The Big Picture Approach to Healthcare Analytics
by Mason Mercy, on Feb 20, 2018 6:00:00 AM
Hospitals all face the same problems: operational hurdles, denied claims, value-based care imperatives. Yet there have been fundamental differences in how they’ve approached these challenges. Many have faced them by adding people or attempting to update processes, yet have ultimately failed to stem the tide of revenue loss and inefficiency. Others, a class of early adopters, foresaw the value in taking a different approach: one that relies on the fundamental advances in computing technology to bear the brunt of data analysis and information gathering.
Working with these early adopters, XSOLIS has been evolving our approach to better accommodate the changing needs of hospitals and health systems, taking a precision approach to utilization management. Our proprietary Care Level Score (CLS) assimilates discrete data – labs, vitals, meds, and the like – to summarize the likelihood of inpatient status in a numerical score: the higher the score, the greater the likelihood that the patient should be admitted as inpatient.
Building upon this traditional CLS, we now utilize machine learning – the type of data science that discovers patterns and learns as it accumulates more information – to further assess the probability of the appropriate status based on documentation – a key missing component when relying solely on the discrete measures previously mentioned. These machine learning models build upon the tens of thousands of individual cases and outcomes from our clients to better assess risk and probability to provide actionable intelligence. This actionable intelligence in turns allows our clients to manage utilization ‘by exception,’ only manually reviewing cases that fall within our grey zone.
The adage of work smarter, not harder certainly applies to this approach, yet doesn’t begin to do it full justice. In just one instance, a large health system saw >90% accuracy across their inpatient population and across their payers, with the data aligning at 12hrs or less into the patient stay. Such a high degree of accuracy allowed them to get the status “right” up front and helped decrease their risk of denied claims. This automated approach also frees up their staff to use their expertise to tackle care transitions, discharge planning, and the facets of care delivery that benefit most from a human touch, yet that often are overlooked due to time constraints.
For hospitals looking to adopt this approach, the timing has never been better. Adopt the technology that allows you this type of opportunity, work with your internal teams to define an acceptable threshold for manual and automated review, implement the approach, and begin to reap the rewards.
The outcomes – increased revenue, heightened productivity, data-driven care determinations – speak for themselves. If you’d like to see how this approach can support you and your organization, schedule a demo with our team.