What does it mean for an organization to pursue a data-facilitated journey into value-based care? Bill Gillis, CIO of the Boston-based Beth Israel Deaconess Care Organization, shared his learnings and perspectives on that subject, in the opening keynote presentation of the Health IT Summit in Boston on Tuesday morning, sponsored by Healthcare Informatics. Speaking to an audience gathered at the Courtyard Boston Downtown Hotel, Gillis spoke on the topic “Managing the Risk Madness: The Value-Based Return on Data Quality.”
As the description of the session noted, “Value-based and accountable care organizations [ACOs] rely on accurate data, analysis and reporting to provide the critical information the organization needs to understand how it’s providers are performing to meet risk-based contract agreements. The extreme challenge is how difficult it is to accurately capture and report on the numerous and disparate performance metrics individual to each risk-contract across commercial, Medicare and Medicaid markets. The lack of standardization across payers is further complicated by a lack of standardization across EHR vendor systems and quality measures for the healthcare industry overall. This is layered on top of the foundational challenge that much of the information health care organizations collect and store for data analysis is incomplete or imprecise.”
Speaking of the Beth Israel Deaconess Care Organization, or BIDCO, a joint hospital-physician membership organization that has brought physicians and hospitals together to deliver care to 225,000 covered lives across 300-plus geographic locations and 35.6 million annual patient encounters, Gillis told his audience on Tuesday morning that “When I talk to our peers across the country, sometimes I think BIDCO is unique. And I’m envious of folks on a single Epic instance; but even those people have docs or systems in their communities that are on something different. And sometimes, they just give up. The reality is, you can get there. But this data resides in different systems. We have 46 different EHRs [electronic health records], and 150 different installs. There is data in core EHRs, but also in lab, radiology, and infusion systems. You want to try to get your hands on all that data, and that can be a challenge in and of itself. And the data itself can be very messy.”
Looking at the oceans of data available to use on behalf of accountable care and risk-based contracts, Gillis noted that the landscape of that data is extraordinarily complex—and yes, messy. There are essentially four categories of data available for use in that context, he told his audience: structured, coded data; structured, uncoded data; unstructured, coded data; and unstructured, uncoded data. “Structured, coded data,” he said, “is the really good stuff; but it represents 1 or 2 percent of all the data you’re going to get. Meanwhile, 5 percent of the data is structured and uncoded,” and needs intensive mapping and normalization. “The really dirty stuff is unstructured and uncoded data; and 85 percent of our data falls into that category. And we don’t really have a way to use that data. So we really try to focus on the first three categories.” The bottom line for all those working with the data, he said, is that “The burden of the normalization and validation of that data falls on you.”
What about continuity of care documents, or CCDs? “Will CCDs solve our problems?” Gillis asked. “The reality is that every one of those systems stores the data differently, meaning that their CCD payload is different. So you’ve got to take all the data from those CCDs, and normalize it and validate it”—which presents a major challenge for those working in accountable care and risk-based contracts.
“We’ve learned a lot, and we’ve gotten beaten up a lot,” Gillis said of the experiences that he and his colleagues have had so far on the journey. One of the things that he said he hears all the time is the idea that “It’s just interfaces, right? And I can just connect them and extract the data, right? The reality,” he said, “is that you get into a network with varied systems, vendors, payloads, and transport mechanisms—and it turns out that everything is all over the place. It feels a lot like organized chaos, and it’s very challenging. Even if you’re on a unified platform, it’s challenging,” he said, noting that his core hospital organization has been on a unified platform for several years already. “In the ideal world,” he added, “you’ve got the data points connected, got the pipes connected, and the data’s flowing, and you’re proud. But is that data usable? The reality is that no, it’s not. We pulled all the data in five or six years ago, and pulled it apart; it’s pretty messy stuff. The fact is that so much of the data is uncoded and unstructured, and everyone one of the [EHR] systems stores its data differently.” As a result, he said, “It just takes some time to work through” all of the necessary processes.
Making meaningful connections with physician practices
Get the latest information on Health IT and attend other valuable sessions at this two-day Summit providing healthcare leaders with educational content, insightful debate and dialogue on the future of healthcare and technology.