Data drives decision-making in healthcare
Amid the turmoil of reform, clinical and administrative progress is being made in healthcare. Providers increasingly are able to mine patient data to help choose the best course of treatment, and sophisticated software solutions are allowing them to fine-tune their claims process.
Sequoia Hospital is a community hospital in Redwood, Calif., with a range of specialties, including particularly strong programs in cardiovascular services. In those and other programs, the hospital draws from throughout northern California. For more than a decade, the hospital has been contributing to national cardiac registries developed by the American College of Cardiology and the Society of Thoracic Surgeons. The data is used by hospitals for a variety of purposes, ranging from accreditation to ensuring clinical best practices.
To derive maximum value from the data in the cardiac registry, Sequoia uses IBM SPSS predictive analytics software—specifically, IBM SPSS Modeler data mining workbench—not just to provide the statistics on past outcomes that are needed for reporting purposes, but to manage patient risk through predictive analytics. “We knew about SPSS from sister hospitals, such as St. Joseph’s in Stockton, Calif.,” says Audrey Fisher, manager of cardiovascular outcomes and quality, “and were impressed by their positive comments about it.”
Sequoia collects and stores the information in a local database using heartbase software. “We add some fields to our version of the database in order to be able to look longitudinally at factors such as long-term survival,” Fisher explains. “The techniques we use are very specific to our hospital, and we want to be able to look at our own population.” In addition, the hospital can easily evaluate the impact of new techniques for producing desirable changes, such as reduced post-operative infection rates.
Prior to using IBM SPSS predictive analytics software, Sequoia needed to route requests for such analyses through the IT department, which took several weeks. Now, a team of analysts in Fisher’s department carries out the requests at near real-time speed. Wireless technology has provided another option that serves physicians and patients well. “This new capability allows me to take my laptop to a surgeon’s office while a patient is there and assess the risk given specific factors affecting the patient’s condition,” Fisher says. “It is not unusual for me to get a call from a doctor over the weekend, asking for information about a procedure.”
Sequoia Hospital has been using SPSS software for nearly a decade, and each year is able to enhance the value of its data. “The size of our database increases steadily so we have a wealth of data to draw on,” Fisher continues. “For example, we discovered from our data that diabetic patients with a high level of a certain hemoglobin component had worse outcomes for cardiac surgery. If they are treated by an endocrinologist before surgery, the outcomes are better.” The detailed data allows a more nuanced evaluation of prospective treatments. “We are able to look not only at critical factors, such as mortality,” Fisher says, “but also at quality of life, because our patients also are thinking about that.”
In fact, the hospital holds the best record nationally for survival from valve replacement over the past six consecutive years. Fisher is convinced that the hospital’s aggressive commitment to data-driven decision-making, combined with its physicians’ skills, has produced that achievement. In addition, Sequoia Hospital ranks in the top percentile for patient satisfaction. “We believe we are on the forefront of information collection and use,” Fisher says, “and our patients are regularly benefiting from this approach.”
Much of the basis for analyzing outcomes originated with efforts to detect fraud in Medicaid claims, according to Bill Haffey, technical director for the public sector at SPSS. “Association analysis was being used to find data that did not match the usual pattern of diagnosis and treatment,” he says. “But the same type of analysis could be used to see if a set of events, such as a visit to a provider followed by a prescription for a particular medication, was associated with an event such as a long interval before the patient needed to be seen again.”
Although many valid conclusions can be drawn from analyzing existing structured data contained in a registry, the growing availability of electronic medical records (EMRs) will lead to increased evidence-based medicine and smarter healthcare. “Much of the pressure for EMRs is actually coming from physicians who want to be able to query a database of their patients against various questions, such as whether someone has been vaccinated or not,” Haffey says. “These records will have a lot more detail about individual patients than is currently available for analysis, and will open up the potential for large-scale population studies as well, to promote improved outcomes management.”
Wright State Physicians is a large subspecialty practice in southwest Ohio. Its physicians support the clinical, educational and research activities of the Wright State University Boonshoft School of Medicine. The organization was looking for a new practice management system that would eliminate the need to maintain software in-house, and allow easy interconnection of the organization’s multiple locations.
The practice selected software from athenahealth because the product line is delivered as a SaaS product and includes both practice management software, called athenaCollector, and an EMR, athenaClinicals. “As an academic institute that is training medical students and residents, we wanted to implement EMRs before they became more widespread,” says Jocelyn Piccone, COO at Wright State Physicians.
Meanwhile, Wright State Physicians is benefiting from improved claims processing, thanks in part to the ease of modifying rules. “Insurance companies have a variety of rules for their claims,” Piccone explains, “such as the way certain procedures are controlled. Some are specified in the contract, and some are not. In addition, the rules can change each year.”