A few weeks ago I posted an article called “Six Ways to Reduce Workers’ Comp Costs Using AI”. In the article I mentioned that by using AI model to predict the likelihood of applicant attorney involvement, one of our customers saw a drop in applicant attorney involvement from 13% of all claims to 4%. This statistic generated a lot of interest and a bunch of questions.

The interest generated by this dramatic decrease in the attorney involvement rate relates to the value inherent in reducing the rate. Many studies have shown that the added cost of applicant attorney involvement on a case-mix-adjusted basis is around $25-35K for lost time claims. The impact of the decline from 13% to 4%, if you have 5,000 claims and 25% of them are lost-time claims, is a savings of $2-to-4 million dollars over the life of the claims.

The questions generated by my article basically boiled down to, “What exactly was done to cause this impressive decrease in the applicant attorney involvement rate?” Here are three takeaways from our customer’s experience:

1. Get the math right

The first step is to use AI models that can accurately isolate claims that are heading towards attorney involvement early in the process (preferably based on data available on Day 1). Predicting this likelihood is not trivial and requires advanced data science/AI and sophisticated case mix adjustment. Our AI models have been refined iteratively over a ten-year period and have become increasingly accurate using only the most basic claims data available. Specifically, the accuracy of our attorney involvement models runs at 90%+ based on Day 1 data. I am sure there are others with highly accurate models as well. Both accuracy and early detection are important.

2. Build an early warning system

Once you have isolated the claims that are heading towards attorney involvement, you need to put in place a system to attend to the list of “problem” claims fast. This system needs to get the right information to the right individuals in a timely manner. It could vary in complexity from simple email-based lists to more complex API-based integration into existing claims systems. The important point is to make sure the communication reaches the team that will respond, which leads us to the final step.

3. Operationalize a rapid response mechanism

First let’s review some of the factors that increase the likelihood that an injured worker will feel the need to retain an attorney.

The Workers Compensation Research Institute (WCRI) reported that workplace trust issues, fear of claim denial and injury severity are the main reasons for attorney representation in workers’ compensation claims. The WCRI study found there was higher attorney representation if workers thought their claims would be denied either because of a denial notice, a delay in payments or perceived through communications with the claims manager. In those instances, workers were 34 percent more likely to retain attorneys versus 5 percent where there was no perceived denial.

Notice that merely a delay in paying a medical bill is enough to stoke the worker’s fear that his/her claim was going to be denied, thus leading them to retain an attorney. The most effective way to counteract this fear is through reassurance and the building of trust.

So, what exactly was done to cause the decrease in our customer’s applicant attorney involvement rate from 13% to 4%? The customer began by enjoining us to develop the optimal attorney involvement prediction model using sophisticated AI. The second step for the customer was to change its approach to dealing with those injured workers who rated a high likelihood of attorney involvement. This customer appointed their most personable in-house attorney and a part-time UR nurse as their “customer” care team; the customer in this case being the injured worker. Whenever an injured worker rated a high likelihood of applicant attorney involvement, the in-house attorney would contact the employee and speak with him/her to answer any questions they had about any procedural or legal aspects of the case. The UR nurse fielded any questions the worker had about their treatment, including the reasons for any treatment denials or delays. In a sympathetic manner, both members of this team focused their intervention on answering questions and reducing the injured worker’s uncertainty, fear and anxiety.

Some will say that this effort is unnecessary because dealing with injured workers is the claims adjuster’s purview and therefore injured workers already have someone available to answer questions. However, claims adjusters often have a hundred or more claims to manage at any one time and some can be indifferent or hostile to injured workers with a lot of concerns. it is quite a different experience having an attorney’s or a nurse’s undivided attention to answer questions vs. the harried attention of the claims adjuster.

The success of the customer care team in dealing with high risk injured workers depends on the team’s ability to form and nurture a relationship with the injured workers. Harvy Simkovits has written in the Complete Business Wisdom Library about ways to build and sustain trust in business. He cites the following nine factors as crucial to building trust:

  • Rapport – finding things in common with another person, even by mirroring and matching your body language to theirs, can help to create fertile ground on which to build trust
  • Honesty – always telling the truth about how you see things; offering your true perspective on matters at an opportune time (when the other person is open to your thinking)
  • Sincerity – demonstrating caring and unconditional positive regard to other person’s point of view, even if you disagree with their perspective
  • Respect for Self and Others – always talking to and dealing with others as equals and never as if they were lesser than or greater than you; i.e., never criticizing them or belittling yourself; saying what you agree with before you say what you disagree with; never laying blame nor creating guilt, etc.
  • Openness – fully hearing and understanding the other person’s viewpoint; allowing yourself to be impacted by their needs and ideas
  • Competency – demonstrating your knowledge and know-how around matters of importance to the other person; also demonstrating your ability to get to what is most important to the other person and in being able to differentiate your point of view from theirs
  • Mutuality – always working to serve all parties’ best interests; not being out just for yourself, or for just one or a few others; creating conditions where every party can win will work to bolster everyone’s positive regard of you
  • Integrity – walking your talk (having alignment between your words and actions); taking your promises seriously and working like heck to keep them not only to the other person but also to yourself
  • Reliability – being consistent in your behavior or in your way of being or acting; being someone others can depend on in fulfilling commitments.

In the final analysis, the best managed claims are those where there has been a relationship built between someone affiliated with the payer and the injured worker that promotes engagement of the injured worker in his/her treatment and recovery. If claims adjusters are too busy or are not relationship-minded, the payer should seize the opportunity to engage the injured worker by designating other appropriate individuals to build and nurture such a trust-based relationship.

In summary, if you want to reduce applicant attorney involvement rates, use the best possible AI to forecast risk of applicant attorney involvement, and then most importantly, work to build real trust with the injured workers.