THE OPPORTUNITIES — AND THE BARRIERS
The opportunity for using predictive analytics to make better decisions in healthcare is high and expansive, according to surveyed healthcare executives. Direct clinical and financial outcomes are the most valuable data to predict, with clinical outcomes leading ( 55%) and costs— whether per patient, per episode of care, or through another lens— coming in a close second ( 52%). Less critical, but still considered valuable, are the following predictors from data: reimbursement ( 35%), hospital readmissions ( 35%), staffing and workforce needs ( 32%), and patient demand and population shifts ( 28%). [ Figure 1]
There are notable differences between payers and providers. Payers are more than twice as likely as the survey average to choose patient behavior and diagnosis as valuable outcomes to predict. They also place far less emphasis on staffing and workforce needs, with only 8% identifying this as a valuable outcome to predict vs. 31% of providers.
Where there are opportunities, there are also challenges. When asked to identify their organization’s biggest obstacle to implementing predictive analytics at their organizations, healthcare executives cited incomplete data ( 20%) and insufficient technology ( 19%). These are not unexpected, as the industry’s slow acceptance of technology compared with other industries has caused a lack of structured, organized data— both of which are key to leveraging predictive analytics. Interestingly, almost as many healthcare executives don’t know the top obstacle their organization faces, uncovering an absence of strategy or urgency around using predictive analytics. [ Figure 2]
Payers and providers exhibit notable differences in this question, as well. Hospitals and health systems are more likely to lack the sufficient technology ( 23%) than payers ( 3%) or medical groups/ clinics and nursing homes ( 14%). Payers, on the other hand, are more likely to encounter incomplete data, with 31% noting this as their top obstacle, vs. the survey average of 20%. Payers are also less likely to face any barriers at all, with 15% citing no barriers vs. the survey average of 4%. Additionally, medical groups/ clinics and nursing homes are twice as likely to lack the skilled employees needed for predictive analytics.