Improve Healthcare with Data Analytics
The healthcare industry is awash in data. Anyone who has walked into a doctor's office and seen the reams of paper records realizes that's nothing new. As that data is increasingly delivered in digital form, though, the ability to collect, collate and analyze that data has opened new ways for providers to run their businesses more efficiently. That in turn improves treatment outcomes for patients.
Finding ways to analyze this data to generate meaningful information healthcare providers can use to improve care is a growing focus for the industry. Just understanding basic factors such as the exact makeup of the people using the emergency room, or who is being readmitted to hospitals for treatment, can result in big gains. Even bigger gains are thought possible with the broader application of predictive analytics to cut costs, manage patient and staff workflow, diagnose diseases and coordinate care.
With greater adoption of electronic health records (EHR) and other electronic reporting systems, provider organizations can now use clinical data to improve performance, the Healthcare Information and Management Systems Society (HIMSS) states in a recent report. "Successfully leveraging these data along with pharmacy and laboratory data has enabled organizations to track and deliver a superior quality of care than was possible in the past," according to the report. "In conjunction with claims data, these organizations are also developing a greater understanding of utilization of resources and optimizing their costs."
The increasing demand for providers to shift from a fee-for-service approach to one that puts the onus on value encourages providers to deliver the best care at lowest cost. It also pushes using predictive analytics to more closely track outcomes, and show where and how to make changes.
Barriers to achieving even a decent level of using analytics are substantial, however. Ensuring the broadest data set requires myriad data sources such as pharmacy, lab reports, claims and billing, as well as EHRs. That in turn means ensuring interoperability between these data sources and being able to store the data. It also requires standardizing, normalizing and integrating data invariably produced in various different formats.
Beyond the technical challenges, healthcare organizations also have to tackle policy and cultural concerns. The "not invented here" syndrome is frequently mentioned in connection with implementing analytics. An organization may resist lessons from other organizations in building their analytics platforms. That can lead to more costly and less effective solutions.
The possible returns from a successful implementation are substantial. At the federal and state level, the Department of Health and Human Services (HHS) was one of the first organizations to commit to a broad use of analytics to reduce fraud and waste in the Medicare and Medicaid programs.
Since June 2011, the Centers for Medicare & Medicaid Services' (CMS) Fraud Prevention System (FPS) have used predictive analytics to investigate all claims. So far, they have saved around $820 million in payments, a 10-to-1 return on the investment CMS made to implement the system.
HealthInfoNet, Maine's state health information exchange, noted significant improvements across its network in 2015 because of predictive analytics and modeling. It helped St. Joseph Healthcare, for example, reduce emergency visits by 15 percent; cut its 30-day emergency return rate by 9.5 percent, its 30-day readmissions by 13 percent, and its hospital mortality rate by 37.3 percent.
It's still early in the adoption process for data analytics in healthcare and success rates will continue to vary. As legislative and business pressures continue to push providers to be more efficient in delivering better care, data analytics will become a pervasive need.