Harnessing clinical and administrative data for hospital performance improvement
This article was originally published in Health Data Management:
Data creates a ripple effect. Activity data from a smartphone can dictate what rebates customers get from their insurance providers, or what ads they see while scrolling through Twitter. Clinical data is no different. When a patient enters an emergency department (ED) for treatment, receives an outpatient procedure, or lays in a hospital bed, the clinical insights generated from each step of that patient’s care can become key to helping improve care.
Today’s advanced analytics have the power to help clinicians quickly isolate common patterns in their patient populations, so they can intervene sooner and more effectively, all while streamlining costs and improving patient care. Nearly all of today’s hospitals may data inputs necessary to conduct this type of analysis using their own operational and clinical data, many can find it challenging to develop the processes required to use this data.
That is starting to change. Spurred by the value-based care movement, more healthcare providers have begun the process of realigning their administrative and clinical models to focus on delivering the highest possible quality at the lowest possible cost . Along the way, key metrics such as 30-day readmission rates, bundled payment benchmarks, and hospital acquired infection rates have become the Holy Grails of hospital performance measurement .
As hospitals get better at tracking this data, it is now becoming possible to analyze the multitude of variables that are linked to each discreet data point to quickly and accurately identify avoidable care, spot gaps in the process of care, and even tailor interventions based on warning signs in that data.
Identifying Clinical Opportunities
The first step in leveraging analytics in the clinical setting is to accurately identify the biggest opportunities for performance improvement. To do this, hospitals will want to work with a large repository of patient discharge data that captures several years of system-wide patient care information and corresponding risk-adjusted data.
The critical component here is depth of information. Hospitals have access to an amazing array of data points, but often have a fairly myopic focus when it comes to overall performance. What’s even more important is the ability to develop a granular view of the process of care. To truly spot the kinds of trends and anomalies that can shine a spotlight on opportunities to improve care, it is important to know each step along the way that went into achieving that performance, good or bad.
For an example of how this can play out in a real-world setting, consider a recent case study involving a hospital quality improvement initiative focused on chronic obstructive pulmonary disease (COPD).
Hospitals everywhere are struggling to find ways to improve medical errors which has become the third-leading cause of death in the United States . For Schneck Medical Center, a 93-bed community hospital in Jackson County, Indiana – where the COPD population is roughly double the national average – the disease was creating both clinical and operational challenges .
After analyzing its data, hospital administrators found that it had a raw readmission rate of nearly 14 percent and that 10 percent of those readmissions were due to COPD. Those COPD readmissions alone were costing the hospital nearly $300,000 per year .
Moving from Data to Action
Once the initial heavy lifting of data programming and analysis was completed and these red flags were identified, Schneck was able to start the process of adjusting its approach to COPD. This included developing a long-term care practice that makes weekly respiratory care visits, incorporates sleep studies into its observations, and conducts patient discharge planning. In addition, the hospital put in place new protocols that included the installation of a transition team to help with patient discharge, follow-ups with recently discharged patients, and annual facility education regarding COPD.
Over a three-year span of operational and clinical tweaks to its COPD process of care, the hospital saw its unplanned readmissions rate decrease by 80 percent. Even more impressive, the total number of COPD hospital admissions fell by 55 percent . Ultimately, because Schneck was taking better care of its patients in the community – through focused interventions that were tailored specifically to the COPD population – people with the disease were not requiring as many hospitalizations in the first place.
The Road Ahead
As technology evolves, with high-powered analytics surfacing even more insights, the potential to realize results similar to Schneck Medical Center grows exponentially. With more and more organizations leveraging these powerful tools, the faster and more intuitively performance insights will emerge from seemingly disparate sets of data. That has the potential to unlock huge value for organizations, and ultimately, a better level of care for patients everywhere.
[4,5,6] Leventhal, R. (2018, October 9). A Data-Driven Effort to Tackle Indiana’s COPD Problem. Retrieved from https://www.healthcare-informatics.com/article/analytics/data-driven-effort-tackle-indiana-s-copd-problem