his labs and meets with his doctor, she seems hurried and spends most of the time looking at her laptop as she updates Gary’s chart. Although she notes his A1Cs have risen, she opts to continue the same regimen until his next appointment. Gary leaves feeling uncertain about the effectiveness of his medication, and has doubts on whether the side effects of heartburn and indigestion are worth it.
Now, let’s look at a version of this same narrative where AI enables a better all-around experience.
After Gary misses his first lab appointment, the AI-enhanced EHR offers to reschedule at a location one stop away from his house. He attends that appointment and keeps his follow-up with his doctor. Prior to this appointment, his doctor reviews an AI-generated clinical summary, which highlights key factors to consider for Gary’s treatment and confirms that she has reviewed the latest research relevant to Gary’s condition and history.
She greets Gary when he arrives and asks how he’s doing with the medication. When he mentions heartburn and indigestion, she acknowledges they’re common side effects, and recommends that he be diligent about taking it with food and using an antacid until the side effects diminish.
She turns Gary’s attention to a large wall-mounted screen showing a patient-optimized view of his health record. She uses a tablet as a second screen to direct the conversation via a physician view. Gary sees his blood glucose trends and notices that the side effects he just mentioned are in his record. As they discuss additional medication to help control Gary’s rising A1Cs, the doctor asks Gary to confirm the accuracy of the displayed list of medications and supplements. Gary mentions he has also started taking low dose aspirin.
Shortly after he says this, aspirin appears on the screen. The physician view on her tablet alerts her of new research indicating an interaction between aspirin and a candidate medication. With this information, she recommends an alternative combination drug, and Gary walks out with a new prescription and confidence in managing his diabetes.
In the second scenario, AI enabled the following technologies and associated benefits to provide an alternate and improved experience:
- Prescriptive analytics based on clinical and socio-demographic perspectives of EHR data: offloads intent from the patient and reduce barriers to care;
- Context-aware clinical natural language processing: offloads data entry from the physician, and allows the patient to more naturally participate in the treatment conversation;
- Cognitive computing to assess medical evidence: allows physicians to more easily review information relevant to a specific patient.
These enablers are neither new nor novel. However, combined they provide unobtrusive interventions that reduce the clerical and cognitive burden on physicians and provide improved opportunities for patient engagement. With time to think and a renewed focus on the person sitting in front of them, physicians can return to an empathy-driven encounter, and everyone wins.
About the Authors:
Basil Hayek is responsible for digital strategy and delivery for Sapient Health. He supports a broad portfolio of clients, with a focus on health plans, pharmacy, and retail health. He gets excited about bringing together his technology, data, and product background to drive engagement and deliver business results for companies and better health outcomes for individuals. Basil graduated from Cornell University with a BS in Computer Science.
As a Manager of Business Consulting at Sapient Health, Paul Penta draws on his experience building technology for patients in a clinical chronic care environment to enable digital change in healthcare organizations. With a focus on digital and technology strategy, Paul always keeps the patient at the center of the experience. Often taking on a cross-functional role, Paul excels in leading the strategic merger of process and product to achieve impactful metrics. Paul received an MBA from Boston University.
HIN Disclaimer: The opinions, representations and statements made within this guest article are those of the author and not of the Healthcare Intelligence Network as a whole. Any copyright remains with the author and any liability with regard to infringement of intellectual property rights remain with them. The company accepts no liability for any errors, omissions or representations.