USING PREDICTIVE ANALYTICS
Within the U.S. healthcare industry—in this survey, composed of 78% providers, 12% payers and 10% other organizations— fewer than half are currently using predictive analytics (43%). Nearly the same amount (42%) are not using predictive analytics, and 15% of respondents are unsure of their organization’s status. This even split represents an honest picture of an industry that has historically made business decisions differently than other industry sectors. The fact that healthcare is provided regardless of a patient’s ability to pay for services is just one factor that makes healthcare business decision making unique.
When diving deeper into the data, we find disparities among the different sectors within healthcare. A large majority of payer organizations in this survey use predictive analytics (80%). That number dives to 39% for medical groups/clinics and nursing homes, and even further to 36% for hospitals and health systems. Payers arguably operate more like businesses than providers, basing many of their decisions and systems in actuarial science, like other insurance operations. This helps make sense of their much higher use. [Figure 3]
Belonging to an accountable care organization ( ACO) affects whether a healthcare organization uses predictive analytics. Of organizations that are part of an ACO, 52% are using predictive analytics vs. 28% who are not using them. Why are ACO-related organizations more inclined to use predictive analytics? ACOs need analytics to evaluate risk, more so than the average healthcare organization, because ACOs tie provider reimbursements to quality metrics. The better able organizations are in predicting outcomes, the better able they are to create positive results.
What type of information is being predicted at the 42% of organizations that say they’re using predictive analytics? The most common outcomes being predicted are hospital readmissions and costs, both reported by 55% of respondents. Inventory needs are the least common to be predicted, with 13% of respondents predicting this category. [Figure 4]
Differences emerged again between the industry sectors on this question. Medical groups/ clinics and nursing homes were more likely than other sectors to predict adverse events ( 39% vs. the survey average of 28%) and staffing/ workforce needs ( 50% vs. the survey average of 31%). Payers, interestingly, are more likely to predict clinical outcomes ( 57% vs. the survey average of 49%). And hospitals and health systems are more likely to predict hospital readmissions ( 62% vs. the survey average of 55%). While nearly half of hospitals and health systems predict costs, a much higher percentage of medical groups/ clinics and nursing homes ( 78%) and payers ( 77%) do so.