Modernizing health care in a consumer-based economy

Dr. Roy Beveridge, Humana’s chief medical officer, recently posted this blog entry on LinkedIn.

 

As a physician, I am always amazed at the amount of data that is collected on us – something like 2.5 quintillion bytes of data every day.

Some industries are better at tapping into this invisible web of data to create experiences that delight and satisfy. Take the retail sector as an example.

A retail company can tell from your smart watch that you’re exercising, so it may send you ads for a nearby yoga studio or recommend a documentary on marathon running. That same company could also recommend a new shoe because it knows the average life of a running shoe and how far you run every day. And if you don’t buy shoes on its website, you could start receiving ads for knee braces or Tylenol.

But what if a company used data to predict a person’s likelihood of comorbid conditions, like heart disease and diabetes, and sent information to prevent the diseases from developing or progressing – like healthy cooking options, chair aerobics, sleep apnea machines or videos on meditation?

How can we take this consumer-centric approach and apply it to health care so we anticipate what people need before they need it?

How health care integrates data and analytics today

Take Gino, a 75-year-old man who has diabetes, like 25 percent of people over the age of 65. He had a heart attack five years ago and now has congestive heart failure (CHF), which is the medical condition causing the highest burden on our health care system in terms of resources and cost.

He’s a widower who lives alone but has a daughter who lives a few hours away with her family. Gino just spent a week in the hospital because his CHF was out of control, and he developed pneumonia.

Now back at home, Gino is taking five maintenance prescriptions; but he’s unaware of three new ones because they were sent to his pharmacy electronically. Seventy percent of patients leave the hospital uncertain of their medicines. Even when medicine reconciliation is done in the hospital, 40 percent of patients still have at least one medicine discrepancy, which doubles their chances of returning to the hospital within 30 days.

Gino developed a pressure wound on his heel while in the hospital, and he’s unsure how to care for it. Pressure wounds are one of the most common preventable conditions, affecting more than 1 million people annually, and they increase the cost of care by more than $10,000 per patient per year.

Gino’s primary care physician (PCP) doesn’t know about his hospital stay, and Gino did not mention it when he scheduled an appointment. This unfortunately is the situation 30 percent of the time, and when PCPs are unaware of a hospital stay, patients are twice as likely to report having problems. Gino also has limited fresh food in his refrigerator, like one in eight Americans who are food insecure, and that increases his risk of mortality and isolation.

When Gino’s daughter calls him, he tells her things are good because he doesn’t want to be a burden. Fast-forward one week, and Gino is at his PCP’s office. He is confused and has an infection in his foot, hypotension, a fever, and hyperglycemia. Gino is readmitted to the hospital.

At the hospital, Gino is told he needs emergency surgery that might lead to an amputation. He’s in septic shock, and surgery brings high risk for complications or death. Sadly, one in 20 diabetics endures an amputation because of a foot wound.

All along the way, data is being collected on Gino – by his physicians, his pharmacy, the hospital, diagnostics, labs, his health plan, even his daughter – but it’s of little use to Gino because there is no inter-connectivity and no real-time access to the data.

Let’s see how data, analytics and technology could have helped Gino.

This time, within 24 hours of being discharged, Gino’s care coordinator calls him because he was identified through predictive analytics as being at high risk of returning to the hospital because of his diabetes and congestive heart failure. Twenty-three percent of people with diabetes return to the hospital within 30 days of discharge.

His care coordinator tells Gino he has three new prescriptions and has the pharmacy deliver them to his house. She does a medicine reconciliation to make sure Gino is taking his medicines correctly.

The coordinator sends the reconciliation to Gino’s PCP and asks the hospital to send its records to the PCP. The coordinator changes Gino’s PCP appointment to bring him in earlier – within 4 days vs. a week – because predictive analytics say the ideal follow-up time after a hospitalization is 3-5 days after discharge. The PCP orders home care because she has predictive modeling that shows Gino at high risk of rehospitalization, so she wants him to have a face-to-face interaction within 72 hours.

The next morning, the home health nurse arrives with a mobile clinical workstation containing Gino’s history. She teaches wound care, orders supplies, and does a physical exam using an AI virtual assistant to record her findings. With a voice command, she reports Gino’s vital signs to his PCP, and the PCP conducts a telemedicine visit with Gino using the home health nurse’s mobile workstation. The PCP discusses the pneumonia, glucose control and the wound, and how they’re all connected.

The home health nurse queries Gino’s smart fridge and smart pantry and automates Meals on Wheels delivery. She has predictive modeling that tells her people who are food insecure have a 50 percent higher incidence of diabetes. She sets up Gino with a wearable device and Bluetooth scale and connects it to Gino’s central communications hub to monitor his glucose and weight. The data is sent to Gino’s PCP in real-time, with notifications and alarms for irregular values. The nurse uses Gino’s smart speaker to set daily reminders for Gino to check his blood sugar and weight.

On the third day, the care coordinator checks in virtually with Gino to schedule transportation to his PCP’s office.

The care coordinator uses predictive analytics to determine Gino’s risk of loneliness. She then pulls up her data visualization map, types in Gino’s address, and connects him with the YMCA and Silver Sneakers to improve his social interactions.

On the fourth day, Gino arrives at his PCP appointment with his glucose and blood pressure controlled. His weight and fluid are in check because he’s taking the right medicines and has the right food in his fridge. His wound is healing, and he feels supported at home, where 89% of people older than 50 want to be.

Let’s build a health system designed around the consumer.

It’s this type of coordination — through data/analytics, predictive modeling, the PCP, care coordination, home health services, and in-home technology – that can improve Gino’s conditions. And because of the secure collection, storage and sharing of this data, Gino can give access to the people who need it in real time – even his daughter.

So, with these connections, is it possible to prevent people like Gino from being hospitalized in the first place? I’d like to think the answer is: yes.

Leave a Reply

Your email address will not be published. Required fields are marked *

You might also like