EVENTS

Date:	Thursday, March 31, 2005
Time:	9:15 AM -- 11:30 AM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

March 2005

Is There a Crisis in Medical Malpractice? New Evidence from Texas

The president, several governors, and most physicians are again calling for reform of the medical malpractice system. Proponents of reform claim that the periodic spikes in medical malpractice insurance premiums are threatening to drive physicians out of practice and can only be reduced by limitations on malpractice awards and other changes to the current system. Opponents of reform maintain that there is no crisis and that increases in insurance premiums are caused by low returns on investments and other factors internal to the insurance industry, not by unreasonable malpractice settlements. This March 31 AEI Health Policy Discussion presented a new study of medical malpractice outcomes based on fifteen years of experience in Texas. The study's authors argue that settlements are not the main factor driving soaring malpractice premiums. Do their data support this finding? Following their presentation, three experts on medical malpractice--an economist, a lawyer, and an insurance executive--gave their own assessment of the new study and the case for medical malpractice reform.

David Hyman
University of Illinois

Medical malpractice has caused sharp political controversy between Republicans and Democrats, and between doctors and lawyers. Our recent study of Texas’s medical malpractice claiming system has provoked widespread interest and some vehement criticism.

The data used comes from a public database maintained by the Texas Department of Insurance (TDI). All available years with reliable data were investigated and reported. The project was funded by the authors’ home universities, and not by either side in the malpractice debate. Among the four authors, there are two Democrats and two Republicans, though politics was immaterial to this undertaking. The study’s results were surprising even to the authors, who expected to find more significant changes.

This project focused on Texas because the TDI makes comprehensive data available on all closed insured malpractice claims. Data from Florida are also publicly available, but another team of researchers is studying that state. In 1999, Texas began to experience a serious spike in malpractice premiums, and in 2002, the American Medical Association identified Texas as experiencing a malpractice crisis. Texas has the second biggest population, the third largest health care budget, and the highest uninsured rate of any state in the United States. It has long been viewed as plaintiff-friendly, though in 2003, it adopted various tort reforms, including caps on non-economic damages.

The three largest liability insurers in Texas raised rates an average of 110 percent from 1999 to 2003. The largest percentage rate increase was observed for TMLT, a physician-owned insurance company. The sharp premium increases observed in Texas are characteristic of other states as well.

Can soaring malpractice premiums be explained by problems in the claiming system? Public discussions often point to rising numbers of claims, skyrocketing settlements, and out-of-control juries. Our analysis of Texas’s experience with malpractice claims provides no support for any of these theories.

The TDI has collected information on closed claims since 1988. For claims resulting in payouts greater than $10,000, the department maintains a detailed individual record, with even greater detail available on claims larger than $25,000. Smaller claims and zero-payment claims are reported as aggregates. The TDI database includes information on settlements, jury verdicts, defense costs, and other pertinent information. Studying closed claims instead of lawsuits provides a more complete picture of medical malpractice. Most settlements are not publicly known, and jury verdicts are rare. While most jury verdicts favor defendants, a few outlier plaintiff awards tend to distort the public’s impression of the performance of the medical malpractice system.
     
Bernard Black
University of Texas

Our analysis found no time trend in the number of claims with large payouts (greater than $25,000 in 1988 dollars) in Texas when adjusted for population growth. While increasingly intensive health care use might have induced greater malpractice claims, the pattern becomes flatter when adjusted for the number of physicians or real health care spending--two imperfect proxies for health care utilization. Total claims adjusted for population were stable from 1995 to 2002, and claims per one hundred physicians actually declined over the same period. Paid claims per one hundred physicians also declined from 1990 on, though large paid claims per one hundred physicians remained flat over the period. This finding suggests that smaller claims are being forced out of the medical liability system.

Data from Texas also showed stability in average awards for large claims over time. Moreover, payouts of over $1 million occupied a consistent share of all claims from 1988 to 2002--roughly 6 percent of payouts greater than $10,000.

Total annual awards, driven by both the frequency of claims and the average payout, was stable and even declined by some measures. Adjusting for Texas’s gross state product revealed a downward trajectory for total malpractice claim payouts. This measure is the best gauge of the social burden of medical malpractice. Jury awards were highly variable, but a per-claim analysis suggests an increase of 2.5 to 3.6 percent each year.

Unlike various other metrics that proved surprisingly stable, defense costs per large paid claims increased from 1988 to 2002. Over fourteen years, defense costs increased by 80 percent in real terms. The total cost per large paid claim increased about 1 percent each year, mostly due to rising defense costs, since payouts were relative stable. The aggregate cost of all claims, including the cost of defending against them, revealed no time trend.

This study is limited by its focus on closed claims, which are broader in scope than lawsuits but omit data on open (unresolved) claims. Since TDI’s database includes only insured claims, self-insured groups like the large University of Texas hospital system are excluded from the study. Data on defense costs for zero and small payout claims are also unavailable. The study is limited to Texas, and it does not consider county-level trends. The TDI database does not provide specialty-level detail, obscuring any trends within and across physician specialties. The study’s objective was not to explain the rise in malpractice premiums, but rather to probe some common explanations. Liability caps, adopted in 2003, are outside the window observed for this study, so their impact is not measured. The starting year varies somewhat across the different measures employed in the study. This was done to ensure the reliability and completeness of the data considered and does not affect the findings presented in the study.

Many of the criticisms of this study target our adjustments for inflation and population, but it is necessary to make these adjustments to determine whether trends are attributable to factors external to the malpractice claiming system. Many of the claims made by other commentators on medical malpractice are unreliable because they fail to make such adjustments. For instance, if payout trends fail to account for the decline in the number of smaller paid claims, the average payment per claim increases by 40 percent between 1988 and 2002. Without adjusting for inflation, that measure jumps to 112 percent. Malpractice insurers and tort reform advocates have an obvious incentive to report the larger figure, despite its obvious biases.

Other critics object to the exclusion of open claims data. Those data are unavailable, but the paper discusses several reasons why we do not believe the unavailability of open claims data undermines our findings. If the number of open claims were rising dramatically, insurers could disclose those data, and they have not done so. 

The absence of data on defense costs for zero or small payment claims could impact our results. Small claim defense costs are probably rising (as are large claim defense costs), but their number is declining. If the cost of defending zero and low payout claims were included in estimates of the total annual cost of malpractice, those estimates would likely rise, though slowly.

We plan to use the TDI data for further research investigating jury verdicts, different types of damage awards, county-level trends, the impact of liability caps, and various other issues.

Medical malpractice premium spikes in Texas were not caused by an increase in either the number of claims or the size of the awards. Our conclusion that claiming practices in Texas were stable is strengthened if one adjusts for medical intensity.  Malpractice premium spikes may have resulted from insurance market dynamics such as low investment returns, stress in reinsurance markets, volatility in claims projections, or the vulnerability of an undiversified business, like malpractice. Tort reforms, then, may not affect the current trajectory of medical liability premiums.

Better data is critically important to devising effective solutions. Other states should collect the same type of data as Texas, and those states that currently collect such data but do not publicly release it should make it available to researchers.  

Randall R. Bovbjerg
Urban Institute

The medical malpractice debate has long been dominated by doctors, lawyers, and insurers. Policymakers seldom hear other voices. In particular, patients are overlooked, though they occupy the most central role. Not only do they finance the liability system through their taxes and health insurance premiums, but they are the ones ultimately protected--or not--by the liability system.

The authors’ work represents a tremendous undertaking in a difficult field. The trends they report for Texas closed malpractice claims seem quite similar to those seen in the closed, paid physician claims of the National Practitioner Data Bank. Their conclusions are also a useful reminder that policies made after declarations of “crisis” can indeed be shortsighted.

Nonetheless, the study has major shortcomings, some of which Hyman and colleagues recognize, though only in the small print. In the big print and in their op-ed column, they have trumpeted a major conclusion that seems premature--that Texas’s malpractice system is not in crisis, contrary to the evidence of rapidly rising premiums, and that President George W. Bush’s tort reform proposals are thus overly broad and misdirected. The data presented do not yet support such conclusions. Had this manuscript come to me as a peer reviewer, I would have recommended that it be revised and resubmitted.

The authors aim to explain why malpractice premiums in Texas shot up so drastically. They contrast with disfavor the sharp premium rises during 1999-2003 with the slow change they found in number of closed claims or average payout during 1990-2002 (after various adjustments). There are several problems with this approach. One is that those crying crisis do not generally blame a sharp rise in frequency of claims, except in occasional states, but rather point to a significant upturn in severity of claims.

A bigger problem is the presentation of data on trends in payouts. For premiums, rises in payouts must be projected forward, but closed claims like those in this paper look backward. For example, cases closed in 2002 consist almost entirely of injuries incurred and reported some years before. This is why insurers call malpractice a “long tail” line of coverage and why premium setting can be so difficult. Disproportionately, the slow-settling cases feature larger awards, and these bigger cases are disproportionately missing from the closed claims the authors observe in the crisis years since 1999. Put another way, too little time has passed for the large 2000-2002 cases to have been resolved.

The crisis years on the premium charts reflect insurers’ expectations that claims costs will rise rapidly in the years ahead. Those must be financed by premiums set now, but the “same” years in closed claims show mainly cases from earlier years, when premiums were lower. The authors’ presentation shows premiums consistently above payouts. It comes from a news article rather than the paper but is included to bolster the paper. As the authors recognize, premiums must exceed payouts in order for insurers to cover their other costs, including the sizeable costs of investigating and defending claims. The graphic, like the paper, appears to compare each year’s premiums with the same year’s claims payments instead of matching premiums with the corresponding long tails of payouts to come.

The authors would have benefited from discussing their general approach with insurance analysts before deciding on their methodology. They could not have expected insurers to share data on open claims or even on closed claims if compiled under promises of confidentiality, but analytical insights are shared more openly.

Hyman and colleagues report stable payouts from 1990 to 2002. Drawing from a much larger, multi-state database, the Physician Insurers Association of America reports significant (I believe unadjusted) growth in awards over that time. As seen in closed claims, it will take considerable time to tell which is right. Some other discrepancies might exist. The authors could be inadvertently mis-measuring claims rates, for instance, if the share of physicians covered by the reporting insurers has changed over time. Average payouts can also be affected by the policy limits that defendants buy, which may differ in Texas from elsewhere or may have changed over time. When analyzing claims, the paper oddly appears to lump together multiple claims from the same incident, which could affect trends if the number of defendants per case varies over time.

Moreover, if claims trends were so stable, why were insurers leaving Texas rather than flooding into the state? Why, over time, would insurance premiums rip off doctors with excessive rates, when the dominant insurer in Texas is physician-owned?

This study does very well to expose the need for better data, and its publicity may prompt greater disclosure. As in health care, we should pursue “evidence-based” policymaking for law and insurance, but public data are deficient for both sectors. Hyman and colleagues lay out a promising research agenda for future work.

Jonathan Klick
AEI and Florida State University

More empirical research into medical liability is valuable, but there is a serious disconnect between the authors’ hypothesis and their test. Their data is overly aggregated, which strips out most of the interesting variation, such as that seen across counties and physician specialties.

A measure that captures the variability of awards (which is absent from the authors’ design) has great explanatory power for premium changes over time.

During Texas’s crisis period, eleven insurers left the state, which affected one-third of physicians. Capacity constraints for the remaining insurers may have led to premium hikes. Relative to other states, Texas is losing physicians. Depending on the strictness of previous Texas regulations, physicians may have been able to repeatedly challenge increases in malpractice premiums. Their continual suppression may have motivated some insurers to leave Texas.

This paper would have benefited from two more years of data and analysis. It more closely resembles a pre-test analysis rather than a complete study. While its release has received much attention, it is better to be accurate than to be marketable.

Donald Zuk
SCPIE

SCPIE was founded thirty years ago as a physician-owned, not-for-profit insurance company. In 1997, it became a public company on the New York Stock Exchange. SCPIE has experience in both Florida and Texas’s liability insurance markets.

In 1998, SCPIE wrote seventeen insurance policies for physicians in Texas. That number swelled to 491 in 2000 and dropped again to forty in 2002. Facing loss ratios of 210 percent, SCPIE exited the Texas market in 2002, one year before the state adopted liability caps. The firm may consider reentry into Texas in the future.

While the frequency of malpractice claims being brought may indeed be down, their severity has increased dramatically. An actual case may illustrate the pressures that forced SCPIE out of Texas. An eighty-year-old patient that underwent a hip replacement experienced respiratory arrest after the procedure. After being revived, the patient claimed to have suffered memory loss. The claim was settled for $1 million because there were no caps imposed on damage awards at the time.

SCPIE was not the only insurance company to exit the Texas market. In 2000, the National Association of Insurance Commissioners named medical malpractice in Texas the least profitable business for insurers among the top fifteen states.

Liability reform discourages frivolous lawsuits, makes payouts more predictable and therefore premiums more affordable, and encourages competition. With tort reforms in place for thirty years, California has maintained a stable medical malpractice system. It currently has the lowest median claim payments and has experienced very moderate liability premium growth compared to the national average. Even in Texas, insurers are beginning to cut their rates as a result of the 2003 liability caps.

Before Texas enacted tort reform, the number of medical malpractice insurers dropped from seventeen to three. Since damage award caps were adopted in 2003, twenty-four insurers have contacted the TDI to inquire on operating within the state.

Medical malpractice torts cost $27 billion in 2003 and have grown at an alarming rate. By some estimates, caps on non-economic damages could save the federal government more than $50 billion each year. Recognizing their very positive record, many state legislators are adopting liability caps to contain skyrocketing malpractice costs.

Insurers do not raise premiums to increase their investment gains. Only 3 percent of SCPIE’s assets are invested in the stock market, and all insurers face limits on what share of their assets they can invest. Premiums are set only to capture future losses, not to make up investment shortfalls.

When premiums are stable, insurers face incentives to reduce them to compete with new businesses entering the market. Insurers should, as a rule, maintain premiums to ensure continued market stability.

AEI research assistant Ximena Pinell prepared this summary.