Patient Registries: Putting Disease Management on the Map

The percentage of the U.S. patient population with multiple, complicated and chronic medical conditions is expected to rise, at the same time many healthcare practitioners are reaching retirement age. New ways of managing chronic conditions are at the heart of both quality improvement and cost containment efforts nationwide.

Chronic disease management is getting a lot of attention from all corners of healthcare. And for good reason: Chronic conditions such as diabetes, congestive heart failure, chronic obstructive pulmonary disease (COPD) and others are expected to cost billions of dollars annually. For example, the total cost of diabetes in the U.S. in 2007 was $174 billion—$116 billion for direct medical costs and $58 billion for indirect costs including disability, work loss and premature mortality, according to the American Diabetes Association.  

Federal efforts to address chronic conditions include the Stage 1 criteria for meaningful use of electronic health records (EHRs), which call for the use of clinical decision support (CDS) applications to facilitate disease management. Stage 2 criteria are expected to further emphasize CDS and disease management via the wider use of EHRs, according to the Centers for Medicare and Medicaid Services.

Identifying these patients and collecting condition-specific information is the first care step toward determining best care practices and comparing outcomes, both for individual organizations and for nationwide improvements in care. Enhanced patient registries are already enabling these improvements. These registries consist of familiar tools: a database of patient information, a data model for populating and organizing information, and software that allows users to mine the data and create reports from the information within the registry.

Data from multiple sources

Enhanced patient registries pull information directly from EHRs and other sources into a comprehensive database, says John Hutton, MD, vice president and director of Biomedical Informatics at Cincinnati Children’s Hospital Medical Center. In addition to having the ability to extract data needed for patient management directly from the EHR, enhanced registries can accommodate input from sites where EHRs are not in use, via direct upload or file transfer, Hutton says.

Cincinnati Children’s is using a $12 million grant from the Agency for Healthcare Research and Quality (AHRQ) to create a disease registry that will be tested by the ImproveCareNow network, which concentrates on inflammatory bowel disease, a condition affecting nearly 100,000 children in the U.S. The registry is expected to come online in stages over the next three years, says Hutton, principle investigator for the project.

The aim of the enhanced registry project is “to develop better ways, from technological point of view, of accruing and managing data related to specific types of patients,” he says.

“A registry fundamentally is a way of collecting information necessary to study, manage or take care of patients who share some characteristic. For example, you could have a registry for patients with diabetes to study how to improve quality of care, how to compare treatments, to look for ways to improve care,” he says.

“We were developing the registry [through] new IT techniques … to accumulate data from EHRs that are relevant to a subset of patients, to keep track of it, to display it to the healthcare providers, to look at quality across sites, compare quality and treatments,” Hutton says. “The registry part has been done here at Children’s hospital, [and we] are working to improve it even more and implement it in the ImproveCareNow network.”

Most of the registry technology was developed in-house, and was built on Informatics for Integrating Biology & the Bedside (i2b2), a scalable framework developed at Partners HealthCare and supported by the National Institutes of Health. “i2b2 is an open-source platform we modified, and is one of the multiple pieces of the improved registry,” says Hutton.  

“We can get information from our EHR; we do that here. Because we have developed the technologies, the intent is to test that and expand it to additional sites. It’s designed to be interoperable with Epic or Cerner or other [EHR] systems.”

Privacy and security are handled by individual sites, he says. Participating sites’ Internal Review Boards (IRBs) reviewed the registry proposal for collection of information and the nature of the permission the patient is giving. For facilities that have EHRs, “the intent is to keep the data at the local site so it doesn’t get moved centrally, and then to query it as de-identified data at a distance,” Hutton says.

“Other sites don’t yet have an EHR, or the capability to implement that form of a registry. ... They have the option of entering data directly into a secure data warehouse at our institution that is designed to segregate their data and keep them secure, but we would [have] capabilities to de-identify them or manipulate them in accordance with the restrictions from the site of origin.” Data collected are generally demographic information about the patient, he says.

Data that are entered into the ImproveCareNow national database are almost completely de-identified, says Richard Colletti, MD, at Vermont Children’s Hospital at Fletcher Allen Health Care at the University of Vermont, and director of ImproveCareNow, based in Burlington, Vt. “However, it’s important that data not be made completely anonymous, because when sites get monthly reports, we use the reports to help manage individual patients.”

The enhanced registry will have all regular demographics: age and gender, along with diagnoses, therapies, ICD-9 codes, CPD codes, pharmacy information, problem lists, according to Hutton. “By contrast, in a conventional patient registry you can’t add additional data. You might want to add genetic markers, go in and pull data directly from EHRs.”

Cincinnati Children’s is planning and implementing a large number of registries to improve the care of children with, for example, liver disease, arthritis, and certain kinds of esophagitis, according to Hutton.

The enhanced registry’s modular design is not intended to disease-specific, “and it’s not specific in terms of the type of EHR your organization has,” says Hutton. “It has interfaces to pull data from a variety of types of records and upload data in different ways. Because it’s modular, you may change the types of data; you may want to upload different kinds of information. The information needed differs from disease to disease, but basic types of the registry technologies remain the same across registries, he says.

“We have a plan to develop new registries for a whole variety of diseases, but we also recognize that some disease management is not so complicated—there are registry capabilities built into some of the EHRs, for example, that are not sophisticated enough for some diseases, but they work fine for others. You have to choose your registry depending on the management needs of a particular type of patient.”  

Reining in diabetes

As part of its effort to adopt a certified Patient-Centered Medical Home (PCMH) care model, Crystal Run Health, a multispecialty physician group in upstate New York with approximately 1,300 employees, implemented NextGen Ambulatory EHR and NextGen Practice Management software organizationwide. This enables about 200 providers to keep track of patients within the system and monitor multiple chronic conditions, says Greg Spencer, MD, chief medical officer of Crystal Run, based in Middletown, N.Y.   

“About three years ago, we [looked closely] at the population of patients who we had and identified the ‘big three’ medical conditions we see,” says Spencer. “We identified hypertension, hyperlipidemia and diabetes. These are high-risk, high-dollar issues. For patients with diabetes, for example, we report various measures quarterly, including hemoglobin A1C, micro-albumin determination, also subsets of diabetics who have hemoglobin levels greater than 9, so are considered poorly controlled.

Disease management data are based out of the NextGen EHR, “so people don’t have to change their workflow. We do provider report cards, currently [using] Microsoft Reporting services and Crystal Reports, augmented with Excel spreadsheets for clarity and portability. Reports are prepared through care management, and go up to the clinical divisions for providers and also through committee structure with the physician quality committee and bigger supercommittee, which is Quality and Patient Safety. Those are reviewed on quarterly basis.

“We have a data warehouse, data are extracted to Microsoft reporting services. We have hypertension and diabetes registries. We’re moving to more information cubes [which will enable] providers [to] slice and dice data as they see fit. It’s all on a reporting platform so it’s not impacting performance, which is crucial.”

Templates within the EHR make various data available at the point of care, “so when something needs to be ordered, you’ll be able to think of it and do it,” Spencer says. “All of this is tied in with policies and procedures. Our current effort is to have standing orders for various conditions so physician doesn’t have to think to order it.  

Crystal Run’s registries are populated only with its patients’ data, and the system doesn’t report data out to external registries, says Spencer. Some of the high-risk conditions that area hospitals are monitoring, such as congestive heart failure, aren’t emphasized at Crystal Run because they are considered to be high-risk are also high-risk for us, but [because] the quantity they see [is so much larger]… it’s more important from a numbers standpoint for them—congestive heart failure being the big one,” says Spencer. “We don’t have a ton of congestive heart failure patients in our practice, and as part of the big picture for us, it’s not as important as it is to the hospital. I think as things move forward, those capabilities will be overlapping with the hospital component.”

“We see that as a piece of this bigger care management mechanism,” says Spencer. “The goal is to keep all information in one place.”

From clinical decisions to clinical trials

Nationally, getting all that information into “one place” may still be well beyond the horizon, but doing so will move disease management efforts forward, says Aaron Seib, CEO of the National eHealth Collaborative, a Washington, D.C.-based public-private organization advocating adoption of a secure, interoperable, nationwide health information system.

“We’ve been doing disease management for 15 to 20 years now, using healthcare claims data to do patient identification and do some intervention as far as case management is concerned,” says Seib. In the past, the disease management industry used claims data to identify patients for inclusion in disease management programs, but “there are a lot of known limitations to the quality of the data from claims, and because of the paucity of the data, there ends up being a lot of false-positives and missed cases,” he says.

“What I hope we’re starting to see is more and more of the specialty associations trying to see how they can be involved, because that’s where they have a lot of clinical expertise to apply. In the future, as the NHIN [Nationwide Healthcare Information Network framework] matures, I believe we will see clinical experts developing patient identification protocols that can work with the clinical data along with the provider and independent of the financial data processes. That’s one of the things I hope that the new framework, this overall architecture, will enable, because it will benefit everyone involved,” says Seib.

When the NHIN framework is in place—in 2015 and beyond—healthcare organizations be able to not only deploy disease management efforts and enhanced patient registries, “but start getting to near-real-time outcomes management … to the point where we’re able to say, [for example], white males age 40 to 45 who have congestive heart failure respond better to this [treatment] compared to that,” says Seib.

The NHIN framework also will enable standardized content to be pulled directly from the EHR, so clinicians are not only identifying the right patients for certain treatments, but also are finding patients who might be at higher risk for complications, Seib says. “We can do population science [with] unsupervised algorithms if the data are properly de-identified, and see what unusual things are happening. We can get beyond where we are today, with a lot of controlled [Stage 4 Clinical Trials research] on a very specific population, and be able to explode it out to a whole population, to where it really matters.”

Enhanced Registry Helps Improve Care for Pediatric Patients

An enhanced patient directory is being put to the test by members of the ImproveCareNow network, says Richard Colletti, MD, at Vermont Children’s Hospital at Fletcher Allen Health Care, University of Vermont, in Burlington, Vt. Colletti is director of ImproveCareNow, a collaborative dedicated to improving care for pediatric inflammatory bowel disease patients.   The network, which started in 2007 with eight pediatric gastroenterology practices, has grown to 30 practices. “Everyone was committed to using quality improvement methods to improve the way we were providing care to our patients,” says Colletti.   The collaborating physicians came up with best practices, which were called Model IBD Care, Colletti says. “From that, we derived 15 measures that we use to track, measure and evaluate our performance for patients with inflammatory bowel disease,” which includes Crohn’s disease and ulcerative colitis. “The patient care encounter doesn’t change,” he says. “What’s different is that during or at the end of each visit, data are collected on about 80 different data elements and these are entered into a centralized national database in Chapel Hill, N.C. We use web-based data entry, but it doesn’t just sit there: It’s used to manage care, so every month, and soon to be twice a month, data are extracted and go to participating ImproveCareNow practices.” “Each practice gets a report on how they’re doing and how the collaborative as a whole is doing. We set our standards, we measure how we’re doing, and each site gets results, sees how they’re doing, and then implements changes to the way they deliver care to try to get even better results,” Colletti says. Results are improving: “When we first started [the ImproveCareNow network], 50 percent of patients were well or in remission. That’s increased to 70 percent,” he says. Although all but one ImproveCareNow site have an electronic medical record of some type, data collected at most others in Vermont are collected on paper, Colletti says. “The new [AHRQ grant, announced in October] is going to help us demonstrate that it’s possible to get all sites in a collaborative like this to enter data electronically and then transfer that data to our database, so they no longer requires manual data entry,” he says. The five-year goal for the project is to get 80 sites into the network. About 3,000 patients are currently enrolled and that number is expected to double by the end of 2011, according to Colletti. “We have data on 15,000 visits in the database, and that will almost double as well by the end of next year,” he says. The enhanced patient registry model would be “a natural relationship” for other disease management initiatives, says Colletti. “The process we’re using is definitely applicable to other disorders. We’ve been chosen to be the pilot project [for an enhanced registry developed by the Cincinnati Children’s Hospital Medical Center], but the idea is that it could be something that could be applied to diabetes, congenital heart disease and others.”