Time matters when you’re injured at work. Medical costs in the United States have historically made up 60% of total workers’ compensation expenses, and any delay in approval of these expenses can add months to the worker’s recovery, delaying their return-to-work (RTW). If there is any place where artificial intelligence (AI) can offer the most benefit in terms of its speed, it may be the worker’s compensation claim. 

Earlier intervention means higher quality claims

AI medical insights can help insurers and risk management professionals to ‘see the future’, metaphorically. If insurance can become a protective talisman to policyholders, perhaps its  predictive analytics that offer the same comfort to claims. Predictive analytics can estimate when exactly an injured worker is expected to return to the job, depending on type of injury and how their condition impacts the nature of their work. This information is invaluable to insurers concerned about risk, since it provides a direct insight into the cost of the claim. 

Workplace injuries costs American employers over $44,000 annually. Around 10% of these injuries will result in prolonged or permanent withdrawal from the workforce. Each day of delay in returning to work can spike the cost of a claim, sometimes substantially. Short term absence from operational duties may require some shifting around of your workforce, but longer term absences may require temporary workers or modified duties. Injuries not resolved quickly can lead to litigation, permanent disability, and longer times off work (which means more wage replacements, medical bills, and legal fees to pay). None of this is pleasant for claimants and their families, nor employers who have to backfill the injured employees duties during recovery.

Technology Streamlines the RTW Process

Traditional models for predicting return to work slot patients into timelines based on return-to-work guidance from similar workers in their industry. Today’s AI tools are more capable of keeping up with the ‘human speed” of recovery in these claims, and where insurance companies might need to intervene. For example, using a dataset built from 10 years of Ohio Bureau of Workers’ Compensation claims, researchers at the University of Michigan built a deep learning model to update its predictions of whether the worker would return in 7, 14, or 30 days. The model updated daily, giving insurers greater precision when deciding whether or not to intervene, with prediction values that better matched with actual outcomes. 

The longer an individual is out of work, the less likely they are to return, and a multitude of variables might contribute to this outcome. Although RTW predictions are made by physicians who specialize in occupational health, the patient themselves will probably see their family doctor or emergency room physician. Occupational health providers are more effective at getting patients back to work than family physicians, since the family physicians treat the symptoms, but do not necessarily consider RTW as a treatment goal. 

Since many generalist physicians don’t have training in occupational health or disability though, their care can unintentionally delay the patient’s return to job duties by overlooking opportunities for modified duties or not encouraging early attempts to return. However, the growth in disability claims, coupled with the popularity of RTW programs implemented by employers, has significantly increased the amount of administration these physicians must do (including provide documentation at the request of claims teams).  

Faster Claims, Smoother Recoveries

Automated claims documentation, real-time insights, timeline views of medical events and treatment options can streamline these processes. This helps claims teams to get through paperwork faster, and doctors spend less time answering questions. Today’s technology can triage, segment, and automate claims in a fraction of the time: leading to faster RTW outcomes, happier doctors, and patients who don’t need to wait months for documentation delays.