Medical Provider Networks for Care of Injured Workers Have Been Around for Many Years

The data supporting the creation of medical provider networks is starting to pile up.
Risk & Insurance – Peter Rousmaniere Dec. 18, 2012

Medical provider networks for care of injured workers have been around for many years. But only recently have researchers and practitioners come forth to proclaim their value. Some claims payers that use medical provider networks are reporting reductions of 20 to 50 percent in medical costs and duration of disability.

To get these results, claims payers say they need to evaluate and promote high-performing doctors. One of the holy grails of workers' compensation reform among employers and insurers is to wring from state legislators the right to pick the doctors who treat injured workers.

As many as 28 states today set limitations on employee choice of initial provider. Most of the states with strong employer choice, such as California, Connecticut, Florida, Georgia, Indiana, New Jersey and Texas, had to wage legislative reform battles to create the legal provision.

Payers -- insurers, TPAs and employers -- assert they can use claims data and network contracting to find and promote use of superior doctors far better than can individual injured workers on their own.

But for years, the proof of impact of provider networks was missing in action. One reason for the absence of proof is that not many people could use data to prove that a well-chosen provider network actually did what it was supposed to do: cut the duration of work disabilities and remove unnecessary medical expenses from the system.

That was then. The pace of research on the value of careful provider selection has picked up in recent years. The expertise to measure provider performance is now more available.

Except for a few state funds, most claims payers are coy about revealing how well their provider network management can deliver results. According to Mark Pew, senior vice president of Prium, a managed care company, "public comments about the details of network management might stir some distrust among medical providers about the payer."

However, apart from the recalcitrance of payers, researchers have scrubbed medical and claims data to create an accurate picture of costs and outcomes on a per-doctor basis. The findings in some half-dozen published studies are remarkably consistent among far-flung states and uniformly support the use of carefully designed networks.

In the studies, some doctors emerge with the good reputation of generating low medical costs and early return to work. Other doctor profiles are tainted by high medical costs and prolonged patient disability.

The New Hampshire Workers' Compensation Advisory Council has drawn upon these studies in drafting a law for a pilot program that authorizes employers to select a panel of providers for the first 90 days after an injury. Thomas Callahan, chairman of the council, said "better cooperation between employers and medical providers will improve access and treatment outcomes. And it helps resolve medical reimbursement issues."

RESEARCH STUDIES ON NETWORKS

In a recent issue of the Journal of Occupational and Environmental Medicine, the Louisiana Workers' Compensation Corp. reported on the performance of a small, premium quality network of medical providers, OMNET Gold. The network contained only 131 physicians at the time of the study.

This premium quality network, compared with the insurer's other providers in this employee-choice state, brought lost time injured workers back to work 45 percent faster, with claims costs that were 46 percent lower.

LWCC's findings arose from what Larry Yuspeh, LWCC's director of strategic risk and strategy management, called an "extensive and ongoing participation in research with The Johns Hopkins University School of Medicine."

Texas enacted in 2005 a law authorizing claims payers to introduce Health Care Networks (HCN), providing a measure of employer choice whereas the state before was employee choice. The state's Workers Compensation Research and Evaluation Group recently reported that networks generally had lower medical costs, lower utilization and higher return-to-work rates than achieved by non-network care.

California became a base for network development after the introduction in 2004 of a medical provider network (MPN), provision giving substantially all provider selection authority to the employer. Broadspire and Sedgwick, two large third-party administrators, have summarized their experience in MPNs in California as creating savings of around 20 percent. The two companies presented their findings at the National Workers' Compensation and Disability Conference® & Expo in Las Vegas in 2011.

"The average paid total per claim is 20 percent less, and for lost time claims the costs decline by 30 percent. That is done by partnering with and delivering care through the highest performing providers. We treat the injured worker as a consumer of health care. Diagnostics, pharmacy and doctor spends are less. Litigation is less. We can get these results in employer choice as well as employee choice states," said Kimberly George, senior vice president, managed care practice and client service for Sedgwick Claims Management Services Inc.

Erica Fichter, senior vice president for medical management at Broadspire, said that its in-house integrated model of claims and medical management helps get workers into using their networks. They use data mining to "pinpoint providers with successful utilization outcomes," resulting in a 20 percent reduction in loss costs.

How do these and other claims payers do it? By more careful selection of physicians and specialties. The range of performance among doctors is significant, which other studies lucidly documented. Experience or a specialty in workers' compensation also seems to be a significant factor.

For instance, in 2003, the California Workers' Compensation Institute found that the experience of the provider in treating work injuries had a huge effect on outcomes. It reported that the claims treated by providers with very large injured worker caseloads were 56 percent less costly than the claims treated by providers with smaller caseloads.

The average duration of disability for most experienced providers was 17 days vs. 30 days for the least experienced providers.

The CWCI did another study in 2011 on patterns in prescribing powerful pain killers, or Schedule II drugs such as OxyContin, to injured workers. It reported that a very small number of physicians, around 3 percent, were responsible for more than 65 percent of the Schedule II payments in California. Over half of Schedule II prescriptions were used to treat minor back injuries.

In 2010, the Louisiana Workers' Compensation Corp. and researchers at Johns Hopkins published an analysis of about 1,000 medical providers, looking at variations in duration of disability and medical costs.

The result was stunning. The analysis found that less than 4 percent of treating specialists caused the entire claims costs of the insurer to increase by half without any evidence that those physicians took on more severe cases.

None of these studies leaned upon negotiated priced discounts from providers as the source of claims savings, a traditional and increasingly discredited way to cut costs.

AVOIDING PROVIDERS

Most claims payers are not saying anything specific in public about their success with provider networks. This is partly stemming from competitive secrecy. It is also because they don't want to attract the wrath of critics of selective provider networks.

Consider how a network might use the doctor profiles to cut doctors out. A co-author of several medical provider network studies, Dr. Edward Bernacki of Johns Hopkins, minces no words about how to improve provider networks. He thinks it is not worthwhile trying to help some poor performers to improve. "It is very difficult to turn people around; it is better not to include them in your network. In fact, you probably do not need a network if these outlier or cost intensive physicians can be avoided."

Some individual providers and their professional associations are hyper-alert to the threat medical provider networks present to their earning ability. A decision by a claims payer to exclude a provider from its network based on an incomplete profile could easily flare up into a controversy, or more prohibitively, a lawsuit.

In California, battle was joined in 2008 between an occupational medicine clinic, Palm Medical Group, and California's State Compensation Insurance Fund in a suit brought by the clinic when the fund refused it admission into its provider network.

Palm Medical claimed that SCIF's behavior amounted to a violation of the state's populist "fair procedure" law and won a damage judgment of more than $1 million.

Margaret Wagner, CEO of Signature Networks Plus, of Cypress, Calif., a network developer, notes that MPN managers need to be very attentive to consistency in their review of provider performance and in admission decisions. State law requires the managers to disclose to state regulators their "economic profiling" practices, in which they measure the economic impact of clinical decisions, such as disability durations and total medical costs. The disclosures need, according to state law, to cover "policies and procedures that are used in utilization review, peer review, incentive and penalty programs, and in provider retention and termination decisions."

A defensible economic profile of doctors needs more accurate data on individual providers than most claims databases can deliver. Some California-based analytic firms such as Wagner's, Harbor Health and Axiomedics have grown to meet claims payer needs for MPN development. The analytic techniques honed there could be applied throughout the country. Other firms that support provider performance analysis in the workers' compensation space are Enclarity, Acrometis, a spin off from Medrisk, and Medmetrics.

Jeffrey Austin White, director of medical management practices and strategy for Lansing, Mich.-based Accident Fund Holdings, said his employer, with operations in more than 20 states, has been working on the accuracy of provider data for more than three years.

Since 2009, the use of individual provider identifiers, called national provider identifiers or NPI, went from 65 percent to more than 90 percent of providers, due to federal government mandates for electronic billing. For White, these identifiers are "essential" for evaluating provider experience.

White lays out three possible strategies to make a provider network more valuable, each requiring accurate tracking of doctor performance.

One option, mirroring the work of Johns Hopkins' Bernacki, with whom he co-authored a recent article on opioid use in Michigan, is to "avoid the poor performers."

Or, he says, "Go for middle of the curve, encourage the above average doctor, and not be so confrontational with the poor performers." Or third, "hand pick the best providers, probably those most highly trained and experienced with workplace injuries."

With proof of the value of provider networks, and the proliferation of expertise in doctor profiling, the atmosphere appears to be improving substantially for the creation of medical provider networks.

PETER ROUSMANIERE can be reached at .

December 18, 2012