Shaping Emergency Department Utilization Through Predictive Analytics

In the United States, one percent of healthcare users account for 21.4 percent of total U.S. healthcare spending. “Frequent flyers” to emergency departments (EDs) heap a considerable cost burden onto the system, especially considering that reducing unnecessary ED visits could save an estimated $38 billion.

Despite the huge expense of unnecessary emergency department utilization, there is a silver lining – these encounters produce valuable data that can ultimately be used to curb healthcare costs. With every mouse click and keystroke, EHRs are generating discrete data points that can help physicians and system leaders better understand the drivers of utilization in the communities they serve.

As hospitals and health systems prepare for Stage 2 of meaningful use, they have the opportunity to employ data analytics in building and enhancing their business intelligence capabilities.

For a system planning for population health, the nature of ED care creates a data-rich environment that offers insight into the health issues (such as the flu) affecting a large and diverse population. The organization can also capture a snapshot of the most acute patients in the region, such as chronically ill patients, who routinely use the ED for conditions that could be effectively treated in the primary care setting.

Predictive analytic tools can be used to glean critical information from the encounters and patient records of high-frequency users. By systematically sifting through the data, software can help clinicians discover and flag risk factors for repeat admission and use this information to design targeted triage and follow-up care protocols geared toward reducing utilization. Additionally, collecting and analyzing this data can reveal opportunities to fill gaps along the healthcare continuum.

For instance, in the example below, leaders in a busy east coast-based ED recognized an uptick in re-admissions and were looking for ways to identify high-risk patients at the beginning of their visit. After discussions with physicians and clinical staff, analysts collected and examined data from clinical and financial systems.

What resulted was a decision tree that could identify patients at high risk for readmission and trigger the appropriate follow-up care protocol at triage.ED_Readmission_Decision_Tree_14M10D09

However, simply purchasing software is not enough. To realize the full potential of predictive analytics in the ED, health systems must be prepared to create a culture that embraces a data-driven approach to operations.

Specifically, organizations must be prepared to:

  • Embrace the EHR – For many providers and their clinical teams, effective implementation of EHRs requires a change in “how we’ve always done things.”  Like any change, it might not be immediately welcomed. However, in order for many analytic tools to be effective, using the EHR as it is designed is essential. Working around the EHR limits the efficacy of any analytics platform.
  • Validate data integrity – Missed key strokes and typos can frustrate English teachers and editors, but in an ED, they can significantly affect care quality. Department leaders must make high-quality documentation a priority. Tracking deficient charts and following up with providers and clinical staff is a mandatory competency.
  • Fit the tool to the problem – While many analytic platforms may offer solutions out of the box, it’s important to first understand exactly what problems exist. Leadership and physicians need to work together and target areas of opportunity before choosing their analytical approach.

EDs are gold mines for rich, actionable data – data that offers organizations an inside view of the health behaviors of patients and communities. Developing competencies in business intelligence and predictive analytics will allow progressive hospitals and health systems to leverage this data in responding to and shaping utilization patterns to ensure that patients are seen in the right setting by the right provider at the right time.

This post was originally featured on the athenahealth Health Care Leadership Forum – September 19th, 2014.

This entry was posted in Data Analysis, EHR/EMR, Emergency Medicine, Performance Improvement and tagged , , , , by John Budd. Bookmark the permalink.

About John Budd

John has extensive experience working in hospital-based physician practice management and project management. His ability to discover value in a wide range of service areas, ranging from revenue cycle to clinical operations, through qualitative and quantitative analysis helps clients save money and streamline operations. Prior to joining ECG, John worked in hospital-based physician practice management with a broad range of experience in emergency medicine and revenue cycle operations at a three-facility health system in Virginia. He possesses a bachelor of arts degree in English and a master of business administration degree with an independent research focus in management science and applied analytics in healthcare, both from the University of Mary Washington.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s