As hype cycles go, predictive analytics made its transition from hyped to plateau of productivity long ago. Some organizations have moved onto the next stage of analytics maturity with prescriptive analytics, and data science in healthcare as a whole has developed beyond simply being able to predict what is to come.
Predictive analytics is defined by SAS as “the use of data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on historical data.” With this approach, clinicians and administrators can be warned of future developments before they happen, allowing them to make more informed decisions in situations where lives might depend on the ability to make the informed decisions quickly.
Additionally, areas such as detection of claims fraud, risk assessments and medication adherence offer healthcare providers, insurers and pharmacy companies opportunities to realize savings.
Investment in Predictive Analytics
Healthcare executives have come on board with analytics efforts in recent years. The reason for their investment is simple; many believe they will save money in the long term. According to a 2018 report from the Society of Actuaries, 60% of healthcare executives believed that using predictive analytics would save their organization 15% or more over the next five years.
Furthermore, 87% of providers and 83% of payers said they were currently using predictive analytics or had plans in place for near term implementation.
The reason for this is that understanding patient populations is considered one of the vital elements of any cost reduction plan. For example, it’s role in risk scoring is extremely important. Identifying those at risk of developing chronic conditions early in the disease’s progression helps treat the condition more effectively, thus avoiding more long-term health issues that increase cost and the complexity of treatment needed.
“The use of predictive modeling to proactively identify patients who are at highest risk of poor health outcomes and will benefit most from intervention is one solution believed to improve risk management for providers transitioning to value-based payment,” read a 2016 report titled “High-Risk Patient Identification” from the Association of American Medical Colleges.
That risk scoring can also be applied to readmission rates, predicting patient no-shows, suicide risk and patients susceptible to deteriorating conditions while in the hospital.
Predictive Analytics in Hospitals
One of the most common places to see predictive analytics being used is in hospital settings, particularly emergency rooms where there is a high volume of preventable activity that occurs. But use cases are proliferating as analytics practices become more commonplace.
One area showing promise is the elimination of operating room bottlenecks. By implementing an analytics system with complex event processing algorithms, hospitals are having success at improving workflows that streamline operating room processes. With a large number of individuals working in an OR, delays can be a tough issue to tackle. But by implementing processes that smooth the handoff process between tasks, staff can decrease turnover time between procedures by 15-20%, resulting in an estimated $600,000 annually in savings.
Care transitions is another area of promise, particularly following knee and hip replacement procedures. In the era of value-based care and alternative payment models in which quality and outcomes are highly scrutinized, readmissions and length of stay have become areas of focus. Focusing on discharge delays, analytics efforts often attempt to identify which patients can recover at home rather than an inpatient rehabilitation setting.
