ArborAdmin

Building a successful clinical registry is a three-step process that involves acquiring data, transforming it with advanced analytics that identify what matters, and informing clinicians and other stakeholders what and how to improve.

Let’s focus on the first step: acquiring data. For a high-achieving registry, you need to collect a large amount of quality data.

Data for your registry can come from a variety of sources, such as medical records and health insurance claims. Regardless of the source, you need to be sure that the data you’re collecting will both enhance your registry and help you meet your registry goals.

With so much data available, it’s easy to feel overwhelmed. As long as you keep your focus on your goals, you’ll be able to compile a healthy dataset that will serve your organization’s purposes successfully.

Let’s explore the various data sources you can use to build your registry.

In this post we will cover:

• Clinical Data
• Patient-Generated Data
• Cost & Utilization Data

• Public Health Data

For each, we’ll define what it is, where and how to acquire it, and address any common challenges or limitations of the data source.

Clinical Data

Clinical data is at the core of many registries. It has a powerful ability to summarize clinical experiences and capture information needed to inform healthcare progress.

What Is Clinical Data?

Clinical data helps establish who a patient is – their demographics, family history, comorbidities, procedure and treatment history, and outcomes. The breadth and depth of clinical data opens up opportunities to advance quality improvement initiatives, research, registry-based studies and virtual trials, and other stakeholder activities.

Sources of Clinical Data

The key source of clinical data resides in a patient’s medical record, which collects nearly all care delivery and treatment decision details, regardless of treatment setting. The record resides in outpatient and inpatient facilities and contains information from various ancillary services, such as chemistry labs, pathology or radiology departments, and pharmacies.

Because a patient’s medical record aims to show as complete a picture of the patient as possible, the data abstracted from those records is often more comprehensive than data from other sources. Additionally, medical records follow patients over time, allowing registries to collect and analyze longitudinal information. Medical record data is therefore extremely valuable to budding registries.

Collecting Electronic Health Record Data for Registries

Medical record data can be highly useful, but gaining effective access to that data isn’t always simple.

For clinical registries, successful medical record data acquisition relies on the combined effectiveness of trusted relationships with the record stewards and technologies used to transmit data, such as electronic health record (EHR) systems.

The vast majority of hospitals in the United States use electronic health records (EHRs). There are many different EHR vendors, which can cause difficulties with data sharing. Given the variety of EHR solutions and the varying policies at different sites, having a flexible data integration approach for your registry is key.

Here at ArborMetrix, we integrate with hospital and clinic EHRs and other data systems with minimal impact to IT and clinical resources. Whether healthcare organizations use cloud-based EHR systems, custom-defined layouts, or anything in between, our registry data platform seamlessly collects the necessary data.

We support industry standards for interoperability, as well as custom data formats. Industry standards such as HL7 FHIR® and C-CDA serve to bridge documentation from disparate EHRs in ways that facilitate data aggregation across populations.

Limitations of EHR Data for Registries

Sometimes, data collected in EHRs lacks the clinical specificity needed for quality improvement and research done through registries.

Many EHRs are not configured to collect the details on the high resource intensive conditions. This leaves the data with gaps that prevent it from being used to make impactful changes in care delivery. Consequently, there is the need to supplement the EHR data with specialized and targeted assets required for real clinical improvement.

Some registries use technology-enhanced clinical abstraction to address this and create a more complete dataset. For example, our electronic case report forms (eCRFs) allow registry participants to enter cases on-site or remotely. Once all the required data have been entered into the system, data analytics and dashboard reports are available immediately.

Leveraging eCRFs to fill any gaps in EHR data creates more complete data set.

Patient-Generated Data

Driving high-value healthcare often requires data that goes beyond an exam room or treatment facility. Your registry can generate many healthcare insights through patient-generated data.

What Is Patient-Generated Health Data?

Patient-generated health data (PGHD) “are health-related data created, recorded, or gathered by or from patients (or family members or other caregivers) to help address a health concern,” according to the National Learning Consortium at HealthIT.gov. [1]

A few examples of patient-generated data include a person’s health and treatment history, symptoms, biometric data, and patient-reported outcome measures.

Sources of Patient-Generated Data

A core source of patient-generated data comes from patient-reported outcomes (PRO). PRO are health data sourced directly from patients, without any interpretation or alteration by a clinician or other individual.

Because iInpatient documentation in the hospital medical record stops when patients are discharged or transferred, and outpatient documentation occurs longitudinally but is sporadic and only captures new data clustered around patient encounters., Aadequately capturing patient data outside of the walls of a hospital or outpatient clinic requires PRO.

Collecting PRO Data

A user-friendly PRO solution encourages patients and caregivers to share details of their experiences, quality of life, and outcomes in ways that extend the period of data acquisition following inpatient care. andIt also fills the gap in patient communication between outpatient encounters.

Important pieces of information that come from PRO include:

• Type, frequency, and severity of symptoms
• Nature and severity of disability
• Impact of condition on daily life

• Patients’ perceptions and feelings about their conditions and/or treatments

Including PRO in their registries can help healthcare organizations track outcomes, support shared decision-making, develop guidelines, inform best practices, and calculate predictive analytics.

Usually, organizations acquire PRO data through surveys and other technology-enabled digital platforms. Other sources of patient data include wearables, population health measures, and patient registries.

Challenges of Collecting PRO Data

A question that sometimes comes up with PRO data is whether it is accurate. Fortunately, research shows a high correlation between patient-reported data and clinically documented chart abstracted data. [2] This means data collected from patients is accurate and trustworthy.

A second potential challenge in collecting PRO data involves patient engagement. Patients who are engaged in their care can share information and communicate more easily with their caregivers, making PRO acquisition much more effective.

Organizations that have successfully engaged patients have employed a few helpful tactics, including:

• Treating patients like consumers, by making surveys and feedback platforms responsive and user-friendly.
• Recognizing the role of technology to help meet patients where they are at with a simple and effective survey experience.
• Delivering a tailored experience so surveys can be customized to different patient populations and flexible and responsive to their unique needs.
• Being creative and compelling by giving patients and their families access to other resources, such as a shared decision-making platform.

Collecting PRO Data from Patient Devices

Other patient-generated data sources include passive data collection from patient devices, such as wearables or smart healthcare electronics. These systems often have data connectivity tools, such as APIs, to security transfer patient datapoints to a registry platform. Real-time, streaming data can further augment information gathered from PRO surveys and clinical data systems, without burdening patients or providers.

Cost & Utilization Data

Many healthcare and clinical registry pursuits are driven by value-based care initiatives — improving outcomes while controlling for costs. These programs require measuring health outcomes against the cost of delivering those outcomes.

What Are Cost & Utilization Data in Healthcare?

Cost and utilization data helps to assess care value for preventing and treating health problems.

Healthcare cost has a specific meaning, but its “interpretation often depends on whose perspective is being considered.” According to the AMA Journal of Ethics, the definition of cost means:

• To providers: the expense incurred to deliver health care services to patients.
• To payers: the amount they pay to providers for services rendered.
• To patients: the amount they pay out-of-pocket for health care services.” [3]

Healthcare utilization “refers to the use of healthcare services. People use healthcare for many reasons including preventing and curing health problems, promoting maintenance of health and well-being, or obtaining information about their health status and prognosis,” according to the Encyclopedia of Behavioral Medicine. [4]

Sources of Cost & Utilization Data

Key sources of cost and utilization data are collected by health insurers, governmental organizations, and public payers. They include claims data regarding patient treatments, as well as public datasets from organizations like the Centers for Medicare and Medicaid Services (CMS) and the Agency for Healthcare Research and Quality (AHRQ).

Payer and regulatory data are especially useful for registries examining the economic aspect of healthcare. Specifically, by including these types of data, registries can more easily perform cost and quality analyses due to the clear records of a patient’s treatment history. Also, especially for government data, the dataset may be larger-scale and include information from a broader patient population.

These data can be acquired from private and public health insurers, and many regulatory datasets are available for public use directly from the sponsoring governmental agencies. For example, we use the Healthcare Cost and Utilization Project (HCUP) dataset from AHRQ and the Limited Data Set from CMS to calibrate propensity-score-based risk adjustment models and to provide external benchmarks for our registry clients such as the Michigan Value Collaborative.

Challenges with Healthcare Cost & Utilization Data

While cost and utilization datasets can be extremely valuable, they also have their limitations.

Dealing with claims data, whether they come from commercial or public payers, can be subject to reversals and adjustments, completion lags, and coding variation and changes that can present considerable challenges when incorporating such data into a new registry.

Overcoming these challenges requires registries to use sophisticated analytic tools and have deep expertise with payer and regulatory data.

For example, we developed and use an elastic episode-based model that lends itself to more appropriate assessments of longitudinal cost. Our model also allows us to more fully assess the outcomes of medical treatments by looking at longitudinal episode cost across a variety of conditions, settings, and care types. This approach captures episodes of care, episodes of illness, and health maintenance episodes to illustrate cost effectiveness across a variety of components. Through this model, we can track care in systematic ways to make comparisons on cost and outcomes.

Public Health Data

There are a variety of factors that significantly influence healthcare outcomes, with clinical care and patient health behavior only representing a portion of the influences. Up to 50% of health outcomes are contributed to community and population health factors. [5]

Consequently, efforts to advance health care requires a full view of all factors that influence health outcomes. This view is critical to understanding the pathways to advance the direction of health care.

Registries can be powerful for providing data to understand the multitude of pathways influencing patient health and using that data to support the development of evidence-based clinical interventions, community guidelines, even advocacy for policies and regulations.

Sources For Public Health Data

There are numerous public health data assets for evaluating community and population health factors. Government data sources are easily available and can layer valuable information on the prevalence and severity of health conditions across different populations in a community. Some sources of public health data include:

• Behavioral Risk Factor Surveillance System (BRFSS): Includes data at the county and census tract-level related to health outcomes, clinical care, and health behavior.
• National Center for Chronic Disease Prevention and Health Promotion’s (NCCDPHP) Chronic Disease and Health Promotion Data and Indicators Open Data Porta: Includes numerous data access tools across a wide range of chronic disease data, risk factor indicators and policy measures.

• National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention – NCHHSTP AtlasPlus: Includes more than 15 years of CDC’s surveillance data on HIV, viral hepatitis, STD, and TB. AtlasPlus also provides access to indicators on social determinants of health (SDOH) allowing users to view social and economic data in conjunction with surveillance data for each disease.
• National Environmental Public Health Tracking Network: The National Environmental Public Health Tracking Network (Tracking Network) brings together health data and environment data from national, state, and city sources.
• CDC National Vital Statistics System (NVSS): Includes health outcome data, such as mortality rates by cause of death at the county level.
• CDC Wide-ranging Online Data for Epidemiologic Research (CDC WONDER): Includes nearly 20 collections of public-use data for U.S. births, death, cancer diagnoses, tuberculosis cases, vaccinations, environmental exposure, population estimates, among other topics.
• USDA Economic Research Service (ERS) – Food Access Research Atlas: offers census-tract-level data on food access that can be downloaded for community planning or research purposes.

Public health data sources add critical information to registry systems to enable registry stakeholders to best serve the needs of the populations they serve. Public health and clinical information can be blended together within a single registry to have a comprehensive, multi-factorial view of health. This creates a platform that can best fulfill the meaningful and diverse healthcare objectives registries pursue.

Set Your Registry Up for Success

Whether you’re building a condition-specific registry for quality improvement or expanding your registry to address new research priorities, you need to be thoughtful about what data you include. Further, you need the tools, expertise, and technology to ensure your registry can achieve your goals.

‍

There are widespread, longstanding, systemic, and well-documented preventable health disparities among racial and ethnic minorities and amongst those in LGBTQIA+ communities. Minority populations experience a multiplicity of burden in regard to their health including poorer overall health statuses, worse outcomes, and disproportionate diagnosis of health conditions due to the intersection of structural inequity.

Barriers to access, differences in clinical need, and differences in patient treatment exacerbate these poorer health outcomes. Given this unfair burden, healthcare stakeholders have an ethical obligation to change healthcare delivery to address health inequity.

The Council of Medical Specialty Society (CMSS) provides a forum to catalyze healthcare improvements across U.S. medical specialty societies. They held the CMSS Leadership Summit on Equity and Antiracism in April 2021 to focus on methods to reduce health disparities, address issues perpetuated through dominant culture dynamics, and improve equity within medical specialties. Key objectives included:

1. Increasing awareness of longstanding, health inequity and health disparities.

2. Identifying and measuring disparities.

3. Forming action to address disparities.

In this post, we share key learnings from the summit and expand on the role of clinical registries in this imperative work.

Increasing Awareness of Health Inequity and Disparities

The initial step to increasing awareness around inequity and the role of bias in health disparities is acknowledging the extent to which these systematic processes harm health outcomes.

These processes permeate healthcare at every level: from the language we use to talk about racism and bias, to algorithms built on racist data that are used to perpetuate myths about genetic and physiologic differences between patients of different races [1], to the exclusion of minorities in medical schools, to the lack of provider diversity, and ultimately in patient care.

This work begins with the way we speak about bias in general, and racism specifically. The language used to describe processes that exclude people who do not conform to dominant sociocultural norms indemnifies the very groups that perpetuate these processes. [2] This camouflages how detrimental bias is to both wealth-building and health outcomes. In changing how we talk about persons that do not hold power and influence in society, firstly we recognize that there is a difference in treatment based on power dynamics and we open the door to changing that dynamic.

Increasing awareness must be a systematic process occurring at every level of healthcare. This begins with recognizing race as a social construct that has physical implications [3] and treating social data with the same gravity as traditional medical data and enacting policy to support the collection and bidirectional transfer of these data. It also includes creating psychologically safe environments that encourage allyship amongst clinicians of dominant sociocultural groups and marginalized people. [4]

To dismantle racism and bias within healthcare, we must prioritize diverse perspectives amongst clinicians and the healthcare delivery system as a whole and normalize diversity, equity and inclusion initiatives. [5] To increase awareness surrounding health inequity, we are educating and training our team, focusing on ways to integrate the use of social data in registries, and supporting our clients in doing the same. We are also developing partnerships with and learning from experts who can help accelerate our work in measuring and improving health equity so that we can in turn support organizations like CMSS.

Identifying and Measuring Healthcare Disparities

Advancing health equity work in the pursuit of reducing and eliminating health disparities involves detection and measurement in addition to understanding the determinants. Clinical data registries can be instrumental in these tasks by assessing the causes and impact of health disparities on patient outcomes.

According to the Centers for Disease Control and Prevention (CDC), only 10 percent of factors affecting premature death are related to clinical care, and 30 percent of factors relate to genetics. This means that 60 percent of factors impacting premature death are based on a combination of social/environmental factors (20 percent) and behavior (40 percent). Ensuring the delivery of high-quality, patient-centered care requires understanding both the needs and circumstances of the populations served. Registries can collect and analyze patient, clinical, social, and geographic data to provide a more complete view of the type of advantages or disadvantages certain patients have in attempting to achieve optimal health. [6]

Fully identifying and measuring disparities and their root causes necessitates data that goes beyond clinical data elements. This includes the prioritization and comprehensive collection of social data such as:

• ICD-10-CM codes (Z55-Z65) that detail patient social risk factors
• Utilizing public health datasets such as the American Community Survey, Behavioral Risk Factor Surveillance System and CDC National Vital Statistics System
• Patient-reported outcomes data

Integrating comprehensive social data elements in the United States Core Data for Interoperability (USCDI) and enforcing mandatory collection within electronic health records (EHRs) are two steps that make up a meaningful process in measuring the extent and impact of social determinants of health.

Forming Action to Address Healthcare Disparities

Data and measurement are only as valuable as it is usable and actionable. Advancing health equity begins with understanding the sources of disparities, but it can only be achieved through effective action.

Once disparities are assessed and the causes are understood, impactful interventions can be developed using data-driven approaches.

However, implementing informed interventions is only the first step. Next, interventions must be evaluated for effectiveness, changed to address problems as necessary, and finally, best practices must be disseminated to scale.

Leveraging Clinical Registries to Improve Health Equity

To promote optimal patient care, it is our ethical obligation to understand the prevalence and severity of social and health inequities, utilize comprehensive data collection and analysis, and take informed action to end them. We must evaluate and implement interventions that effectively improve outcomes for the most vulnerable populations, which in turn improves conditions for all populations.

Clinical data registries are valuable tools for pursuing these efforts. Registries can collect a high-volume of data from different sources and deliver high-quality, data-driven insights. Reduction and eventually elimination of health disparities require data to detect and measure impact, identify the complex and interconnected determinants, and inform interventions and best practices.

We’re interested in hearing how your organization is addressing these important issues and what challenges or questions you are finding. If you have insights to share or questions to ask, reach out to me at cprice@arbormetrix.com.

I’d like to thank my colleague, Bradley Moore, M.P.H., Director, Registry Solutions, for his significant contributions to this post.