Predictive modeling: property/casualty writers can learn to offset rising medical costs by importing a simple profitability tool fromgroup health.

In the race to sustain profitability, property/casualty insurers might take note of the new ways group health is containing ever-escalating medical costs. This could be especially helpful to workers' compensation writers, who have had to accept the line's increasing claims severity.

The medical care portion of lost-time claims--claims serious enough to require time off from work--increased by an estimated 10.5% in 2004, according to the National Council on Compensation Insurance. The average yearly increase for the eight-year period 1996-2003 was 9.3%, compared with 3.9% over the period 1991-1995.

And medical costs are likely to continue rising as newer, more expensive prescription drugs, particularly for pain management, are used to treat injured workers and doctors significantly increase their use of hospital services and complex diagnostic tools, such as MRIs.

Insurers are focusing on their claims adjudication to contain rising severity. Carriers are adopting expert systems to help evaluate and manage each bodily injury or workers' comp case. Benchmarking tools such as the ISO Claims Outcome Adviser database include more than 13,000 medical conditions and 14,000 occupations.

But as health-care costs continue climbing at an ever-steeper rate, more will be needed. Going back to 1950, the Consumer Price Index for medical care has increased 1,950%--almost three times more than the 680% increase for all items in the index.

No wonder workers' comp writers have struggled to achieve underwriting profitability. The workers' comp combined ratio--a key measure of losses and other expenses per dollar of premium--has lagged significantly behind the combined ratio of property/casualty as a whole. In 2004, while the property/casualty industry's combined ratio improved to 98.1 (a two-point improvement from 2003), workers' comp writers were at an estimated 105, meaning they were paying out $1.05 for every dollar of premium they took in. And that was the best underwriting result for the workers' comp line since 1997.

To compete successfully, insurers will have to do a better job of predicting high-risk cases that are likely to become high-expense cases.

Predictive modeling has taken off as a cost-containment solution in group health. Analytic software based on medical and pharmacy claims information is now being used to predict future health-care needs. Health plans, providers and employers are using this new technology for underwriting, medical management and performance evaluation.

ISO's DxCG unit has developed a modeling capability based on the superior predictive power of diagnostic and drug data to determine future health-care costs and hospital utilization. Scores produced by this analytic capability also predict whether the duration of a claim may extend over long periods of time.

Typically, when a worker gets hurt on the job, insurers look at employment conditions and extent of injury to determine time out of work. By adding comorbidity (other existing medical conditions that may affect the injured person's recovery), a model can provide a more accurate projection of time lost and time in the hospital. A person with a history of heart disease. for instance, may take longer and require more extensive treatment and medication to recover from a workplace injury than someone who is healthier. An obese person is statistically more likely to be out of work longer recovering from a knee or hip injury than someone whose weight is normal.

Modeling reveals a great deal about the likely trajectory the cost of a claim will take. Knowing how comorbidity will affect recovery costs can be a valuable aid to an insurer by helping to determine the value of a case and whether to implement an intensive recovery program--instituting aggressive treatment and closely managing and tracking the patient's therapy.

Predictive health-care modeling can help insurers explain settlement values to employers with high deductible and experience-rated plans. Other uses include helping establish a triage program to segregate claims according to their complexity, so the most complex go to senior claims handlers.

As the group health industry continues to profit from the benefits of predictive medical modeling, property/casualty insurers are sure to apply this new capability to tame their own raging medical costs.

Frank J. Coyne, a Best's Review columnist is chairman, president and chief executive officer of ISO. He can be reached at [e-mail].