Readmissions are a major problem for many healthcare organizations in the U.S. According to AHRQ’s Healthcare Cost and Utilization Project (click here), the overall readmission rate in the United States was 13.9% in 2016, with the average cost across any type of principal diagnosis being about $14,400. Given this strain on the healthcare industry, it is essential to develop strategies to reduce readmissions, and some organizations have risen to call to drive improvement.
Atrium Health—a system of 40 hospitals and 900 care locations—is one organization that has successfully used its data analytics department to identify root causes to its readmissions and implement targeted interventions to drive reductions. To learn more about the power of data analytics, the HBI Cost & Quality Academy spoke to Dr. James Hunter, senior vice president and chief medical officer; Jennifer Shore, director of business operations for the quality department; and Dr. Jarrod Bullard, a senior health services researcher.
Q: What kind of readmissions-related data does Atrium Health gather and analyze?
Dr. Bullard: We use most of our readmission data from our hospital data that we process every month. We use an algorithm that mimics what CMS does—called QCC readmissions logic—which looks at a facility’s 30-day unplanned readmissions. The data is pulled by SQL from a cloud-based warehouse and put into a statistical software program called SAS with a combination of other packages like Microsoft Power BI. They are all used to clean and manage the data, and the retrieved product can be used to build a dashboard or conduct different analyses.
The health system also has patient-centric dashboards that are particular to different facilities, and we use the algorithm to track unplanned 30-, 60-, and even 90-day readmissions—which we especially make a point to do for problem areas like COPD or sepsis. We can also create dashboards that show an analysis of our patients’ admission rates per individual facility within the health system, as well as different patient demographics, discharge dispositions, and zip codes. We have different corporate statistical mapping techniques within the dashboards.
Q: What data can be used to prevent readmissions from occurring?
Dr. Hunter: We have criteria in place that are based on the risk stratification of patients—how likely they are to be readmitted. We’ve developed a program that arrives to a number from analyses, saying whether a patient is at low, medium, or high risk, as well as the desired follow-up time frame. For high-risk patients, follow-up is five days, medium is two weeks, and low is 30 days. Risk stratification information is visible in the EHR to both doctors and nurses.
Using these stratification levels, doctors can determine the best method to follow up with a patient. For instance, a group of nurses do phone calls and send reminders about follow-up visits no matter the level; however, to bridge the gap between discharge and follow-up appointment for a high-risk patient, we have transition clinics in a couple of our communities like behavioral health. Staff in these clinics may treat patients for up to one month before returning them to their primary care doctor.
Q: What additional actions can hospitals do to ensure readmissions are reduced?
Dr. Hunter: The most crucial step is the healthcare team‘s education and awareness about readmissions and our tactics to care for at-risk patients. Awareness of a nurse or caregiver alone can help with care delivery or education at some of the critical interfaces with patients and families.
Showcasing the value of teamwork is equally important, so we make sure to share progress we’ve made via metrics such as readmissions. That means that when departments bring forward specific work to drive improvement—like the oncology or COPD—we need to track the results of those efforts and share data that illustrates the improvement.
Another aspect is to identify bottlenecks in the flow of work and think of new solutions. I can give you an example: Once primary care physicians at Atrium Health started seeing patients faster, the discharge summaries were not available on time. To avoid this, the new rule was for discharge summaries to be prepared within 24 hours and the actual document was restructured. The doctors got together to streamline it from a three-page dictation that had the plan on the last page to a template where the care plan is listed on the top of the very first page.
Shore: The last practice each healthcare organization should adopt is looping in all parts of the continuum of care. They should be involved in the effort and brought to the table to hear the conversations taking place about the data, barriers, and improvement opportunities. Improvement efforts are more effective when an issue is tackled with everyone at the table discussing it together.
Interested in hearing more about data analytics and readmission reduction? Reach out to the HBI team today by filling out the form below!