Skip to main content

Carolinas HealthCare Mines Its Data Warehouse to Improve Care

Issue

Imagine highly detailed, computer-accessible patient records that go well beyond basic medical history to include information about:

  • Diagnoses, procedures and outcomes;
  • Complications and countermeasures;
  • Post-procedure care;
  • Readmissions;
  • Financial/insurance resources;
  • Demographic and geographic correlations to common illnesses found among the patients’ respective populations;
  • Tests or care patients may need in the future.

Access to these kinds of information—plus the ability to assess physician and nursing staff performance, productivity and success rates—is at the heart of Carolinas HealthCare System’s (CHS) Enterprise Data Warehouse (EDW). Housed in a server pool that can hold up to 10 terabytes of data, the EDW is central to CHS’ efforts to deliver more effective, more successful healthcare among its 39 hospitals and 900 care locations and in partnership with third-party providers. It was developed for the company by CHS’ Dickson Advanced Analytics (DA2) and Information Services departments.

A Holistic View of Patients

“Our EDW gives us a 360-degree, holistic view of our patients,” says Chris Danzi, CHS’ IT leader of business intelligence/data warehousing, Web and process improvement. “This gives our healthcare providers the ability to provide more focused and comprehensive care while at the same time not wasting resources on less-effective options and approaches.”

“One of our priorities is to reduce the number of patients being readmitted within 30 days of being discharged,” says Seth Nore, CHS’ director of information services, EMR management. “Using data analysis, we can identify each patient’s level of risk of requiring readmission, and determine pre-release care options and post-release support that can reduce this risk.”

Identifying and Reducing Readmission Risks

Take, for example, 78-year-old Edgar, hospitalized for an exacerbation of his COPD. Edgar responded to treatment and was due to be released after 3½ days. But CHS’ data analysis showed he had an 80% likelihood of being readmitted in the next 30 days, based on six indicators:

  1. COPD diagnosis;
  2. Exacerbation;
  3. Three emergency room visits in six months;
  4. No primary care physician;
  5. History of hypertension;
  6. No use of home oxygen.

Edgar had moved to the area a year earlier, after his wife died, to be closer to his son and daughter-in-law. Although it was written on every set of discharge instructions, Edgar hadn’t connected with a primary care physician. When he was anxious, coughing or short of breath, he called his daughter-in-law, who—having no medical training or knowledge—often drove him to the ER late at night.

Armed with this data, Edgar’s care team scheduled him for a follow-up appointment before he left the hospital, to occur within three days of discharge. His daughter-in-law enrolled in “caregiver college” so she would know how to meet his health needs more effectively. And Edgar got an appointment for pulmonary rehabilitation to learn how to manage his disease. Edgar now had a team focused on providing him the best quality of life. The result was better health, fewer ER visits, and fewer hospital stays and unplanned readmissions.

Impact on Patient Care

Edgar’s story may be an idealization, but CHS’ real-life results speak for themselves. Using the EDW, CHS physicians are able to predict their patients’ 30-day readmission risk with nearly 80% statistical accuracy. “It’s not just a matter of identifying high-risk patients. Our analysis also helps us spot patients with apparently low risk levels who nevertheless could be readmitted due to often-unnoticed secondary medical conditions, home environmental factors and inadequate after-hospital support,” Danzi says.

In addition to its individual patient records, CHS’ EDW has access to a wide range of identity-obscured population data that can create accurate contexts for assessing a patient’s overall prognosis and after-hospital challenges. “We can model the health risks in a given geographic location—such as air pollution triggers for asthmatics—contextualize that against risk levels for specific age groups and people with certain pre-existing health conditions, and create a backdrop against which we can more appropriately treat patients from these places,” says Nore. “The result is better care, delivered more efficiently and cost-effectively. Everybody wins.”

Uncovering Opportunities for Improvement

CHS’ EDW data analyses can also provide interesting opportunities for management. “Say we find that one of our most successful knee surgeons is using the most expensive knee replacements, even though less-expensive alternatives exist that deliver comparable performance and durability,” Danzi says. “We can steer them to these alternatives—still satisfying their need to deliver top-quality care—and maintain the same level of outcomes while reducing costs.”

These are just some of the many benefits being delivered by CHS’ EDW system. “It did take a significant commitment to create the CHS EDW,” Danzi acknowledges. “But the system provides CHS with tremendous insight into optimizing care delivery for the benefit of our communities, and positions us to reduce the cost of healthcare in the region.”

James Careless is a freelance writer with extensive experience covering computer technologies.



Back to Top