Data analytics is a key driver in the development of business strategy and workers’ comp claims are a goldmine of information. Yet, when not used properly, the results can fall far short of expectations. Here are five common mistakes:
- Relying solely on the insurance company Some employers rely solely on the insurance company to analyze their claims and make recommendations to prevent injuries and control costs. In recent years, insurance companies have beefed up their analytics and embraced predictive analytics to manage claims. They use information from years of past claims to build models that will predict what may happen next in a particular claim. Indeed, such information benefits employers.Insurance companies also are a great resource for claims information in your industry. They can provide helpful guidance for how you stack up versus your peers.But it’s important to have realistic expectations and remember that the insurance company’s goal is to leverage data to improve their profits. This can lead to aggregate information or a cookie-cutter approach that falls short of your needs.
- Data such as injured-worker demographics, department, type and severity of injury, frequency, timelines and money set aside for reserves of claims, and if the claim ends up in litigation can all help employers guide future outcomes. Smart employers regularly review their loss run reports from the insurance company that includes this information, not only to ensure it is correct (errors mean increased premiums) but also to identify trends that lead to actionable insights. What are the main drivers of incidents in the organization and what can we do to change are the key questions to ask in analyzing data.
- Observing metrics at face value Each year, Risk & Insurance identifies “All Stars” who stand out from their peers by overcoming challenges through exceptional problem-solving, creativity, perseverance, and/or passion. One of the 2018 All-Stars was Kevin Farthing, environmental health and safety manager for Florida-based Sparton Electronics, a 600-employee company manufacturing sonobuoys for the navies of the world.The company faced a high number of musculoskeletal injuries and annual workers’ comp claim costs exceeding $500,000. Multiple modifications to the production processes and attempts to control ergonomic risk factors had not solved the problem.Digging through the data, he discovered that 40 percent of the musculoskeletal injuries were occurring during the first three years of employment. The company was hiring workers who were not capable of performing the physical demands of the job.
- He then took the logical next step and worked with a company to design specific post-offer, pre-employment tests to make sure candidates were up to the physical challenges. But he did not stop there.
- The failure rate on the test was high – 50%. Rather than lowering the demands of the tests, he identified which tests individuals were failing most and modified the actual work tasks. For example, they no longer require employees to manually move certain types of heavy loads. Coupled with other changes, a two-year investment of $174,000 has yielded an expected savings of more than $950,000.
- Not being objective or hanging on to old beliefs Commitment to the status quo or leadership thinking may limit taking action on data. Some rationalize that the incident rate is acceptable and changes will mean lower production. Or a belief that “injuries are part of the job” or simple complacency. Buy-in from management can take effort and tenacity.For many years, it was believed (and documented) that inexperience and inadequate onboarding put younger workers at increased risk and they were more likely to suffer a workplace injury. On the other hand, older workers would experience fewer injuries but would take longer to recover and have more costly claims. Recent research from the National Council on Compensation Insurance (NCCI) dispels this conventional wisdom and finds that younger workers are getting injured less often than their older peers.The workforce is changing and processes are becoming more automated. While the number of workers under 55 has remained more or less stable, the number of workers who are 55 or older has doubled since 2000. Women make up more than half of labor force growth. Relying on old data or beliefs leads to ineffective and costly programs.
- Year-over-year analysis will show how claims are changing. This will tell you if initiatives are working or if a new direction is warranted.
- Failing to segment An important finding of the NCCI research was that key injury risks vary by age group. Younger workers are prone to injuries from contact with objects or equipment, while overexertion injuries are most vexing for employees in the middle of the age spectrum. Meanwhile, slips, trips and falls disproportionately affect those over 55.There’s clear value for employers to mine their own claim data correlating type of injury with age and gender of workers. When younger male workers are experiencing a higher incidence of injuries from contact with objects or equipment, a change to interactive and technology-based training, rather than a dry manual, could be an effective way to improve safety.It’s not just age subsets that can help employers to be tactical in the way they manage their safety budget. Comparing similar departments can identify why one department may be functioning at a higher level than the others and then apply the best practices to other departments.
- Not looking beyond the data Although there are many sophisticated data tools, programs cannot rely on data alone. There is a myriad of subjective factors that affect incident rates. Production pressure, management safety practices, limiting mind-sets, and fear of automation are just a few.These factors cannot be quantified with statistics. Instead, organizations need to have subjective methods to review these factors that represent the “heart” of their workers’ comp program.
For Cutting-Edge Strategies on Managing Risks and Slashing Insurance Costs visit www.StopBeingFrustrated.com