With healthcare reform, HITECH, and the ICD-10 transition all moving providers in the same direction, it's high time to begin applying analytics to performance improvement
There is widespread agreement among healthcare leaders that a lack of clinical data has never been a problem; indeed, hospitals, medical groups, and health systems have for decades been awash in clinical data. Rather, the problem has been organizing that data, and leveraging it to do things. For, compared to most other industries, healthcare remains woefully behind in its ability to use its core operational data to drive core operational performance; and of course, healthcare's core operational data is fundamentally clinical, and its core operations are its care delivery processes.
As pioneering organizations have been venturing forward on a variety of paths in recent years, one term that has gained widespread use in the past few years is clinical intelligence, which combines business intelligence with clinical data. Not surprisingly, the same patient care organizations that have tended to be very far ahead in the area of clinical intelligence have also tended to be national leaders in quality, patient safety, and performance improvement work as well.
IN TERMS OF THE USE OF ANALYTICS BASED ON AGGREGATE DATA, I'M FINDING THAT MANY ORGANIZATIONS HAVE UNDERGONE SOME KIND OF PLANNING INITIATIVE AROUND BUSINESS INTELLIGENCE AND ANALYTICS IN THE PAST FEW YEARS. -MITCH MORRIS, M.D.
But now-with the passage of the PPACA, or federal healthcare reform, last March-particularly with its establishment of ACOs, the medical home model, and value-based purchasing under Medicare-the situation on the ground is changing rapidly. And, when coupled with all the provisions around data collection and analysis in the federal HITECH Act, the time has come for leaders of all hospitals, medical groups, and health systems to move forward quickly to extract the intelligence hidden in their reams of data in order to analyze gaps in their care quality, patient safety, cost-effectiveness and efficiency; develop integrated and coordinated care and contracting models; and prove their performance to purchasers and payers.
The most advanced patient care organizations nationwide have been deeply involved in a continuous performance improvement loop for years now, but they remain in the minority. Indeed, while industry experts see things moving forward now, it's not at the pace or level of consistency one might think.
GOOD PHYSICIANS, GIVEN GOOD DATA, WILL MODIFY THEIR DATA TO IMPROVE QUALITY AND EFFICIENCY.-WILLIAM MOHLENBROCK, M.D.
“In terms of the use of analytics based on aggregate data, I'm finding that many organizations have undergone some kind of planning initiative around business intelligence and analytics in the past few years,” says Mitch Morris, M.D., national leader of health IT at the New York-based Deloitte. “They're developing a strategic plan, and budgeting for things; but this thing of analytics almost always falls below the line, and rarely gets full funding, so adoption has been slow. But meaningful use will speed this up, as well the shift towards accountable care organizations, and the push towards quality-based purchasing.”
Morris sees the coming transition to the ICD-10 coding system as a key factor for analytics. “I think that with ICD-10 implementation, we'll have a far greater level of detail of documentation,” he says. “In the past, for example, you've had one procedure code for lumbo-sacral fusion, which is spinal fusion for lower back pain. Soon you'll have more than 200 procedure codes for that same procedure, based on small variations on the same theme. So we'll be challenged in many ways, but we'll also have the opportunity for a much richer level of analysis in terms of clinical performance and operations; and you'll be able to make much finer distinctions among individuals and groups.”
For some provider CIOs, there's no time to waste. Tina Buop, CIO at the 105-physician Muir Medical Group, based in Walnut Creek, Calif., the need to compete in one's local market and government mandates are already making the use of clinical analytics tools compulsory. “The burden,” says Buop, “is that the government is telling you to collect an increasing number of data points about patients.” But, given the flood of data that patient care organizations are already collecting in some form, she says, the core issue is focusing on both the requirements and the opportunities, in a strategic, systemic manner.
“Where clinical intelligence needs to go is to be able to use technology to pull together disparate data, and give more intelligence at the point of care,” she says. “And when you have 18 different specialties contributing to your chart, you've got a lot of relevant, non-electronic data as well. “Buop and her colleagues have a forged a unified EMR across 18 sites of care, all of which are independently owned. And they've used data continuously for many months to determine how successful their physicians were using the IPA's disease management templates in several key areas, including diabetes.
On the inpatient side, organization's like the 871-bed Northwestern Memorial Hospital are moving forward into more and more refined areas of analysis for performance improvement. For example, says David Liebovitz, M.D., medical director for clinical information systems at Northwestern, his organization is moving forward quickly to provide real-time clinical performance reports to clinicians based on clinical data collected in a very timely way.
William Mohlenbrock, M.D., chairman and CMO at the Las Vegas-based Verras, and a practicing surgeon, has spent years working with hundreds of hospitals nationwide to do variation analysis, probing for answers as to why some physicians' care is more or less expensive and has better or worse outcomes than others. And while the range of costs and outcomes, risk-adjusted, is inevitably tremendous when his consulting firm first comes into a hospital organization, with data analysis and coaching, it often quickly turns around. The lesson of his experience is a simple one, Mohlenbrock says: “Good physicians, given good data, will modify their data to improve quality and efficiency. “
The kind of work that Mohlenbrock has been doing with his clients will inevitably become more widespread, as, under the PPACA, hospitals will be penalized for unnecessary readmissions, and preparing to meet whatever new standards are set forward by the federal government will be impossible without being able to use business intelligence tools to analyze clinical and financial data countless different ways. The same will be true in terms of supporting ACO and medical home development, competing under value-based purchasing regimes under Medicare, and ultimately, experts agree, staying in business going forward.
Healthcare Informatics 2011 March;28(3):28-32