Health System Cuts Workers’ Comp With Best Docs
Posted on 02.09.18
Health systems can significantly reduce their workers’ compensation expenses by identifying and using high-performing physicians in medical networks, according to the experience of operators of one system in California, which saw overall claims costs reduced by almost 20%.
The premise of the program is that if one begins with treatment rendered by top-performing physicians who have demonstrated better outcomes, he or she will enhance the overall outcome of the entire book of business, says Linda Lane, president of Harbor Health Systems in Newport Beach, CA.
“Over time, that can lead to a significant impact with both a reduction in total medical spend and in total indemnity,” she says. “It’s a program that takes advantage of data analytics and allows our risk managers and payer partners to utilize what they may already have in place.”
Harbor Health Systems identified the highest-performing physicians and directed injured patients to them. Harbor manages or owns 154 medical provider networks (MPNs) nationwide, with 1.7 million patients enrolled. The company developed a proprietary software suite, MD 360, that examines 20 different variables within the MPNs treating workers’ compensation claimants.
Algorithms Identify Top Docs
The effort began in 2001 when the company founder’s physician wife was frustrated at the inability to promote her better outcomes with treating workers’ comp cases, Lane explains. MPNs seek to be broad and deep, but in past years they have not focused enough on outcomes data to allow one doctor’s results to stand out.
In response, Harbor Health asked a group of biostatisticians to create a proprietary set of algorithms that look at 20 points from three sets of data: claims data from a payer, network data to identify the treating physicians, and billing data to identify the treatment and billing patterns from those physicians.
“We mine those three data sources and run them through the algorithms to assess impact on disability days, total medical, litigation, and a number of other criteria,” she says. “The model allows us to assess one particular provider in a specialty, and the outcomes associated with that provider, in comparison to his or her peer group. We’re looking at a specialist in one geographic area compared to peers in that same area.”
Harbor Health does not evaluate medical skills of individual physicians, but uses a scoring method based on a retrospective analysis of closed claims, Lane explains. Among the criteria examined are cost of treatment, duration of treatment, patient recidivism, and litigation involving the claim. All the measures are objective, rather than including any subjective measure of medical performance.
“The elements that we use really rely on data being clean, and in our industry when you talk about billing and network data, that’s one of the biggest challenges,” Lane says. “Wrapped around our model is also a very stringent identity management piece, which is crucial to identifying that single individual doctor or that unique group. If you can’t do that, you can’t apply the measures.”
Not Easy to Compile Data
One challenge is accurately identifying physicians so one can associate their outcomes data. It is common for there to be a PPO network, but also other provider networks to ensure accessibility wherever a covered claim might exist, Lane notes. A particular doctor might be a member of one’s primary network, but also several others. In each of those networks, the same physician may be listed differently: John Smith, John A. Smith, J. Smith, J.A. Smith, J. Smith-Jones, etc. The same physician also may list multiple addresses and multiple associations with treatment groups.
“Getting down to that individual doctor and being able to collapse all the information about that individual into one set of data is the first critical piece of our work,” she says. “We’ve overcome that with some processes that cleanse that data to get to what we call a golden record, which allows us to know we’re talking about the same person no matter where data originates. You have to do the same thing with billing, which is just as challenging.”
The scoring model has continued to evolve, and recently Harbor Health began to factor in case mix — large loss claims vs. bumps and bruises. This accounts for the fact that some physicians treat only serious injuries, which can skew their outcomes data. Comorbidities such as diabetes and obesity also are considered.
Claims Costs Cut 20%
The goal is to assess the likelihood that a particular spine surgeon in one community will produce as good, better, or worse an outcome with a patient than other spine surgeons Harbor Health could choose instead. Then, clients use Harbor’s software suite to schedule patients for the higher-scoring physicians, and to manage and measure the outcomes.
The program also focuses on helping workers’ comp patients stay on track with treatment, Lane notes.
Harbor has cut overall claims costs by nearly 20% and indemnity-related costs by nearly one-third, compared to some other workers’ compensation management firms. In one instance in which a third-party administrator used Harbor Health’s platforms, closed claim percentage increased by 50% and the average paid claim amount dropped by 9.1% within nine months.
“This is a direct result of not allowing patients to go to any physician in a broad network, and instead directing them to top performing physicians,” Lane says. “You can effect change in these categories just by highlighting and utilizing those good habits that those top-performing physicians innately have.”
- Linda Lane, President, Harbor Health
Systems, Newport Beach, CA.
Telephone: (949) 536-7068.
Email: Linda Lane at firstname.lastname@example.org
A California health system has cut workers’ comp costs by concentrating on using the physicians with the best treatment results. The program examines 20 different variables.
- Overall claims costs dropped almost 20%.
- Indemnity-related costs dropped by nearly one-third.
- Cost of treatment is only one variable considered.