Data - Why Is It So Hard?

Over time, I’ve written a lot about data; and it has been about almost all aspects of organizational management. That’s because to thrive as an organization in an increasingly complex contracting, payment, and valuation environment, data is key. More accurately, information is key. And they are by no means the same. This year’s (as well as in past years) respondents to the Behavioral Health Industry Trends Report, said they can organize their data by program, service, and setting (59%), but only 50% said they can effectively use that data to support a value-based care model. This is the case even though 74% say they use their EHR to tie service delivery to claims submission. The challenges cited include a lack of real-time reporting capability, the inability to create reports without going to the IT staff, and not having defined metrics to benchmark and measure. Only a third of organizations said they had access to “big picture” and drill-down data needed to report on key metrics.

Three Types of Data for Value-Based Care

Human service organizations have three large buckets of data that need to be aggregated to provide information essential to a value-based care model: data related to the organization’s human resources (data that can assess the productivity, cost, tenure and deployment of staff),  information about finances (revenues and expenses), and service data (population data such as demographics, data about service level and volume, and data abut service impact). The challenge is integrating data into information to permit the organization to evaluate how well it is providing value while thriving financially and culturally.

Defining Your Measurement Terms

Understanding the underpinnings of a value-based care model is critical. One definition of value is the ratio of benefit to cost. The definition of the terms of that equation, however, can differ between an organization’s various stakeholders. Leadership teams need to make sure that they define the terms in ways that help them address the concerns of numerous external constituencies and their board. Leadership may define value as making sure that they are meeting patient/client needs while creating a robust bottom line that supports growth and investment in their staff. For an insurance company paying for services, “benefit” is often measured in terms of reduced use of deep-end services, which reduces their cost. Community stakeholders may define benefit in terms of community tenure or decreased reliance on other community resources such as jail or homeless shelters. Patients/clients and their families and caregivers are likely to define benefit in terms of their subjective experience of distress or wellbeing and cost in terms of affordability. All this information is important, but how it is calculated and reported is not always easy to glean.

In their various forms, these definitions require answering these basic questions in terms used by the varying stakeholders: What are we doing? For whom? With what outcomes? And at what cost relative to our resources? So, the first step in developing informative data is focusing on questions critical to the organization’s success in the eyes of each stakeholder group. Who is being seen, and are they within the targeted population? What services are they receiving and at what level of care? How much service are they receiving and what does each one cost? And are those services effective? Efficient?

Integrating Data to Gain Better Insights

To answer these questions usually requires integrating data from multiple systems. The ability to calculate the cost of a service requires that you evaluate the cost of the direct staff time that goes into it and the infrastructure that supports that time. That requires data from an HRIS system, the clinical record, and the financials that have the cost of all the ancillary supports for that clinical service (facility, HIMS, admin, etc.). This kind of integration requires technology that can link information from separate software systems, but it is not an IT problem. The questions answered by the data must come from leadership, stakeholders, and payers. And it must allow for real-time queries. It should not require IT to “write” a new report for each question. Drill-down is essential. Standard reports are fine, if there is sufficient flexibility that key variables can be filtered in or out to address relevant operational and outcome questions; and at different levels of aggregation (e.g., program, department, overall, etc.).

Benchmarking is another challenge reported and this is a complicated issue also. If two organizations, or an organization and a stakeholder, have differing definitions of an outcome, a service, or even the allotted time for a service, then comparisons can result in erroneous conclusions. A simple example is the common metric “time to first appointment.”  The start and end of that calculation is based on when the organization classifies the client as seeking services. Some use the referral call from another agency, others the contact when the client seeks services either by phone or as a walk-in. Is “first” the first available or the first accepted? What if the person fails to keep that appointment, does that delay “first?” Comparison requires very clear definition of terms and the flexibility to drill into data varying those many potential definitions to create meaningful comparisons or benchmarks; this is not frequently what has happened with poor results for improvement in organizational effectiveness.

The Importance of Solutions that Offer Flexibility and Responsiveness

The solutions to the challenges posed by the need to have information useful in a value-based compensation and service delivery model are complex and costly. Careful planning is essential so that terms and questions are clear and relevant. The organization must make an investment of time and financial resources in technology to provide accurate data that is entered and available in real time. Each of those elements requires cross-departmental participation in iterative process that is the only way to ensure consensus and that can result in a system that is flexible and responsive to organizational needs over time.