Tackling Ed Utilization: Multiple Facets Requiring Multiple Solutions

Alexander Koster, Associate Vice President, Value Transformation, Nemours

If you are in a technology leadership role supporting any value-based care or population health efforts, you are likely aware of the specific needs and challenges of those programs. Achieving success in alternative payment models and population health initiatives requires advanced analytics integrated with emerging and new data sources, tailored EHR functionalities focused on specific populations (or new/supplemental systems), and innovative approaches for engaging patient populations through diverse communication channels.

As with most analytic related efforts, it is critical to understand what problem the health system is trying to solve

ED Utilization, and more specifically avoidable ED utilization is a critical area of focus in value-based care programs. Effective programs to manage and address ED Utilization require a variety of different analytics approaches, governance and project structures, and technical solutions tailored around each intervention.

Remember to design your analytics solution to allow exploration for individual patients/encounters as well as population patterns and trends. This is critical when designing analytics for VBC and Population Management

As with most analytic related efforts, it is critical to understand what problem the health system is trying to solve and define the problem statement. Are you tackling ED utilization to address a gap in chronic disease management for a specific

Your analytic solution should be able to readily provide ED diagnostic patterns, by population segment, PCP, clinical registry, payer and more!

condition or population (asthma or diabetes for example)? Is work intended specifically to be a general driver for overall cost reduction? Is the goal to hit a specific payer target for ED/1000 (Emergency Department Visits per 1000 member months) or PPV (Potentially Preventable Visits) or High Utilizers? Are you addressing a specific issue with LANE (Low Acuity Non- Emergent) ED visits? Are you just trying to find out how you are doing in general, to determine if you NEED to do anything? Perhaps some combination of these, with a different answer depending on the market, payer, or even differing perspectives of the operational leaders. To answer these questions your analytics solution should have ready access to robust ED utilization data from several sources. Claims provide very “clean”, usable data, but have a considerable lag. If your health system manages one of the Emergency Departments that your patients frequent, then you should consider building a parallel set of more timely analytics out of EHR clinical and administrative data that can be used for local operational intervention, as well as for validation against claims. Finally, HIE or external data that is exchanged via ADT feeds, ENS, of similar technology, can provide even more real-time data, particularly about emergency visits that happen away from your health system, though there are limitations as to data quality and robustness. The newer population health platforms, Business Intelligence solutions, and robust EHRs allow all of this data to be integrated into a unified patient view for individual patient analysis.

Do not assume that patient preference or patient behavior is necessarily the main drivers of LANE or “Avoidable” ED visits!

Payers have also upped their game, often relaying alerts, high utilizer reports, and performance metrics using elaborate portals, or even advanced integration options to share data. Payer- driven reporting can provide readily available actionable information, as well as an interesting source of data for validation. For example, to compare system vs payer list for high utilizers on a value contract and identify any discrepancies.

“Remember to design your analytics solution to allow exploration for individual patients/encounters as well as population patterns and trends. This is critical when designing analytics for VBC and Population Management”

Beyond these elements analytics should allow you to explore what happens upstream and downstream of your emergency department to identify patterns and insights that can help eliminate barriers for timely access to care. Did your patient’s call the triage line? What was the disposition? Did they try to get into a primary care sick visit? Were they able to connect to a primary care staff? Was an appointment available? Basically, what was the journey to the ED?

And after it? Was there a follow up outreach? What was the success rate connecting? Was there a follow up visit if one was needed? Factors such as time of day, day of week, geographic proximity to the ED, patient language and interpreter need, health related social needs and more, practice access for same day sick visits, all may shed light on where to intervene, implement a process change, or expand hours available in a non-emergency care setting. The key is to approach the data with curiosity and follow the analysis where it takes you (or your new advanced AI tool!). The answers and recommended interventions may surprise you.

Business Intelligence