Study Supports Benefits of Evidence-Based Medicine

Workers’ comp claims that follow evidence-based medicine guidelines have shorter durations and lower medical costs, according to a new study. The research suggests significantly improved outcomes and cost savings can result when medical providers follow recommendations based on peer-reviewed evidence in workers’ compensation treatment guidelines.

While nearly all jurisdictions either have or are considering the adoption of evidence-based medicine guidelines in their workers’ comp systems, there is almost no published scientific evidence confirming their efficacy or mechanism for improvements. But a team from a workers’ comp insurance carrier and Johns Hopkins University School of Medicine have produced what they believe is the first scientific proof that consistently applied treatment guidelines are effective in treating injured workers.

“We set out to prove or disprove empirically that adherence to EBM guidelines was impactful,” said Jack Tower, senior data scientist at the Accident Fund Holdings Medical Center of Excellence. “We were able to do that.”

The researchers developed a methodology to measure adherence to the Official Disability Guidelines from the Work Loss Data Institute and used an adherence score to compare the outcomes for different case mix adjusted claims populations. They found that claims in which there was at least a 50 percent adherence to the guidelines had 13.2 percent shorter durations and 37.9 percent lower medical costs.

“That kind of gives a strong impetus to implement new medical management strategies based on the results,” Tower said. “Carriers and the work comp industry could benefit from developing programs that embrace the concepts behind EBM.”

Evidence-Based Medicine

The idea of evidence-based medicine is to improve the medical decision-making process by emphasizing the use of scientific research and medical consensus. While it has been around for the last several decades, evidence-based medicine has only recently become widespread in the workers’ comp system.

“The application of evidence-based medicine in workers’ comp is much different from the application of evidence-based medicine in the group health world,” said Jeffrey Austin White, director of Innovation for Accident Fund. “In group health the evidence-based medicine guidelines have been scrutinized by the medical professionals as they are limited in scope and typically used to control cost in a hospital setting by limiting reimbursement rates.”

“This study provides a mechanism for evaluating an EBM guideline and can be used to identify how they might be improved in the future.” — Jeffrey Austin White, director of innovation, Accident Fund Holdings

However, White argues that the workers’ compensation guidelines are much more focused and comprehensive. “Evidence-based medicine [in workers’ comp] encompasses tens of millions of claims having similar incoming diagnoses. The guidelines provide outcome expectations at the diagnosis and treatment level for the majority of workplace injuries,” White explained. “When the diagnosis is made, the evidence-based medicine guidelines define how often a treatment is administered, along with the expected cost and time off from work. It’s a much different way to apply evidence-based medicine than is typically done in the group health setting.”

In addition to the Official Disability Guidelines, the American College of Occupational and Environmental Medicine has also created evidence-based medicine guidelines. A majority of states have adopted or are considering adopting either of the two national guidelines, a combination of the two, or homegrown guidelines that are state-specific to improve consensus around the definition of “necessary and appropriate” treatments for injured workers.

But “there’s a paucity of research around evidence-based medicine and best practice protocols,” said Dr. Dan Hunt, corporate medical director of Accident Fund. “We wanted to use our research to come up with hard facts — things that are true — to help improve the care for injured workers whether it’s Official Disability Guidelines, ACOEM, or another to say ‘here’s objective research that shows these guidelines work.’”

The researchers wanted to show whether and to what extent evidence-based medicine works specifically in the workers’ comp population. “There is a lot of literature that suggests the correct ways to do things medically but many times they are not really proven from an outcomes point of view,” said Dr. Edward Bernacki, professor of medicine and director of the division of occupational medicine at the Johns Hopkins University School of Medicine. “The medical care may be better, but does it really affect costs and return to work?”

The Study

Previous research from Accident Fund in conjunction with Johns Hopkins has highlighted some of the reasons for the increasing use of opioids in the workers’ comp system. One study, for example showed the use of opioids was an independent predictor of catastrophic claims costs while another identified physician dispensing as a driver of the increased use and costs.

“We found that physicians were contributing to [the opioid problem] and asked ourselves ‘Why?’ Our hypothesis was that providers were not using guidelines to help make administration decisions,” White said. “We thought by developing an algorithm or methodology to analyze a historical cohort of claims that we might be able to see a difference in outcomes between case mix adjusted claims that had various degrees of compliance with the guidelines.”

The idea of the study was to develop a technique for testing the safety and efficacy of an evidence-based medicine guideline rather than to drive public policy decisions on treatment practices.

It’s one of those situations where everyone wins — the employee returns to work and medical costs are constrained. To me, it’s a win-win.” — Dr. Edward Bernacki, professor of medicine and director of the division of occupational medicine, Johns Hopkins University School of Medicine

“If a state mandates the use of evidence-based medicine guidelines for the treatment of injured workers we are legally obligated to use them. If there are no mandated legislative guidelines, we are inclined to promote prospective guidelines that have been shown to reduce system costs and positively impact injured worker outcomes,” White said. “It’s important for us to know which guidelines work and why. This study provides a mechanism for evaluating an EBM guideline and can be used to identify how they might be improved in the future.”

Measuring Evidence-Based Medicine

The team developed two separate analytical techniques; one to stratify each claim for medical complexity and another to determine the adherence to the Official Disability Guidelines. The claims were divided into 10 levels of medical complexity and scored based on adherence.

“The number one challenge when doing claims research is being able to group claims into like claims,” White explained. “You don’t want to compare a claim with a broken finger to a claim with head trauma.”

The group started with non-catastrophic, indemnity claims that spanned the years 2008 to 2012 of the insurer’s data. They considered open and closed claims using a two-year development cutoff.

The researchers developed a compliance score to determine adherence to the Official Disability Guidelines. The score assigns a quantitative value to the claim indicating approximately how many of the treatments were consistent with the recommendations from the guidelines.

They case mix adjusted the claims and compared those with greater than a 50 percent adherence to evidence-based medicine guidelines to those with less than 50 percent adherence for the differences in claim durations and medical costs incurred. Using data from Official Disability Guidelines, the researchers identified the adherence of every procedure given a specific diagnosis for each claim based on the following four codes:

Green flags in the Official Disability Guidelines indicate the procedure is recommended based on prevalence, medical consensus, and historical claim outcomes.

Yellow flags indicate the procedure is a common treatment for that diagnosis and should be allowed on a limited basis with a restriction on the number of times it should be performed.

Red flags denote low prevalence in workers’ comp and that the treatment is not necessarily indicated based on current scientific research, i.e., recommendation is to review.

Black flags indicate inappropriate care and possibly denial of service.

“For every diagnosis and treatment, we label it with the corresponding colors; then we determine an adherence score at the claim level,” White said. “For a given claim, you can consider the cumulative number of green, yellow, red and black flags, and you can devise a score that indicates the level of compliance which can be compared against like claims.”

Based on the scores, the claims were separated. Those with mainly green and yellow flags, for example, were deemed as fairly compliant with the guidelines while those with many black flags were noncompliant.

“If you break the claims into two buckets, you can compare outcomes of the compliant group with the noncompliant group,” White said. “So for two broken finger injuries where one received compliant and the other noncompliant care, you can see how they differ in duration and medical cost.”

Results

The average for all levels of medical complexity showed claims in the low compliance group had a 13.2 percent increase in claim duration and a 37.9 percent increase in medical costs compared to the high compliance group, the study found.

The numbers increased as the medical complexity of a claim increased. In looking at the top 10 percent of claims for medical complexity, there was a difference in claim duration of 18 percent and increased medical costs of 38 percent, between the low and high compliance groups.

The researchers also found there were more black flag procedures in the low compliance group — 3.5 times the number in the high compliance group.

“I think our research in essence provides evidence that if you do employ these guidelines the outcomes are better,” Johns Hopkins’ Bernacki said. “This is systematically over time that people return to work faster, for the insurers costs are a little lower, and for folks employing them the premium costs will be lower, so the cost of doing business will be lower. I think it’s one of those situations where everyone wins — the employee returns to work and medical costs are constrained. To me, it’s a win-win.”

Future Research

“It’s awfully exciting to be a part of a landmark study. No one else has done this before,” Hunt said. “The ability to develop an adherence process for claims management will have a lot of applications across the whole health care spectrum.”

Hunt, who called the study a “gargantuan undertaking,” hopes it will lead to additional studies that drill down more into the findings. “Age, jurisdictional differences — there are a whole host of really interesting things we can do now,” he said. “You’re going to see additional papers once this method is established.”

For now, the authors hope the findings will help spur action in states that currently do not use evidence-based medicine guidelines in their workers’ comp systems. With properly worded legislation and effective dispute resolution processes in place, evidence-based medicine guidelines should offer better outcomes for everyone.

They hope workers’ comp practitioners will begin using the methodology they’ve created to further refine evidence-based medicine guidelines. In fact, they have developed a 10-step process for companies to replicate the results.

“It’s like a recipe. With evidence-based medicine guidelines, you can quantify exactly how much of each ingredient you put in and therefore enhance your ability to refine, measure, and improve your results over time. At least that is what EBM tries to do,” White said. “It’s a recipe that applies to, say 80 percent of the population most of the time. The recipe should reduce system costs and facilitate cooperation from both sides of the business — payers and providers alike.”