Using AI to provide personalized healthcare: an interview with Arta Bakshandeh

Written by Arta Bakshandeh (Alignment Health)

In this interview with Arta Bakshandeh (Alignment Health, CA, USA), we discuss using AI to provide personalized healthcare and improve patient outcomes. Dr Bakshandeh describes some of the most successful real life examples of AI providing personalized healthcare to patients in his role as Alignment Health’s Chief Medical Informatics Officer, and what the future holds for AI in personalized medicine. 

What are some of the most successful examples of AI providing personalized healthcare to patients, to date?

When I was Alignment’s Senior Medical Director, a member called into our 24/7 concierge care line with dangerously low blood sugar levels. The call was patched to me, and I was able to click into AVA to see that the patient lived in a low-income area with no nearby access to care or family. I guided him to adjust his insulin pump and asked him to eat something — except there was no food in his fridge. I ordered the gentleman a large pizza. By 9 a.m. the next day, Alignment’s care team was in the patient’s home, and by the end of the day, he was set up with regular deliveries from Meals on Wheels. Thanks to our AI-powered clinical care model, a $20 pizza prevented a potential $2,000 ER cost and, more importantly, more serious health consequences for the man.

This “pizza prescription” is one example of what AI technology can help care teams do. All the data within AVA is near-real-time and actionable. We collect thousands of data points to paint a holistic, accurate picture of a member’s health, from lab reports and whether they have picked up prescription medications on time to detecting the likelihood of hospitalization or predicting and potentially preventing future incidents, like a fall. These early warning signs are critical before they turn into a true emergency. By leveraging the data points, we can be proactive and provide personalized care to our members.

In your opinion, how could AI improve patient outcomes by providing personalized healthcare?

What makes AI technology powerful is what it empowers care teams to do. AVA is ten times more efficient because it gives providers insights and access to the complete picture of a patient’s health at any time. We know who has multiple chronic conditions, is frail or needs medication or social support. This 360-view enables care teams to reach out proactively to the most vulnerable members with timely, coordinated care. We also can predict the risk for hospitalization, readmission and disease propensity before early warning signs turn into a true emergency — and our model typically is 20%– 25% more accurate than traditional methods. AI increasingly will streamline and improve personalized care, save providers and payers money and time and enable payers to reinvest these savings to enrich member benefits.

Are there any ways in which AI could negatively impact patients in this area of healthcare?

Human doctors should and will always have a critical role in health care. That is because rich, insightful data and information can only be effective or impactful when paired with the expertise and experience of a high-quality provider. AI does what it is trained to do, but data scientists must consider the potential of bias in data and its potential for negative effects.

As the use of AI-powered diagnostics and predictive analytics accelerates, we cannot diminish the value of human compassion and empathy. High-tech should never replace high-touch, quality care.

In your opinion, what does the future hold for AI in personalized medicine?

We are at a critical inflection point as technology advances at incredible speed. AI will continue to alleviate administrative burdens on doctors and staff, reducing burnout, improving retention and giving them valuable time back to devote to more patients in need.

Time also will create a more intelligent environment. We will soon be able to connect to more data sources, getting vitals by looking at a screen and through augmented reality stethoscopes. The health care industry has constantly evolved and adapted to meet the changing needs of patients and communities, and that will never stop.

Interviewee profile: 

I began my career as a hospitalist and quickly became a Medical Director, building care models for patients with Medicare and Medicaid. I joined the Alignment Health (CA, USA) team in 2014 to help develop Alignment’s clinical model and clinical command center. I now serve as Alignment Health’s Chief Medical Informatics Officer, working with data scientists to develop and implement Alignment’s clinical data analytics platforms and home monitoring programs for patients with chronic illnesses.

We began developing AVA®, the company’s integrated AI-technology platform, to improve patient experience and outcomes. AVA is our health care ecosystem through a unified data layer that powers over 35 modules and applications. My priority is to work with the end users in developing machine learning models — more than 170 so far — that will help solve real problems and change how we care for patients. For example, our general admission model prioritizes outreach and enrolls patients in our Care Anywhere program, which provides care at home and 24/7 access to clinicians who know the patient’s health journey and history. Also, I ensure our team focuses on rooting out bias in the data and evaluating correlation vs. causality in our models while providing our end users with machine-agnostic explanations to help guide intervention.

Outside of Alignment, I am an Assistant Clinical Professor of Internal Medicine at the University of Southern California’s Keck School of Medicine (CA, USA) and co-CEO of the Global Physicians Corps, a non-governmental organization providing sustainable care in Tanzania and Ethiopia. In addition, I continue to work on the front lines treating patients in the hospital and virtual settings.

Disclaimers:

The opinions expressed in this feature are those of the interviewee/author and do not necessarily reflect the views of Future Medicine AI Hub or Future Science Group.