Companies are hard at work redefining their relationships with customers in the digital economy -- and so are healthcare systems.
Boston Children's Hospital is among them, investing resources into how it can improve the "patient journey." To do that, the hospital is attempting to level data silos as well as build new processes and partnerships to improve the patient experience both within and outside its four walls.
"Our thesis is that patients really expect the same service that consumers expect when they interact with the digital world," said John Brownstein, chief innovation officer at Boston Children's and a professor of biomedical informatics at Harvard Medical School.
In CIO lingo, Brownstein is talking about digital transformation, where data is not just a byproduct of treating patients, but a prime driver in making the patient experience more efficient and in building patient-facing applications that could deepen the hospital-patient relationship. The transformation at Children's is driven by big data, artificial intelligence and the entrepreneurial push to reimagine what a healthcare system looks like by being more accessible to patients at the click of a button -- or at the utterance of a voice command.
Underutilized historical data
Indeed, the backbone of the patient journey is data, a surplus of which exists at Boston Children's Hospital, Brownstein said at the recent Harvard Institute for Applied Computational Science's annual symposium. But "we're not necessarily leveraging data that we collect across the house in ways to really optimize that care. So our goal is to really think about where are places across the house where we could do a much better job," explained Brownstein, a speaker at the symposium, where the theme focused on AI, big data and healthcare.
One area of focus is how to exploit rich repositories of historical data with artificial intelligence "to make better decisions going forward," he said. In radiology, for example, the 30,000 to 50,000 images taken on a daily basis, which include extensive, meticulously labeled examples of rare conditions, provided a ripe source of underutilized data. Partnering with GE, the hospital has started to automate the process of finding and flagging irregularities in images that require deeper analysis. Brownstein said early results of the automated process are "looking incredibly promising."
Brownstein's team also focuses on combining real-time data from the intensive care unit or the emergency room with machine learning to triage patients, simplify complex capacity-planning processes, and predict length of stay. When the ER hospital staff has to determine if a person should be admitted, it's a process that can take hours, according to Brownstein. If a patient is admitted, the hospital staff then has a logistical question to answer: What bed does the patient get? "So why not start to understand how likely that patient is going to be admitted well before the decision is normally made?" Brownstein said.
Taking information such as the patient's presenting symptoms and demographic data, Brownstein's team built a tool that aims to predict the likelihood of admittance. "That can save huge amounts of dollars and also optimize the patient experience," he said. Moreover, this isn't a theoretical tool that the innovation team is piloting to improve the patient journey. Data on the probability of a patient being admitted is fed into the nursing staff's dashboard and used for planning purposes.
The digital ecosystem
But Brownstein is not only focused on how to optimize internal processes. He is also trying to optimize external processes -- a passion of his that extends beyond his innovation officer title to being a public health advocate. How does a patient -- any patient -- access information online? Or get treatment insights from the web? Or make it to their appointments? Technology can vastly improve these aspects of the patient journey, if hospitals can figure out how to use them.
Case in point: An ongoing innovation project at Boston Children's is Thermia, a digital support tool for patients and families dealing with fevers. "We built this tool that gives recommendations back to patients about the fevers that their child or family members are experiencing," Brownstein said. More importantly -- and more critical to digital transformation success stories -- the website is also a place where Boston Children's is gathering new kinds of data. In this case, it's collecting data on vaccination status, location of the patient and symptoms, and this information can be used for deeper analysis or to help track potential outbreaks.
Improving processes like these often means tapping into the larger digital ecosystem. Indeed, collaborating with outside partners is critical to innovation in healthcare and for improving the patient journey, according to Brownstein. "I think part of the issue with many hospital systems is that we start to look inside for solutions and don't realize there are huge amounts of dollars being poured out to the outside world in new domains," he said. "When we think about new tools, we're not trying to reinvent the wheel."
Indeed, Boston Children's Hospital recently piloted a nonemergency ride-share service for patients with Circulation, a Boston-based startup co-founded by Brownstein. The HIPAA-compliant app integrates the Uber API with a patient's electronic medical record, enabling healthcare systems to book transportation for patients and replacing paper taxi vouchers that are hard to track.
"Providing a $10 Uber ride or Lyft ride makes more sense than a missed appointment, which would cost hundreds of dollars," he said.
Plus, Brownstein's team is also investigating new user interfaces such as instant messaging and voice assistants. "We've been very interested in the opportunity around voice and healthcare," he said. In 2016, Boston Children's launched the first Amazon healthcare skill with KidsMD, an app that provides basic health advice to parents.
The data barrier
Brownstein's team has its share of challenges. Like in any enterprise, data sets don't exist in one collective place; they're siloed, sometimes hard to get at, and there aren't a lot of incentives to do things differently. "Especially when you talk to electronic medical record companies where, really, their incentive is to lock down a system in an existing setting and not really start to open up the data," Brownstein said.
Plus, there is an existential problem for healthcare systems, in general -- and it's getting in the way of innovation, according to Brownstein. He used the weather as an analogy. Anyone with a smartphone can know what the temperature is at the touch of a screen -- or by asking their digital assistants. That's not because app developers care more about the weather than healthcare. Nor is it from a lack of data.
It's because the infrastructure for accessing weather data through APIs from sources like the National Oceanic and Atmospheric Administration exist. "We miss those types of concepts in healthcare," he said. "We don't have underlying data sets that are available, we don't have an API to that information, no underlying suite of apps that allow us to tie into these sets."
A patient-centered information commons, the idea for which is the basis for the National Institute of Health-funded Big Data to Knowledge initiative, could begin to tie together data sets across the healthcare industry, Brownstein said.