Healthcare organizations can realize the enhanced potential of electronic health record (EHR) technology by using it as a population health management tool.
In the past decade, many large-scale EHR implementations have led to drops in revenue and operating income, and reductions in physician productivity have been one factor in these declines. Faced with negative ROI on a sizable investment, many organizations have responded by attempting to rebuild productivity through targeted initiatives. Common efforts include provider training, EHR system improvements, and workflow tweaks.
However, this single-focus response presents three problems. First, experience has shown that even the strongest process-improvement efforts never fully regain previous levels of production. Second, the continued focus on workflow change runs the risk of exacerbating provider frustrations with technology. Third, although productivity clearly remains important to revenue, the transition to value-based contracts is altering the basic landscape of revenue generation and patient care management.
For these reasons, healthcare leaders need to think in terms of the EHR’s impact on delivering high-quality outcomes, managing the medical spend, and ensuring efficient medical utilization. The industry must begin to think of an EHR as a tool for managing the health of a patient population, rather than as a technology for documenting a patient encounter.
Such an approach to realizing the value of an EHR involves a complex process that requires new structures and capabilities. Financial and clinical leaders can facilitate the process by helping their organizations grow along three vectors: building a data platform, developing an information framework, and evolving new care models.
Build the Data Platform
Under the productivity model, the value of an EHR is its ability to capture patient encounter data that can then be used to generate a charge. Under the new value model, an EHR is used to manage the health of patient populations through intelligence and insights gained from claims, clinical, and social-determinants-of-health (SDOH) data from across the continuum that are brought into the EHR for providers to use at the point of care.
Assembling the data platform for a high-value EHR is an evolution. Healthcare organizations that have done it successfully have constructed their platform one layer at a time, using the three tiers discussed below.
Claims data. The first tier comprises health plan claims data for a managed population across the entire continuum of care. The value of these data is in providing a comprehensive picture of the health conditions within an organization’s population. This information allows a healthcare organization to begin generating risk scores and segmenting its population based on risk levels, which helps identify potential patient issues. The organization can then begin to design programs and interventions for managing high-risk patients and preventing rising-risk patients from entering the high-risk group. Through full claims data provided by insurers, an organization can measure cost and resource-utilization efficiency opportunities that can be brought back into the EHR for follow-up care.
Clinical data. The second tier consists of clinical data, including data from ambulatory and inpatient EHRs, pharmacy data, clinical laboratory data, and other information sources. Layering this clinical information into the data platform enables the organization to move beyond what is happening in the population to why it is happening. For example, an organization might use claims data to identify its rising-risk diabetic population. It could then use physician EHR and lab data to identify specific care management issues and specific actions that could be taken at the point of care to modify patient risk.
An important part of the second tier is using the clinical data within the EHR to calculate standard quality measures and aligning these measures to the performance measures of the organization’s payer contracts. These measures allow an organization to provide real-time performance dashboards at the point of care, enabling providers and the care team to manage performance and fill identified care gaps.
Interoperability is essential in this tier. EHRs need to capture, exchange, and present data in a way that can drive action.
SDOH data. The third tier of this platform is SDOH data, including data on genetics, behavior, environment, and social circumstances. These data could include information about income, housing status, transportation access, and other issues that partially predict costs and outcomes in a patient population. Some advanced healthcare organizations are beginning to use SDOH data to proactively manage patient cohorts by creating models that predict disease prevalence and total spending in rising-risk populations.
Develop the Performance Analytics Framework
The clinical data repository described above is essentially a part of the overall data management platform integrating the claims and SDOH data to manage and monitor performance. The depository’s function is to ingest, normalize, and analyze the data from across the continuum to derive intelligence. This intelligence is then used to define a performance analytics framework, which consists of performance domains, each of which encompasses multiple measures. One key organizing principle is quality. However, organizations should be aware that clinical quality is not the only focus of an effective analytic framework of patient care. Healthcare leaders should think in terms of performance. The goal is to develop a system of integrated clinical insights focused on value-driven performance measures not just for quality and outcomes, but also for care-gap management, cost efficiency, resource utilization, patient access, patient engagement, provider satisfaction, and other aspects of care value.
The key is to look at value-based contracts as a series of use cases and problem statements. For example, consider a health system that has entered a value-based contract for managing its diabetic population. One of the key elements of the contract is cost reduction for this patient cohort. The corresponding question framing the problem is, “Can we reduce the cost of care for our diabetic population?”
The next step is to determine what information is required to “solve” this problem. First, the organization will need a measure such as per-member-per-month (PMPM) costs for the diabetic population. Second, the organization will need data points to generate the measure. Key elements could include data on physician services, tests, drugs, and procedures. These data will come from inpatient and ambulatory EHRs, as well as from billing systems.
Essentially, the process of creating an information framework is identifying problems in value-based care and then working backward to the measures and data needed to solve those problems. This process should be repeated for all the dimensions of a value-based contract, as shown in the exhibit below.
As a whole, the process helps provider organizations identify the relatively small number of measures and data elements needed to demonstrate performance under a value-based contract.
After the measures and data framework have been established, the next step is to modify the organization’s EHRs to capture all required information.
Returning to the example of a diabetes management program, one of the initiative’s care quality measures is the percentage of patients who receive an annual wellness check. Upon reviewing EHR design, program leaders may find that wellness checks are captured in the ambulatory record, but only in the progress notes. Because these data are unstructured, it will be impossible to extract them in an automated fashion. The solution is to reconfigure the EHR template to include a mandatory yes/no field to capture wellness check information as a discrete data point.
Some process redesign may be required to optimize clinical workflows for the digital environment. In addition, it may make sense at this point to invest in clinical documentation improvement (CDI) to ensure codes are captured for the care that is provided and documented. However, this approach to building an analytic framework will keep process change and clinician retraining to the absolute minimum.
Evolve the Care Model
The ultimate purpose of a data platform and analytic performance framework is to generate process-improvement insights that align with a care model that identifies the major areas of care redesign for value-based care: defining populations, stratifying risk, providing access, engaging patients, managing care, and measuring performance.
In building a data platform and developing a performance analytic framework, organizational leaders identify process improvement initiatives aimed at achieving measurable performance goals in each performance domain. It is important to understand that comprehensive process improvement cannot be mapped out ahead of time in detail and that process improvement opportunities vary by population. Progress must be made iteratively through a series of initiatives building on previous accomplishments. The core of this effort is to create and continuously refine a care model that uses EHR to deliver greater care value.
Plan-Do-Study-Act. With the goal of better outcomes and lower costs in mind, it is useful to think about the entire process as a very large Plan-Do-Study-Act (PDSA) improvement cycle that runs continuously and spans the entire continuum of care, as shown in the exhibit below. This process weaves together existing data, puts the data in the hands of providers via the EHR and other tools, drives new interactions and outcomes, and generates fresh sources of data, enabling continuous care-model refinement.
This process can take place on the level of individual physician practices as they manage their patient populations. It can also help guide efforts to expand new care programs. However, the goal is to transform the way the entire care model functions across the continuum of patient care. If EHR-enabled performance improvement is limited to specific programs and practices, it can never lead to a significant value-based return on technology investments. Clinical and finance leaders need to use the entire system to identify large-scale performance improvement opportunities that transform care and outcomes for the populations served.
Work in iterations. Organizations should approach care model development as a series of experiments. For example, a health system’s goal might be to use a telehealth consult app to improve patient satisfaction. One approach is to roll the tool out to the entire physician staff, along with a system for capturing comprehensive patient engagement measures. However, the likely outcome is that nearly every practice will report unique operational problems—more issues than project leaders will be able to address and resolve.
Instead of planning a mass implementation, one approach is to start with a small pilot group consisting of three physician practices representing common operating environments, for example. The results will help an organization determine what works and what does not work in terms of technology, clinical processes, and patient outcomes. Based on these lessons, the organization can adjust the system until it works in the pilot practices. Once the system is functioning well, more providers can be added to the user group.
This approach, known as the “scrum” methodology, uses rapid innovation cycles to achieve an optimal design state. The scrum philosophy empowers healthcare leaders to experiment with value-based care before making sizable technology investments.
For instance, consider a health system that wants to build centralized care-management capabilities. Instead of investing millions of dollars in building a fully interoperable EHR system to link its entire provider network, the organization could use basic registry software to aggregate select EHR data and then build simple but functional care-management dashboards. This approach would allow the organization to rapidly experiment with care-management projects, learn from the results, modify the approach, and achieve real impact on outcomes and costs.
For the best results from rapid-cycle innovation, healthcare organizations should form a working partnership with their EHR vendors, involving the vendor representatives in all efforts to create value using technology. This approach will create opportunities to influence the development of the underlying EHR technology in a way that aligns with the provider organization’s value-based care goals.
Scrum methodology underscores the importance of a strong data-governance system. Iterative development must be a strategically focused process that aligns with the overall goals of value-based care.
Create a value model. A major challenge of evolving effective care models is making well-informed investment decisions. Executive leaders allocate money to various programs and initiatives, but they are never certain that these investments provide an adequate clinical or financial return. Finance leaders can make an important contribution by developing value-modeling tools for assessing population health investments and ensuring their efficiency under risk-based contracts.
For example, a healthcare organization may have several existing care-management programs for various patient populations. As the organization develops its value-based care systems, one logical option is to hire more care managers. To create a value model for this scenario, the finance department should begin by analyzing existing case-management programs. Key questions could be:
- How many patients did a care-management team interact with during a given period?
- What was the total cost of care for the population before and after the management period?
- How did key patient outcomes change during this time?
- What care-management expenses (salary, overhead, etc.) were incurred during this period?
Capturing total costs can be a challenge; however, it is possible to get a useful cost picture by combining EHR, billing system, and insurer data. The result is an accurate estimation of value that the organization can use to inform investment decisions and build a care model that enables strong performance under risk-based contracts.
A low financial return does not mean an investment should be abandoned. For instance, an organization might find that geriatric care management does not yield a significant return in terms of reducing total costs—yet it could still be the right thing to do for certain populations. The main takeaway would be the awareness that scaling this program would not make a significant impact on care value.
A New Yardstick
Physician productivity is still an important element of strong healthcare revenue. However, production can no longer be the sole metric for evaluating an organization’s investment in EHR. The transition to value-based care is creating a new framework for evaluating the value of these investments. Finance leaders can help build this model by positioning EHR as a tool for enabling population health management.
A healthcare organization should focus on building platforms for capturing rich data, creating intelligence through actionable analytics, and using technology to deliver that intelligence to providers at the point of care. In this way, organizations can drive value creation and realize the full potential of their EHR investments.
Shaillee J. Chopra is principal and chief digital innovation officer, Lumina Health Partners, Chicago.
Julie Bonello is senior vice president and CIO, Presbyterian Healthcare Services, Albuquerque, N.M.