In the seminal 1992 study, conducted by Arntz, De Jong and Van Eck, the researchers assessed the fear levels of test subject who had been told they’d be given an electric shock. The researchers found that subjects who knew they were getting a shock but didn’t know if it would be mild or intense exhibited more fear than those who knew that they would receive a more intense shock.
The underlying point is that we naturally want to avoid feeling uncertain; however, that is difficult to do in the healthcare industry. Imagine, for example, that you are a hospital administrator, knowing every month that change was coming but not how it would impact your hospital’s business. Without a clear picture of the changes that will take place, you are left with no choice but to react after the change has already taken place and affected your business. This pattern is not only inefficient, but also strenuous on the organization as it scrambles to realign with the new state post-change.
Ever-changing Payer Policy Updates
Hospital margins are narrower than ever in today’s industry. Tightening margins mean hospitals cannot afford to allow any payments to fall through the cracks. Yet, a shocking number of claims are unexpectedly denied and eventually reimbursed at a lower rate to hospitals and their providers who have difficulty keeping pace with the frequency of policy changes enacted by private payers.
Millions are lost every year due to professional fee underpayments. Often, this is due to the complexity and quickly changing nature of real-world payer processes. Generally, hospitals find themselves at a disadvantage in their business relationship with payers as it is common to ultimately collect 60 percent of the original claim submitted after it has been denied. One of the ways payers maintain their leverage is with policy updates that result in additional denials of claims submitted using the previous guidelines.
In today’s industry, private payers update their medical necessity and deniability qualifications on a monthly basis. These changes are typically communicated to hospitals in a 30-page document sent once a month. However, hospitals usually only have the bandwidth to sit down and assess these documents every six months.
Hospitals rarely have the ability to properly ask why a claim was denied. On top of that, the flow of claims is almost immeasurable and frequently complex to the extent that, when reports are generated, there are so many factors that the report needs a report. This is an example of how hospitals are data rich and information poor.
Why do these payer policy changes take place so frequently?
Large payer organizations employ an army of actuaries. That degree of resources allows payers to continuously update their policies on a monthly basis, segmenting general risk from modifiable risk in policy updates and handing them back to hospitals in 35-page documents once a month. The vast amount of actuarial recourses that major payer organizations employ allows them to replicate this cycle at scale, something hospitals simply are not able to keep pace with on a month basis.
Further, contracting hospitals aren’t factoring the administrative burden tax. Most organizations aren’t dealing with payers in a way where they understand the impact of time and resources utilized to work denied claims. Were the hours spent adjudicating a denial only to resolve a claim for 60% of its original cost worth it? It’s difficult to say.
However, hospitals do have tools at their disposal that can be used to stabilize this imbalance in power.
Understanding the role of machine learning and predictive analytics
When data analytics and machine learning are applied to this issue, these patterns begin to become very predictable. Yet, the majority of hospitals still aren’t monitoring for it.
In addition to having the data, timing is important. Even if a hospital was able to effectively monitor denial trends and successfully convert their data into actionable information, they would still find themselves playing catchup. Retroactively updating organizational coding guidelines in response to every payer policy update is like trying to drive down the road utilizing only the rearview mirror to navigate. Rather, for hospitals to gain leverage on private payers, it is necessary to get ahead of a payer’s next policy update by anticipating policy changes using prescriptive analytics.
By applying prescriptive analytics in combination with machine learning, it is possible to shift from a blind or reactionary model to an anticipatory model, which could decrease denials and, perhaps more importantly, the resource-intensive administrative burden of working and re-working denied claims. In this way, machine learning may also provide a novel means for hospital administrators to at last decrease their monthly uncertainty by gaining control and optimizing their professional fee reimbursements.