Using a surgical approach to workers' comp claims

Accurate analytics can help insurers identify claims that could escalate and patterns that lead to higher payouts.

Companies should strategically apply any predictive models, warning flags and business rules to claims. (Photo: Shutterstock)

Workers’ compensation insurers across the country are leveraging data analytics, medical expertise and root cause analysis (RCA) to implement innovative ways to identify the right claim, using the right resource at the right time. The success of these efforts can vary drastically, from predictive models that fail to deliver as expected, all the way to data-inspired processes that significantly reduce costs and are built into claims protocols.

Here are some thoughts on avoiding false positives and a look at how workers’ compensation insurers can leverage data to help prevent adverse developments, often referred to as “jumper” or “developmental” claims. When used successfully, policyholders benefit from lower insurance premiums while their employees receive better and more timely care after an accident, helping to avoid unnecessary surgeries, treatment delays and excessive prescriptions.

Avoiding false positives

A false positive occurs when a result shows something not really there. For claim adjusters, nurses, physicians and others, false positives can drastically reduce the morale of the entire team. There is nothing more discouraging and time-consuming than working on improperly flagged claims. When these claims involve the use of outside resources (e.g., physicians, investigators, physical therapists), the costs can add up quickly. A high false-positive rate can cause an organization to lose confidence in the models, flags and business rules developed using today’s mathematical techniques and software.

Companies should strategically apply any predictive models, warning flags and business rules. If the parameters are too wide, claims professionals can be flooded with unnecessary work. Before implementing any automated system or protocol, stress test the impact of any proposed changes from a workload perspective. If adjusters’ current caseload of 125 to 150 cases doubles after implementing a predictive model, and the insights from the model scores and reason messages lack real value, people will fight the change; however, if the company’s automated system or protocol allows professionals to focus on the right claims, freeing them up to add more value earlier in the claim life cycle, the injured worker and the entire organization win.

Human-driven & advanced analytics work

Chesapeake Employers’ Insurance Company (Chesapeake Employers) uses a mix of human-driven analysis and advanced analytics to help reduce the risk to policyholders and injured workers. The team’s in-house doctors, nurses, physical therapists and pharmacist work side-by-side with experienced claim professionals to help ensure medical treatment for injured workers is timely, thorough and appropriate.

The health services team of three doctors, 25 nurses, one pharmacist, and one physical therapist works hard to identify outlier events that may be early warning signs for adverse claim development. For example, the pharmacist works with the external Pharmacy Benefit Manager (PBM) to monitor opioid drug utilization and prescriptions related to the work injury while addressing any safety concerns associated with the prescribed medications. A dedicated pain management nurse monitors claims where there are signs of chronic opioid use and mitigating issues such as non-compliance with treatment plans or a dependence issue. These manual efforts help workers avoid dependency and addiction issues related to opioid overprescribing. This has helped reduce the dollars spent on dispensing opioids by 70% over a four-year period and seen the number of injured workers receiving opioids decreased by 66%.

Although Chesapeake Employers doesn’t use predictive models to identify claims with opioid issues, other companies do leverage them with success. Travelers and Hartford Financial utilize artificial intelligence to identify at-risk claimants or physicians who may be overprescribing opioids. Both approaches, human-driven and advanced analytics, accomplish the same goal of ensuring injured workers do not end up addicted to opioids as a result of a workplace accident.

Analyzing claims to prevent adverse medical events

In the Insurance Thought Leadership article, “Carrier’s Perspective on Large WC Claims,” Mark Walls discusses how two-thirds of all claims Safety National defines as large losses started out as fairly routine claims, including back, shoulder and knee injuries. Safety National calls these claims “developmental claims.”

The Chesapeake Employers medical team performed an in-depth review of over 100 of its largest claims, totaling in excess of $350 million in claim payments. As Walls outlined in his article, a number of these were developmental claims, where multiple failed surgeries drove up claim costs. Although the claim adjusters focused on the right items and maximum medical improvement considerations initially, many of the claims eventually exhibited clear warning signs from events occurring after an injury was reported.

To address this, the medical team worked closely with the claims department to develop an adverse medical events (AME) protocol. For example, several AMEs were identified for escalation, such as a pulmonary embolism, unexpected fall, second surgery shortly following an original surgery and excessive prescriptions. When an AME happens today, the claim adjuster works with the nurse case manager and precertification nurse to notify the claims supervisor, health services supervisor and in-house physician. This process provides the opportunity for continued monitoring and direct attending physician contact to discuss the plan-of-care when indicated.

Data-driven behavioral health referrals

Behavioral health issues impact recovery in a variety of ways that could prolong disability. The issues that arise are not necessarily diagnosable psychiatric conditions but rather driven by an individual’s own makeup and subsequent response to the circumstances surrounding an injury, missing work and the claims process. Chesapeake Employers has had significant success in referring select patients for cognitive behavioral therapy (CBT) or a multi-disciplinary rehab program that includes physical therapy, pain management and cognitive behavioral therapy.  While the potential benefit to this approach is more obvious on older claims where there might be prolonged absence from work and possibly prescription drug dependence, proper selection of patients earlier in a claim presents challenges. There are four major considerations when reviewing earlier claims data:

  1. Most injured workers will not need these services to recover;
  2. These programs can be costly depending upon the duration needed although 10-14 visits of CBT or 4-6 weeks of multi-disciplinary rehab is typical;
  3. There is a risk that a psychiatric diagnosis becomes added to the claim; and
  4. The patient refuses to attend for different reasons, including advice from their attorney or their perception of being referred to a mental health provider.

The use of cognitive-behavioral therapy in the setting of a worker-related injury is not primarily focused on the diagnosis and treatment of a psychiatric condition. The CBT sessions are billed and coded as treatment of the patient’s injury as the treatment is focused on their individual coping skills in overcoming perceived challenges to returning to work, including fear of re-injury and pain, as well as helping the worker to re-establish a structured daily routine which may have been absent during their treatment phase. The result is greater success in helping injured workers reach maximum medical improvement, return to work and with settling claims. Once CBT helps an injured worker gain insight and control over thoughts and feelings influencing the injured worker’s ability to recover, negatively impactful habits can be replaced with positive behaviors that help speed up the recovery process.

Kevin Bingham (kbingham@ceiwc.com) is the chief results officer at Chesapeake Employers’ Insurance Company in Towson, MD., and leads their subsidiary efforts and Statistical Analysis and Research Group.  Dr. Stephen Fisher (sfisher@ceiwc.com) is medical advisor to the CEO at Chesapeake Employers’ Insurance Company and leads the medical analysis and research of innovative new approaches to treating our injured workers to help deliver the right level of care.  

Related: