Patient Registries Turn Knowledge into Outcomes Improvements
Patient registries that make data scalable and usable at the health system level remove important barriers to improving care and advancing research. Registries integrate with enterprise data warehouses (EDWs) and EHRs to make information easily and quickly accessible by non-IT professionals across a health system. This puts knowledge in the hands of the right people at the right time, enabling effective intervention on the patient level and outcomes improvement on the population level.
Scalable Knowledge: A Healthcare Must-Have
Scalable (standardized) clinical knowledge is a priority in the EHR and healthcare EDW space. It is expensive and difficult to build IT systems with the knowledge necessary to help providers improve outcomes by tracking and managing the right patients at the right time. In the area of patient registries, not all health systems and IT vendors are approaching standardization in patient registries effectively. As a result, they have minimized access to actionable data that could support better patient care.
Effective registries, on the other hand, are diverse and agile tools that are standardized enough to reuse across an organization for multiple related purposes. They save cost by reusing existing criteria for targeting patient populations that warrant investigation and intervention, instead of paying experts to determine who to include.
The content should be maintainable and upgradeable without significant investments from what are often limited clinical and professional IT resources. This article describes the qualities of optimized patient registry utilization and how it can enable and expedite better care for specific patient populations.
Understanding Patient Registries
In the context of clinical outcomes improvement, registries are a flexible way of accessing information to meet clinical needs. They are comprised of patient populations based on a group of different filters (rulesets) applied across an organization’s patients to narrow them down to an intervention or research group. Registries give the nontechnical user (e.g., clinician or researcher) access to a broad array of information from the EDW/EHR, which it can boil down to a registry of action—a group of patients with whom they intend to follow up and intervene.
The registry of action may be in a clinical context, such as a chronic disease population (e.g., follow-up with a diabetic population to make sure they’re getting foot care and eye exams and taking prescribed medications). It can also be in a research context, such as enrollment and follow-up on a cohort for an approved study or to identify a group for study.
To create a registry, users define rulesets—the criteria, or filters, that define the patient population in question. The rulesets define which patients belong in the registry. Once run through an organization’s system, they create a list of patients who meet the criteria; that list is the registry. The ruleset and registry process can be compared to membership to a club or sports team. For example, in a high school football team, the membership criteria, like rulesets, determine who’s eligible (e.g., students in grades nine through 12), and the eventual list of who’s on the team is like a registry. Like patient population rulesets, the football team membership criteria are shareable from one school or club to another.
The logic of rulesets can range from simple to complex. Ideally, a ruleset builder provides meaningful feedback to the user as they are assembling the ruleset. As a user is creating filters, they should know how that filter is affecting the population – is a filter overly restrictive or not restrictive enough? Developing an easy, intuitive interface for creating registry rulesets is challenging and most of the tools currently on the market reflect this challenge. An easy-to-use interface that provides meaningful feedback during the build process has been elusive to date. Nonetheless, there are options, both open source and commercial, for building registry population definitions.
Putting Knowledge in Front of the Right People, at the Right Time
In today’s data-rich healthcare environment of EDWs and EHRs, collecting data may not be a problem, but getting to that data—especially in a timely and cost-effective manner—is often a workflow challenge. Historically, health systems filed records away in filing cabinets. When a clinician or researcher needed data, the organization had to pay for expensive manual chart reviews. The process cost time and money; the clinician or researcher had to wait for needed information (sometimes so long that they missed deadlines and had to abandon the project), while their department footed a significant bill before it even got to the heart of the matter (e.g., a treatment decision or research initiative).
Sophisticated, flexible patient registries make data highly accessible to clinicians, quality improvement teams, and researchers, so they can turn that knowledge into action within a required timeframe. This removes barriers to informed decision making and research. Registries based on reusable rulesets make searching the virtual file cabinet of electronic data (EDWs and EHRs) timely and worthwhile. A few of the potential areas that registries can support are:
- Prioritizing opportunities with a deeper clinical version of Key Process Analysis (KPA).
- Pre-IRB exploration of potential de-identified research cohorts.
- Quickly excluding potential intervention areas without significant patients (often as important as choosing promising areas to focus on).
Patient Registries in Action for Research and Healthcare Improvement
When, for example, a clinical researcher wants to pursue a grant for research in a specific population (e.g., patients with diabetes or hypertension), they first need to determine if their facility has the needed population. With a registry interface (such as Informatics for Integrating Biology and the Bedside [i2b2]), the researcher can drag and drop data from the EDW or EHR to see how many people who visited their facility the in past six months had a diagnosis of a particular heart condition. The researcher can then compare their system’s information with definitions from the grant organization to determine if they have the needed population.
With a patient registry, a researcher can gather data directly, refining their investigation at once and generating the information needed to immediately respond to the grant organization. The alternative is to submit their request to an analyst group, which does not have the same level of research expertise to refine the request and requires additional response time. The advantage of easily accessible knowledge also holds true from the clinician perspective, as clinicians can view outcomes across a certain population personally and immediately (versus requesting information from the IT department) and make timely informed treatment decisions.
Patient Registries Offer Next-Level Opportunity Analysis
By enabling more focused cohorts, patient registries help clinicians and researchers more precisely target opportunity areas. Health systems can get started quickly with standard measure/registry definitions that already exist from government/quality reporting agencies. Standard definitions form a basis for comparing and sharing information across organizations. Where necessary, it is also a great starting point for an organization to refine the national standards to adapt to local needs without the high cost of starting from scratch. After all, medicine is an art as well as a science, and things are seldom black and white. So, while standardization is necessary to share knowledge, it’s done on a sliding scale to allow strategic customization.
Sharing Rulesets Across Health Systems Multiplies the Knowledgebase
When health systems can conveniently share content, experts multiply their access to knowledge. One organization may already have a subspecialty (where it has specialists or more time to investigate a subset of patients), in which it’s accumulated significant knowledge. Another health system may not have the same experience or depth of expertise, but want to pursue interventions for that population.
If the people with expertise define the rulesets and make those shareable between organizations via a patient registry definition from a common vendor, the other group can skip the time-consuming step of defining the rulesets. The less specialized group can then use the same clinical knowledge as a starting point for their initiatives. This is particularly useful when national standards are focused on regulatory reporting instead of clinical operations, which often requires a different perspective when defining relevant registry and measure content. Eventually it is hoped that cross-vendor standardization through initiatives like HL7 FHIR, to which Health Catalyst® is aligning, will be widely adopted to enable broad access to population information as well as for individual patients.
Vendors who partner directly in the outcomes improvement process have an advantage in building shareable registries because they build their rulesets specifically for provider engagement. This level of content is more of a real reflection of the clinical setting compared with national content (e.g., CMS reporting) that might be focused on licensure or reporting requirements.
It should be noted there will always be barriers and challenges to achieving fully shareable, reusable, and scalable content. National registries are often defined based on diagnostic codes, which are quite standardized and required for certification. But as clinicians and researchers go deeper (e.g., exclusion criteria and more specific criteria for each population), they generally find less consistent terminology and there will be a need to build localized, custom definitions.
The solution to the standardization–customization problem is a pragmatic approach, in which the vendor distributes a standardized starter registry and allows users to customize it. Every EDW has different building blocks, making a Late-Binding™ registry the best way to account for individuality between health systems.
Advantages of the Late-Binding Patient Registry
The late-binding, customizable registry supports both wide adoption and knowledge sharing across organizations by offering a range of critical functions:
- National predefined starter content.
- Broad, approachable access to data for refinement of the definition (what they need, when they need it).
- Aggregate patient population-based (nonpatient specific) analysis.
- Quick decision making and more frequent access to data so that clinicians and researchers are informed by data, versus working in a data vacuum.
Better Data Access for Better Decision Making
The central goal of the patient registry is to grant more users greater accessibility to understand and find the right information they need to improve clinical and operational outcomes within their organization. More and more often, non-technical clinicians and researchers will need to make more informed choices to manage the complex needs of patient populations in addition to treating immediate patient needs at the point-of-care. Technical users will benefit from the scalability and reusability of predefined registry content that will assist them in initiating and maintaining new analytic projects.
The registry takes data to the population level—facilitating transparency, openness, and approachability (this helps overcome common EDW user barriers). With a more approachable registry, more people are mining, exploring, and experiencing data for themselves. This helps clinical users ask more informed questions and act on the information at the point of need, while freeing up the experienced data analysts to more quickly focus on the right issues.
Would you like to learn more about this topic? Here are some articles we suggest:
- Why Precise, Tailored Patient Registries Lead to Cost-Effective Care Management Programs
- The Real Opportunity of Precision Medicine and How to Not Miss Out
- Population Registries Kick-start Rapid-cycle Clinical Process Improvement
- Why Most Analytic Applications Will Never Be Able to Significantly Improve Healthcare Outcomes
- Quality Improvement in Healthcare: Where Is the Best Place to Start?