Omaha National saves technology for the most difficult workers comp claims

September 21, 2022 6:30 AM

Omaha National Insurance Company headquarters, at 9110 West Dodge Road in Omaha, Nebraska<br/>

Digital Insurance spoke with Reagan Pufall, CEO of Omaha National, a workers compensation insurance provider founded in 2016. Workers compensation fraud costs more than $32 billion a year. The company seeks to stamp out fraudulent workers comp claims and fraudulent related premium applications and medical billing by taking the right "pre-technology," as Pufall puts it, steps such as utilization and information reviews. That way, applying technology such as AI to its workers compensation insurance operations can target what has gotten through a first screening. In October 2021, Omaha National launched a new claims management application that it built in-house.

How does Omaha use technology to find medical overbilling or unnecessary treatments?

In medical billing, we perform utilization review in house with our own staff. When a medical provider seeks authorization to perform a treatment, they get assurance that it will be paid for. We review medical bills ourselves with our own very highly qualified clinical staff. If you perform that operational review to a very high standard of meticulousness with really superbly credentialed, highly qualified people, you prevent the great majority of fraud. It doesn't make it past that operational review level.

That said, certain fraudulent behaviors will be difficult to combat. Our certified coders do a great job of identifying when a particular service has been coded and billed at the wrong amount. It's much more difficult to discern that there might be an entire medical facility or a medical practice that is falsifying treatment records and deliberately pursuing a scheme of fairly sophisticated fraud. That's where the special investigations unit steps in, and that's where you start to see the potential impact of good technology to aid in that effort.

How does “pre-technology” catch things like inaccurate reporting of numbers of employees?

We have well-trained human underwriters reviewing every submission we get regardless of the size of the account, but they also go online and do additional review. Sometimes you find things that are fairly easy to recognize as being significant. If you have a submission from a company and they say they have five truck drivers and you go to their website and see a photo of 10 trucks in the lot, naturally some questions start to spring to mind about who's driving the other trucks.

Most potential fraud is going to be caught at that operational phase. On stuff that makes it through that first net, you can start to use sophisticated technology. It is a frequently made mistake to think that fraud prevention is all about the special investigations unit and a software solution looking for patterns indicative of fraud.

Repeatedly, I've read about a new drug or medical treatment being greatly over-utilized, costing the industry many millions of dollars. I'll run some reports and I find out that we really have not been significantly impacted at all by that, because it just got caught in the normal thorough processes of doing things like utilization review and medical bill review correctly.

Is there a place for AI in special investigations?

Certainly AI is an important aspect of it, but some of it is even simpler than that. Some of it doesn't rise to the level of AI, it's just pattern recognition, exception analysis kind of measures that highlight a claim, a provider or employer we should be looking at. AI is a higher level of sophistication. Exception spotting technology is also highly effective.

We do see companies in the industry that take a delay and deny approach. They're just fighting about one issue after another and claims stay open for quite a long time. And we just don't really go down that path. We close our claims very quickly, among the fastest paces of claim resolution in the industry.

You could say it makes it harder to spot fraud, just because the claim's all wrapped up. But we have found if you take the approach of fighting about one issue after another, then you really are going to have to just be monitoring, monitoring, monitoring those claims for the emergence of fraudulent behavior. If we're going to spot fraud with our technology, it needs to happen very quickly, while the claim is still active and open.

How do you reach a quick resolution, which is not the norm, just in how you operate?

There's a number of ways that we go about that. One of the most impactful is that we are happy to settle a claim much earlier than many other carriers in the industry are. I feel really good about the idea of giving our company's money to the person who has the injury. We are always looking for an opportunity to form a really good relationship with the injured worker.

A lot of it is back to basics. Our adjusters have exceptionally low caseloads. I think right now our indemnity claim per adjuster ratio is about 70, very low compared to the rest of the industry. So we can do really meticulous work and our adjusters can really get to know their injured workers and form good relationships with them, then fairly early in the claim explore the idea of providing a very fair settlement amount and closing out the claim, letting the claimant go forward on their own.

That just forecloses a lot of opportunities for misbehavior, not necessarily even by the claimant. It's just a very happy, healthy way to manage a claim and take care of people while reducing claim costs, because it's an iron clad rule, if you take any book of claims, the sooner they close, the less they're going to cost. But from the perspective of the particular issue that we're talking about today, the shorter period of time the claim is open, the smaller the window of opportunity is for somebody involved in that claim to engage in some improper behavior.

Tech isn't magic. It's only as good as the assumptions and strategic thinking that you put into it. If you're going to use AI or more simple algorithms, like exception spotting, whatever approach you take to anti-fraud efforts, you have to monitor and measure the right things. It helps to have a very sophisticated understanding of medical coding and billing to be able to spot fairly subtle patterns, let's say in a given clinic's utilization of certain treatment methodologies where you'd say to yourself: For a clinic this size, seeing this frequency of treatment does not look right. It does not seem to be explainable by a normal variation in practice patterns. It's not enough to have technology knowledge and expertise. You really have to have fine-grained underlying expertise in the subject matter as well.

How has the migration to remote work changed workers compensation insurance?

We've entered this era where work from home or remote work has become more and more common. We serve small to mid-size companies doing work for folks who work with their hands and therefore have a higher risk of being injured. Of course, people in those industries pay a higher premium.

We are much less affected by the phenomenon of work from home than many carriers are simply because if you're going to do landscaping, you're not going to do it in your living room. Even then, we could be insuring a machine shop in Southern California with 20 employees but now their bookkeeper has decided to move to another state, but they like her. Now she's in Tennessee and she's still doing their bookkeeping. The nature of employment really is changing. Every worker's comp carrier is going to have to be a national worker's comp carrier because employees are scattered all around the nation. A claim from a state might be the only claim you get from that state that year. You don't have a big storehouse of experience. You need nimble staff and terrific information resources.