oversight

Medical Malpractice: Implications of Rising Premiums on Access to Health Care

Published by the Government Accountability Office on 2003-08-08.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

              United States General Accounting Office

GAO           Report to Congressional Requesters




August 2003
              MEDICAL
              MALPRACTICE
              Implications of Rising
              Premiums on Access
              to Health Care




GAO-03-836
                                                August 2003


                                                MEDICAL MALPRACTICE

                                                Implications of Rising Premiums on
Highlights of GAO-03-836, a report to           Access to Health Care
congressional requesters




The recent rising cost of medical               Actions taken by health care providers in response to rising malpractice
malpractice insurance premiums in               premiums have contributed to localized health care access problems in the
many states has reportedly                      five states reviewed with reported problems. GAO confirmed instances in
influenced some physicians to                   the five states of reduced access to hospital-based services affecting
move or close practices, reduce                 emergency surgery and newborn deliveries in scattered, often rural, areas
high-risk services, or alter their
practices to preclude potential
                                                where providers identified other long-standing factors that also affect the
lawsuits (known as defensive                    availability of services. Instances were not identified in the four states
medicine practices). States have                without reported problems. In the five states with reported problems,
revised tort laws under which                   however, GAO also determined that many of the reported provider actions
malpractice lawsuits are litigated to           were not substantiated or did not affect access to health care on a
help constrain malpractice                      widespread basis. For example, although some physicians reported
premium and claims costs. Some                  reducing certain services they consider to be high risk in terms of potential
of these tort reform laws include               litigation, such as spinal surgeries and mammograms, GAO did not find
caps on monetary penalties for                  access to these services widely affected, based on a review of Medicare data
noneconomic harm, such as for                   and contacts with providers that have reportedly been affected. Continuing
plaintiffs’ pain and suffering.                 to monitor the effect of providers’ responses to rising malpractice premiums
Congress is considering legislation
similar to some states’ tort reform
                                                on access to care will be essential, given the import and evolving nature of
laws.                                           this issue.

GAO examined how health care                    Physicians reportedly practice defensive medicine in certain clinical
provider responses to rising                    situations, thereby contributing to health care costs; however, the overall
malpractice premiums have                       prevalence and costs of such practices have not been reliably measured.
affected access to health care,                 Studies designed to measure physicians’ defensive medicine practices
whether physicians practice                     examined physician behavior in specific clinical situations, such as treating
defensive medicine, and how                     elderly Medicare patients with certain heart conditions. Given their limited
growth in malpractice premiums                  scope, the study results cannot be generalized to estimate the extent and
and claims payments compares                    cost of defensive medicine practices across the health care system.
across states with varying tort
reform laws. Because national data
on providers’ responses to rising               Limited available data indicate that growth in malpractice premiums and
premiums are not reliable, GAO                  claims payments has been slower in states that enacted tort reform laws that
examined the experiences in five                include certain caps on noneconomic damages. For example, between 2001
states with reported malpractice-               and 2002, average premiums for three physician specialties—general
related problems (Florida, Nevada,              surgery, internal medicine, and obstetrics/gynecology—grew by about 10
Pennsylvania, Mississippi, and                  percent in states with caps on noneconomic damages of $250,000, compared
West Virginia) and four states                  to about 29 percent in states with limited reforms. GAO could not determine
without reported problems                       the extent to which differences in premiums and claims payments across
(California, Colorado, Minnesota,               states were caused by tort reform laws or other factors that influence such
and Montana) and analyzed growth                differences.
in malpractice premiums and
claims payments across all states
and the District of Columbia.                   In commenting on a draft of this report, three independent reviewers with
                                                expertise on malpractice-related issues generally concurred with the report
                                                findings, while the American Medical Association (AMA) commented that
www.gao.gov/cgi-bin/getrpt?GAO-03-836.
                                                the scope of work was not sufficient to support the finding that rising
To view the full product, including the scope   malpractice premiums have not contributed to widespread health care
and methodology, click on the link above.       access problems. While GAO disagrees with AMA’s point of view, the report
For more information, contact Kathryn G.
Allen at (202) 512-7118.
                                                was revised to better clarify the methods and scope of work for this issue.
Contents


Letter                                                                                                 1
                       Results in Brief                                                                5
                       Background                                                                      8
                       Implications of Rising Malpractice Premiums on Access to Health
                         Care                                                                        12
                       Physicians Reportedly Practice Defensive Medicine, but
                         Prevalence and Costs of Such Practices Are Not Reliably
                         Measured                                                                    26
                       States with Certain Noneconomic Damage Caps Had Lower Recent
                         Growth in Malpractice Premium Rates and Claims Payments                     30
                       External Comments and Our Evaluation                                          38

Appendix I             National and State Provider Associations Contacted                            42



Appendix II            Scope and Methodology                                                         44



Appendix III           Summary of Selected Research Designed to
                       Measure Defensive Medicine Prevalence and Costs                               53



Appendix IV            GAO Contacts and Staff Acknowledgments                                        56
                       GAO Contact                                                                   56
                       Acknowledgments                                                               56

Related GAO Products                                                                                 57



Tables
                       Table 1: Assertions of Numbers of Physicians Moving, Retiring, or
                                Closing Practices in Response to Malpractice Pressures in
                                Five States                                                          17
                       Table 2: State Provider Associations GAO Contacted                            42
                       Table 3: Tort Reforms and Average Rates of Premium Increases in
                                Nine States                                                          45
                       Table 4: State Tort Reform Categories, Based on Reforms in Place
                                as of 1995                                                           49


                       Page i                   GAO-03-836 Medical Malpractice and Access to Health Care
          Table 5: Summary of Selected Research Designed to Measure
                   Defensive Medicine Prevalence and Costs                              53


Figures
          Figure 1: Rates of Medicare-Covered Orthopedic Surgeries in
                   Pennsylvania Have Increased                                          19
          Figure 2: Rates of Medicare-Covered Spinal Surgeries in Five States
                   with Reported Problems Have Recently Increased                       22
          Figure 3: Rates of Medicare-Covered Joint Revisions and Repairs in
                   Five States with Reported Problems Have Not Recently
                   Declined                                                             23
          Figure 4: Rates of Medicare-Covered Mammograms in Florida and
                   Pennsylvania Remain above the National Average                       24
          Figure 5: Premium Rates for Three Physician Specialties Rose After
                   2000, but to a Lesser Extent in States with Noneconomic
                   Damage Caps                                                          32
          Figure 6: Recent Premium Growth Was Lower for Three Physician
                   Specialties in States with Noneconomic Damage Caps                   33




          Page ii                  GAO-03-836 Medical Malpractice and Access to Health Care
Abbreviations

AHA               American Hospital Association
AMA               American Medical Association
CBO               Congressional Budget Office
CMS               Centers for Medicare & Medicaid Services
ER                emergency room
FSMB              Federation of State Medical Boards
HCPCS             Health Care Common Procedure Coding System
HEALTH            Help Efficient, Accessible, Low-cost, Timely Healthcare
                   Act of 2003
HHS               Department of Health and Human Services
ISO               Insurance Services Office
MLM               Medical Liability Monitor
MMCC              Medicare Managed Care Contract
NAIC              National Association of Insurance Commissioners
NCSL              National Conference of State Legislatures
NPDB              National Practitioner Data Bank
OB/GYN            obstetrics/gynecology
OTA               Office of Technology Assessment
PCF               patient compensation fund
PIAA              Physician Insurers Association of America




This is a work of the U.S. government and is not subject to copyright protection in the
United States. It may be reproduced and distributed in its entirety without further
permission from GAO. However, because this work may contain copyrighted images or
other material, permission from the copyright holder may be necessary if you wish to
reproduce this material separately.




Page iii                      GAO-03-836 Medical Malpractice and Access to Health Care
United States General Accounting Office
Washington, DC 20548




                                   August 8, 2003

                                   The Honorable F. James Sensenbrenner, Jr.
                                   Chairman
                                   Committee on the Judiciary
                                   House of Representatives

                                   The Honorable W.J. “Billy” Tauzin
                                   Chairman
                                   Committee on Energy and Commerce
                                   House of Representatives

                                   The Honorable Steve Chabot
                                   Chairman
                                   Subcommittee on the Constitution
                                   Committee on the Judiciary
                                   House of Representatives

                                   Medical malpractice insurance premium rates increased rapidly in some
                                   states beginning in the late 1990s after several years of relative stability,
                                   similar to previous cycles of rising premiums that occurred during the
                                   1970s and 1980s. Between 2001 and 2002, premium rates for the specialties
                                   of general surgery, internal medicine, and obstetrics/gynecology (OB/GYN)
                                   increased by about 15 percent on average nationally, and over 100 percent
                                   for certain of these specialists in some states. In response to these rising
                                   premiums, representatives of health care providers—including physicians,
                                   hospitals, and nursing homes—and the media have reported that
                                   physicians have moved out of states experiencing the highest increases,
                                   retired, or reduced or eliminated certain high-risk services. Policymakers
                                   are concerned that, if these provider actions are occurring, they may limit
                                   consumers’ access to health care. Additionally, fear of malpractice
                                   litigation may encourage physicians to practice “defensive medicine,” for
                                   example, ordering additional tests or procedures, thus increasing total
                                   health care costs. In an effort to mitigate rising malpractice costs, states
                                   have passed various tort reform laws, some of which include caps to
                                   restrict the size of damage award payments and other measures to limit




                                   Page 1                    GAO-03-836 Medical Malpractice and Access to Health Care
costs associated with malpractice litigation, and Congress is considering
similar federal legislation.1

Because of your concerns about rising malpractice insurance premiums
and associated implications for the health care system, we agreed to
examine the following questions:

1. How have health care provider responses to rising malpractice
   insurance premiums affected consumers’ access to health care?

2. What is known about how rising premiums and fear of litigation cause
   health care providers to practice defensive medicine?

3. How does the growth in medical malpractice insurance premiums and
   insurer payments for malpractice claims compare in states with
   varying levels of tort reform laws?

GAO also recently issued a related report that more fully describes the
extent of malpractice insurance premium growth and the factors that
contributed to that growth.2 Its findings are summarized on pages 9
through 11 of this report.

To evaluate how actions taken by health care providers in response to
malpractice premium increases have affected consumers’ access to health
care, we interviewed providers and their representatives, including the
American Medical Association (AMA), the American Health Care
Association, the American Hospital Association (AHA), and many of their
state-level counterparts. (See app. I for the complete list of national and
state associations we contacted during the course of our work.) In the
absence of reliable national sources of data concerning provider responses
to rising malpractice premiums, we focused our review on nine states
selected to encompass a range of malpractice premium pricing and tort


1
 Medical malpractice lawsuits are generally based on principles of tort law. A tort is a
wrongful act or omission by an individual that causes harm to another individual. Typically,
a legal claim of malpractice would be based on a claim that the negligence of a provider
caused injury and the injured party would seek damages. To reduce malpractice claims
payments and insurance premiums and for other reasons, some have advocated changes to
tort laws, such as placing caps on the amount of damages or limits on the amount of
attorney fees that may be paid under a malpractice lawsuit. These changes are collectively
referred to as “tort reforms.”
2
U.S. General Accounting Office, Medical Malpractice Insurance: Multiple Factors Have
Contributed to Increased Premium Rates, GAO-03-702 (Washington, D.C.: June 27, 2003).




Page 2                         GAO-03-836 Medical Malpractice and Access to Health Care
reform environments. Five of these states were among those cited by AMA
and other national health care provider organizations as malpractice
“crisis” or “problem” states based on such factors as higher than average
increases in malpractice insurance premium rates, physicians’ reported
difficulties obtaining malpractice insurance coverage, and reports of
actions taken by providers in response to the malpractice-related
pressures of rising premiums and litigation. The remaining four states
were not cited by provider groups as experiencing malpractice-related
problems.3 In the five states with reported problems, provider
organizations reported through surveys and anecdotal reports several
actions taken by physicians in response to rising malpractice premiums.
Although we did not attempt to confirm each report cited by state provider
groups, we targeted follow-up contacts with local providers where the
reports suggested potentially acute consumer access problems or where
multiple reports were concentrated in a geographic area. In these five
states, we contacted 49 hospitals and 61 physician practices or clinics to
corroborate the reports and explore the implications for consumers’
access to health care. We also analyzed Medicare part B physician claims
data from 1997 through 2002 to assess whether utilization of certain
services deemed to be of higher risk for a malpractice claim, such as spinal
surgery and mammograms, has declined for the Medicare-covered
population.4 Because of limitations in the Medicare data that precluded its
use in analyzing utilization of certain other physician services such as
hospital emergency and obstetrical care, we relied exclusively on the
reports of access problems provided by state provider associations and
our follow-up with local providers to assess access to these services.

To determine what is known about the extent of defensive medicine
practices, we reviewed available empirical studies, including those
examining the costs of defensive medicine and the potential impact of tort
reform laws on mitigating these costs. We also explored the issue with
medical provider organizations and examined the results of recent surveys
in which physicians were asked about their own defensive medicine
practices.




3
 The five states with reported problems are Florida, Mississippi, Nevada, Pennsylvania, and
West Virginia; the four states without reported problems are California, Colorado,
Minnesota, and Montana.
4
 Part B of the Medicare program covers claims for services provided by physicians, while
part A covers claims from hospitals and other institutions.




Page 3                        GAO-03-836 Medical Malpractice and Access to Health Care
To assess premium growth, we analyzed malpractice premium rates
reported by insurers to the Medical Liability Monitor (MLM) for the
specialties of general surgery, internal medicine, and OB/GYN—the only
three specialties for which these data are reported—across all states and
the District of Columbia from 1996 through 2002.5 To assess growth in
malpractice claims payments, we analyzed state-level data on claims paid
on behalf of all physicians reported by insurers to the National
Practitioner Data Bank (NPDB) from 1996 through 2002 for all states and
the District of Columbia.6 We compared trends in 12 states with tort
reforms that include caps on noneconomic damages, such as for plaintiffs’
pain and suffering (4 states with a $250,000 cap and 8 states with a
$500,000 or less cap7) to 11 states (including the District of Columbia) with
more limited tort reforms and to the average for all states. We focused our
analysis on those states with noneconomic damage caps as a key tort
reform because such caps are included in proposed federal tort reform
legislation and because published research generally reports that such
caps have a greater impact on medical malpractice premium rates and
claims payments than some other tort reform measures. We also reviewed
available empirical studies that examined the relationship between tort
reforms and malpractice insurance premiums and claims payments.

We conducted our work from September 2002 through June 2003
according to generally accepted government auditing standards. Appendix
II provides more details about our scope and methodology, and a list of
related GAO products is included at the end of this report.




5
 MLM is a private research organization that annually surveys professional liability
insurance carriers in 50 states and the District of Columbia to obtain their base premium
rates for the specialties of internal medicine, general surgery, and OB/GYN. Annual survey
data were available through 2002.
6
 NPDB, under the jurisdiction of the Secretary of Health and Human Services, is a
nationwide source of information on physicians who have been named in a medical
malpractice settlement or judgment. Insurers are required by law to report malpractice
payments made on behalf of these physicians and are subject to civil penalties for
noncompliance. 42 U.S.C. § 11131 (2000).
7
 The eight states with a $500,000 or less cap do not include the four states with a $250,000
cap.




Page 4                         GAO-03-836 Medical Malpractice and Access to Health Care
                   Actions taken by health care providers in response to malpractice
Results in Brief   pressures have contributed to localized health care access problems in the
                   five states we reviewed with reported problems.8 We confirmed instances
                   in the five states where actions taken by physicians in response to
                   malpractice pressures have reduced access to services affecting
                   emergency surgery and newborn deliveries. These instances were not
                   concentrated in any one geographic area and often occurred in rural
                   locations, where maintaining an adequate number of physicians may have
                   been a long-standing problem, according to some providers. For example,
                   the only hospital in a rural county in Pennsylvania no longer has full
                   orthopedic on-call surgery coverage in its emergency room (ER) because
                   three of its five orthopedic surgeons left in the spring of 2002, largely in
                   response to the high cost of malpractice insurance. Similarly, pregnant
                   women in rural central Mississippi must now travel about 65 miles to the
                   nearest hospital obstetrics ward to deliver because family practitioners at
                   the local hospital, faced with rising malpractice insurance premiums,
                   stopped providing obstetrics services. In both areas, providers also cited
                   other reasons for difficulties recruiting physicians to their rural areas. We
                   did not identify similar examples of access reductions attributed to
                   malpractice pressures in the four states without reported problems. In the
                   five states with reported problems, however, we also determined that
                   many of the reported provider actions taken in response to malpractice
                   pressures were not substantiated or did not widely affect access to health
                   care. For example, some reports of physicians relocating to other states,
                   retiring, or closing practices were not accurate or involved relatively few
                   physicians. In these same states, our review of Medicare claims data did
                   not identify any major reductions in the utilization of certain services
                   some physicians reported reducing because they consider the services to
                   be high risk, such as certain orthopedic surgeries and mammograms.
                   Continuing to monitor the effect of providers’ responses to rising
                   malpractice premiums on access to care will be essential, given the import
                   and evolving nature of this issue.




                   8
                    We define loss of access as the direct loss or newly limited availability of a health care
                   provider or service resulting largely from actions taken by providers in response to
                   malpractice concerns. We did not assess the impact on access that may result from the
                   added costs malpractice pressures impose on the health care system (e.g., the combined
                   cost of malpractice insurance premiums, litigation, and defensive medicine practices) and
                   thus on the costs and affordability of health insurance because data to reliably measure
                   malpractice-related costs in total are not available.




                   Page 5                         GAO-03-836 Medical Malpractice and Access to Health Care
In response to rising premiums and their fear of litigation, research
indicates that physicians practice defensive medicine in certain clinical
situations, thereby contributing to health care costs; however, the overall
prevalence and costs of such practices have not been reliably measured.
Recent surveys of physicians indicate that many practice defensive
medicine, but limitations to these surveys suggest caution in interpreting
and generalizing the results. For example, the surveys typically ask
physicians if or how they have practiced defensive medicine but not the
extent of such practices. In addition, very few physicians tend to respond
to these surveys, raising doubt about how accurately their responses
reflect the practices of all physicians. Some empirical research has
identified defensive medicine practices, but under very specific clinical
situations that cannot be generalized more broadly. For example, one
study examined Medicare patients with two specified heart diseases and
concluded that certain tort reforms that reduce malpractice pressures,
such as caps on damages, may reduce hospital expenditures for treatment
of the two conditions by 5 to 9 percent. However, subsequent preliminary
research that expanded this study to additional Medicare patients with a
broader set of conditions did not find similar savings.

Limited available data indicate that rates of growth in malpractice
premiums and claims payments have been slower on average in states that
enacted certain caps on damages for pain and suffering—referred to as
noneconomic damage caps—than in states with more limited reforms.9
Premium rates reported for the specialties of general surgery, internal
medicine, and OB/GYN were relatively stable on average in most states
from 1996 through the late 1990s and then began to rise, but more slowly
among states with certain noneconomic damage caps. For example, from
2001 through 2002, average premium rates rose approximately 10 percent
in states with noneconomic damage caps of $250,000 compared with
approximately 29 percent in states with more limited tort reforms.
Although payments for claims against all physicians from 1996 through
2002 tended to be lower and grew less rapidly on average in states with
caps on noneconomic damages than in states with limited reforms, the
averages obscured wide variation in claims payments and rates of growth
across states and over time. Moreover, claims payments we reviewed were


9
 Damage caps may apply to three types of damages awarded to plaintiffs in a medical
malpractice suit: noneconomic damages, which compensate for harm that is not easily
quantifiable (such as pain and suffering); economic damages, which compensate for lost
wages and other financial harms; and punitive damages, which punish providers for
especially egregious conduct.




Page 6                        GAO-03-836 Medical Malpractice and Access to Health Care
limited to claims against physicians and did not include claims against
institutional providers such as hospitals and nursing homes. Differences in
both premium rates and claims payments are also affected by factors other
than damage caps, including the presence of other tort reform measures,
the presence of state laws regulating the premium rate-setting process,
and certain market forces, including the level of market competition
among insurers and interest rates that affect insurers’ investment returns.10
We could not determine the extent to which differences in premiums and
claims payments across states were attributable to states’ tort reform laws
or to these additional factors.

We received comments on a draft of this report from three independent
health policy researchers and AMA. Each of the researchers has expertise
in malpractice-related issues and has conducted and published research
on the effects of malpractice pressures on the health care system, and two
of the three are physicians. The health policy researchers generally
concurred with our findings. AMA, however, questioned our finding that
rising malpractice premiums have not contributed to widespread health
care access problems, expressing concern that the scope of our work
limited our ability to fully identify the extent to which malpractice-related
pressures are affecting consumers’ access to health care. We disagree that
the scope of our work limited our ability to identify the extent of
malpractice-related access problems. In the absence of current and
reliable national data on provider responses to rising malpractice
premiums, we used a variety of qualitative and quantitative methods as a
basis for our findings on the effect of provider actions on access to care in
the five states we reviewed with reported problems. While we did not
attempt to generalize our findings beyond these five states, we believe
that—because they are among the most visible and often-cited examples
of “crisis” states—the experiences of these five states provide important
insight into the overall problem. In response to AMA’s comments,
however, we clarified the report’s discussion of the scope of work and
methods used for this issue.




10
 For more information on the factors that influence malpractice premium rates, see
GAO-03-702.




Page 7                        GAO-03-836 Medical Malpractice and Access to Health Care
             In the United States, patients injured while receiving health care can sue
Background   health care providers for medical malpractice under governing state tort
             law, usually the law of the state where the injury took place. Laws
             governing medical malpractice vary from state to state, but among the
             goals of tort law are compensation for the victim and deterrence of
             malpractice.

             Nearly all health care providers buy medical malpractice insurance to
             protect themselves from potential claims that could cause financial harm
             or even bankruptcy absent liability coverage. For example, the average
             reported claims payment made on behalf of physicians and other licensed
             health care practitioners in 2001 was about $300,000 for all settlements,
             and about $500,000 for trial verdicts.11 Under a malpractice insurance
             contract, the insurer agrees to investigate claims, to provide legal
             representation for the health care provider, and to accept financial
             responsibility for payment of any claims up to a specified monetary level
             during an established time period. The most common policies sold by
             insurers provide $1 million of coverage per incident and $3 million of total
             coverage per year. The insurer provides this coverage in return for a fee—
             the medical malpractice premium.

             Medical malpractice premium rates differ widely by medical specialty and
             geography. Premiums paid by traditionally high-risk specialties, such as
             obstetrics, are usually higher than premiums paid by other specialties,
             such as internal medicine. Premium rates also vary across and within
             states. Across states, for example, a large insurer in Minnesota charged
             base premium rates of $3,803 for the specialty of internal medicine,
             $10,142 for general surgery, and $17,431 for OB/GYN in 2002 across the
             entire state.12 In contrast, a large insurer in Florida charged base premium
             rates in Dade County of $56,153 for internal medicine, $174,268 for general
             surgery, and $201,376 for OB/GYN, and $34,556, $107,242, and $123,924,
             respectively, for these same specialties in Palm Beach County. In addition
             to the wide range in premium rates charged, the extent to which premiums


             11
               See Physician Insurers Association of America (PIAA), PIAA Claim Trend Analysis,
             2001 Edition (Rockville, Md.: 2002). Averages are based on a compilation of medical
             malpractice claims data from more than 20 PIAA member companies that insure about 20
             to 25 percent of all physicians. Most claims are resolved out of court. Among the closed
             claims PIAA reviewed in 2001 that resulted in an award to plaintiffs, about 96 percent were
             closed through an out-of-court settlement and about 4 percent through a trial verdict.
             12
              Base premium rates exclude discounts, rebates, and surcharges that may affect the actual
             premium rate charged.




             Page 8                         GAO-03-836 Medical Malpractice and Access to Health Care
                            increase over time also varies by specialty and geographic area. Beginning
                            in the late 1990s, malpractice premiums began to increase at a rapid rate
                            for most, but not all, physicians in some states. For example, between 1999
                            and 2002, the Minnesota insurer increased its base premium rates by about
                            2 percent for each of the three specialties, in contrast to the Florida
                            insurer that increased its base premium rates by about 98, 75, and 43
                            percent, respectively, for the three specialties in Dade County.


Rising Claims Costs         Since 1999, medical malpractice premium rates for certain physicians in
Among Factors               some states have increased dramatically. In a related report issued in
Contributing to             June 2003, we examined the extent and causes of these recent increases.13
                            More specifically, we reported on (1) the extent of increases in medical
Malpractice Insurance       malpractice insurance rates in seven states,14 (2) factors that have
Premium Increases           contributed to the increases, and (3) changes in the medical malpractice
                            insurance market that may make the current period of rising premium
                            rates different from earlier periods of rate hikes. Key findings from that
                            report include the following.

                        •   Among the seven states we analyzed, the extent of medical malpractice
                            premium increases varied greatly not only from state to state but across
                            medical specialties. For example, among the largest writers of medical
                            malpractice insurance in the seven states, increases in base premium rates
                            for general surgeons from 1999 to 2002 ranged from 2 percent in
                            Minnesota to 130 percent in and around Harrisburg, Pennsylvania. Across
                            specialties, one carrier raised premiums for the area in and around El
                            Paso, Texas, during this period by 95 percent for general surgery, 108
                            percent for internal medicine, and 60 percent for OB/GYN.
                        •   Multiple factors have contributed to the recent increases in medical
                            malpractice premium rates. First, since 1998, the greatest contributor to
                            increased premium rates in the seven states we analyzed appeared to be
                            increased losses for insurers on paid medical malpractice claims.
                            However, a lack of comprehensive data at the national and state levels on
                            insurers’ medical malpractice claims and the associated losses prevented
                            us from fully analyzing the composition and causes of those losses.
                            Second, from 1998 through 2001, medical malpractice insurers
                            experienced decreases in their investment income as interest rates fell on
                            the bonds that generally make up around 80 percent of these insurers’


                            13
                             GAO-03-702.
                            14
                             The states are California, Florida, Minnesota, Mississippi, Nevada, Pennsylvania, and
                            Texas.




                            Page 9                         GAO-03-836 Medical Malpractice and Access to Health Care
    investment portfolios.15 While almost no medical malpractice insurers
    experienced net losses on their investment portfolios over this period, a
    decrease in investment income meant that income from insurance
    premiums had to cover a larger share of insurers’ costs. Third, during the
    1990s, insurers competed vigorously for medical malpractice business, and
    several factors, including high investment returns, permitted them to offer
    prices that, in hindsight for some insurers, did not completely cover their
    ultimate losses on that business. As a result of this, some companies
    became insolvent or voluntarily left the market, reducing the downward
    competitive pressure on premium rates that had existed through the 1990s.
    Fourth, beginning in 2001, reinsurance rates for medical malpractice
    insurers also increased more rapidly than they had in the past, raising
    insurers’ overall costs.16
•   While the medical malpractice insurance market as a whole had
    experienced periods of rapidly increasing premium rates in the mid-1970s
    and mid-1980s, the market has changed considerably since then. These
    changes are largely the result of actions insurers, health care providers,
    and states have taken to address increasing premium rates. Beginning in
    the 1970s and 1980s, insurers began selling “claims-made” rather than
    “occurrence-based” policies, enabling insurers to better predict losses for
    a particular year.17 Also in the 1970s, physicians, facing increasing
    premium rates and the departure of some insurers, began to form mutual
    nonprofit insurance companies. Such companies, which may have some
    cost and other advantages over commercial insurers, now make up a
    significant portion of the medical malpractice insurance market. More
    recently, an increasing number of large hospitals and groups of hospitals
    or physicians have left the traditional commercial insurance market and
    sought alternative arrangements, for example, by self-insuring. While such
    arrangements can save money on administrative costs, hospitals and
    physicians insured through these arrangements assume greater financial
    responsibility for malpractice claims than they would under traditional
    insurance arrangements and thus may face a greater risk of insolvency.
    Finally, since the periods of increasing premium rates during the mid-


    15
      State insurance regulators generally require insurers to reduce their requested premium
    rates in line with expected investment income. That is, the higher the expected income
    from investments, the more premium rates must be reduced.
    16
     Reinsurance is insurance for insurance companies, which insurance companies routinely
    use as a way to spread the risk associated with their insurance policies.
    17
      Claims-made policies cover claims reported during the year in which the policy is in
    effect. Occurrence-based policies cover claims arising out of events that occurred but may
    not have been reported during the year in which the policy was in effect. Most policies sold
    today are claims-made policies.




    Page 10                        GAO-03-836 Medical Malpractice and Access to Health Care
                                1970s and mid-1980s, all states have passed at least some laws designed to
                                reduce medical malpractice premium rates. Some of these laws are
                                designed to decrease insurers’ losses on medical malpractice claims, while
                                others are designed to more tightly control the premium rates insurers can
                                charge. These market changes, in combination, make it difficult to predict
                                how medical malpractice premiums might behave in the future.


States Use Tort Reform to       In order to improve the affordability and availability of malpractice
Help Contain Costs              insurance and to reduce liability pressure on providers, states have
Associated with Medical         adopted varying types of tort reform legislation.18 Tort reforms are
                                generally intended to limit the number of malpractice claims or the size of
Malpractice                     payments in an effort to reduce malpractice costs and insurance
                                premiums. Also, some believe tort reforms can lower overall health care
                                costs by reducing certain defensive medicine practices. Such practices
                                include the overutilization by physicians of certain diagnostic tests or
                                procedures primarily to reduce their exposure to malpractice liability,
                                therefore adding to the costs of health care.19 State tort reform measures
                                adopted during the past three decades include

                            •   placing caps on the amount that may be awarded to plaintiffs for damages
                                in a malpractice lawsuit, including noneconomic, economic, and punitive
                                damages;
                            •   abolishing the “collateral source rule” that prevents a defendant from
                                introducing evidence that the plaintiff’s losses and expenses have been
                                paid in part by other parties such as health insurers, or damage awards
                                from being reduced by the amount of any compensation plaintiffs receive
                                from third parties;
                            •   abolishing “joint and several liability” to ensure that damages are
                                recovered from defendants in proportion to each defendant’s degree of
                                responsibility, not each defendant’s ability to pay;
                            •   allowing damages to be paid in periodic installments rather than in a lump
                                sum;
                            •   placing limits on fees charged by plaintiffs’ lawyers;



                                18
                                  States have also experimented with approaches to constrain malpractice-related costs in
                                addition to tort reforms. For example, Virginia created a no-fault compensation program
                                for birth-related neurological injuries, and Maine temporarily used standardized clinical
                                practice guidelines to provide physicians with a defense against potential malpractice
                                lawsuits.
                                19
                                 Physicians may also reduce or eliminate certain services they believe place them at risk of
                                malpractice litigation. Such practices may also be referred to as defensive medicine.




                                Page 11                        GAO-03-836 Medical Malpractice and Access to Health Care
                         •   imposing stricter statutes of limitations that shorten the time injured
                             parties have to file a claim in court;
                         •   establishing pretrial screening panels to evaluate the merits of claims
                             before proceeding to trial; and
                         •   providing for greater use of alternative dispute resolution systems, such as
                             arbitration panels.

                             Among the tort reform measures enacted by states, caps on noneconomic
                             damage awards that include pain and suffering have been the focus of
                             particular interest. Cap proponents believe that such limits can result in
                             several benefits that help reduce malpractice insurance premiums, such as
                             helping to prevent excessive awards and overcompensation and ensuring
                             more consistency among jury verdicts. In contrast, cap opponents believe
                             that factors other than award amounts affect premiums charged by
                             malpractice insurers and that caps can result in undercompensation for
                             severely injured persons.

                             Congress is currently considering federal tort reform legislation that
                             includes several elements adopted by states to varying degrees, including
                             placing caps on noneconomic and punitive damages, allowing evidence at
                             the trial of a plaintiff’s recovery from collateral sources, abolishing joint
                             and several liability, and placing a limit on contingency fees, among
                             others.20


                             Actions taken by health care providers in response to rising malpractice
Implications of Rising       premiums have contributed to reduced access to specific services on a
Malpractice                  localized basis in the five states reviewed with reported problems.21 We
                             confirmed instances where physician actions in response to malpractice
Premiums on Access           pressures have resulted in decreased access to services affecting
to Health Care               emergency surgery and newborn deliveries in scattered, often rural areas
                             of the five states. However, we also determined that many of the reported
                             physician actions and hospital-based service reductions were not
                             substantiated or did not widely affect access to health care. For example,
                             our analysis of Medicare utilization data suggests that reported reductions


                             20
                              On March 13, 2003, the House of Representatives passed the Help Efficient, Accessible,
                             Low-cost, Timely Healthcare (HEALTH) Act of 2003 (H.R. 5); on June 27, 2003, a similar
                             version (S. 11) of this bill was introduced in the Senate.
                             21
                              Provider groups in the four states without reported problems neither cited nor provided
                             evidence of provider actions taken in response to malpractice pressures that could affect
                             consumer access to care.




                             Page 12                       GAO-03-836 Medical Malpractice and Access to Health Care
                            in certain high-risk services, such as some orthopedic surgeries and
                            mammograms, have not widely affected consumer access to these
                            services. To help avoid consumer access problems, some hospitals we
                            contacted have taken certain steps, such as assuming the costs of
                            physicians’ liability insurance, to enable physicians to continue practicing.


Health Care Provider        We confirmed examples in each of the five states where access to services
Actions Taken in Response   affecting emergency surgery and newborn deliveries has been reduced. In
to Malpractice Pressures    these instances, some of which were temporary, patients typically had to
                            travel farther to receive care. The problems we confirmed were limited to
Have Limited Access to      scattered, often rural, locations and in most cases providers identified
Certain Services in Some    long-standing factors in addition to malpractice pressures that affected the
Localities                  availability of services.

                        •   Florida: Among several potential access problems we reviewed in Florida,
                            the most significant appeared to be the reduction in ER on-call surgical
                            coverage in Jacksonville. We confirmed that at least 19 general surgeons
                            who serve the city’s hospitals took leaves of absence beginning in May
                            2003 when state legislation capping noneconomic damages for malpractice
                            cases at $250,000 was not passed. According to one hospital
                            representative, the loss of these surgeons reduced the general surgical
                            capacity of Jacksonville’s acute care community hospitals by one-third.
                            The administrator of the practice that employs these surgeons told us that
                            at least 8 are seeking employment in other states to avoid the high
                            malpractice premiums in Florida. Hospital officials in Jacksonville told us
                            that other providers, including some orthopedic surgeons and
                            cardiovascular surgeons, had also taken leave as of May 2003 due in part
                            to the risks associated with practicing without surgeons available in the
                            ER for support in the event of complications. According to one
                            Jacksonville area hospital official, her hospital has lost the services of 75
                            physicians in total due to leaves of absence taken by the physicians.
                            Hospital and local health department officials said that the losses of
                            surgeons have caused a reduction in ER on-call surgical coverage at most
                            acute care hospitals in the city; the health department official said patients
                            requiring urgent surgical care presenting at an ER that does not have
                            adequate capacity must be transferred to the nearest hospital that does,
                            which could be up to 30 miles away. Within the first 11 days after most of
                            the physicians took leave, 120 transfers took place.22 Although the hospital


                            22
                             Some providers have also reported reductions in certain nonurgent elective services that
                            may require surgical backup in the event of complications, such as cardiac surgery.




                            Page 13                       GAO-03-836 Medical Malpractice and Access to Health Care
    officials we interviewed expected that some of the physicians would
    eventually return to work, they believe timing may depend on passage of
    malpractice reform legislation during a special legislative session expected
    to take place this summer.
•   Mississippi: Reductions in ER on-call surgical coverage and newborn
    delivery services have created access problems in certain areas of
    Mississippi. We confirmed that some surgeons along the Gulf Coast who
    formerly provided on-call services at multiple hospitals are restricting
    their coverage to a single ER and others are eliminating coverage entirely
    in an effort to minimize their malpractice premiums and exposure to
    litigation. Officials of two of five hospitals we spoke with in the three Gulf
    Coast counties told us they have either completely lost or experienced
    reduced ER on-call surgical coverage for certain services. These
    reductions in coverage may require that patients be transferred greater
    distances for services. Some family practitioners and OB/GYNs have
    stopped providing newborn delivery services, creating access problems in
    certain rural communities. An official from one hospital in a largely rural
    county in central Mississippi told us that it closed its obstetrics unit after
    five family practitioners who attended deliveries stopped providing
    newborn delivery services in order to avoid a more than 65 percent
    increase in their annual premium rates. Pregnant women in the area now
    must travel about 65 miles to the nearest obstetrics ward to deliver. Loss
    of obstetrics providers in other largely rural areas may require pregnant
    women in these areas to travel farther for deliveries. A provider
    association official told us that malpractice pressures have worsened long-
    standing difficulties associated with recruiting physicians to the state, and
    providers also said that low Medicaid reimbursement rates and insufficient
    reimbursement for trauma services also influence physician practice
    decisions.
•   Nevada: Reductions in ER on-call surgical coverage have created access
    problems in Clark County. To draw attention to their concerns about
    rising medical malpractice premiums, over 60 orthopedic surgeons in the
    county withdrew their contracts with the University of Nevada Medical
    Center, causing the state’s only Level I trauma center to close for 11 days
    in July 2002.23 The center reopened after a special arrangement was made
    for surgeons to temporarily obtain malpractice coverage through the
    Medical Center and the governor announced his support for state tort
    reform, prompting the return of approximately 15 of the surgeons,
    according to medical center staff. Another hospital in the county has
    closed its orthopedics ward and no longer provides orthopedic surgical


    23
      Trauma centers are designated based on the level of service sophistication, with Level I
    trauma centers equipped to handle the most complex trauma cases.




    Page 14                        GAO-03-836 Medical Malpractice and Access to Health Care
    coverage in its ER as orthopedic surgeons have sought to reduce their
    malpractice exposure by decreasing the number of hospitals in which they
    provide ER coverage, according to a hospital official. Clark County has
    had long-standing problems with ER staffing due in part to its rapidly
    growing population, according to providers.
•   Pennsylvania: Some areas in Pennsylvania have experienced reductions
    in access to emergency surgical services and newborn delivery services.
    For example, one rural hospital recently lost three of its five orthopedic
    surgeons. As a result, orthopedic on-call coverage in its ER has declined
    from full-time to only one-third of each month. At the same hospital,
    providers reported that four of the nine OB/GYNs who provide obstetrical
    care in two counties stopped providing newborn delivery services because
    their malpractice premiums became unaffordable and another left the
    state to avoid high premiums. Some pregnant women now travel an
    additional 35 to 50 miles to deliver. According to a hospital official, the
    remaining four OB/GYNs were each in their sixties and near retirement.
    This hospital reported that the loss of the physicians was largely due to the
    rising cost of malpractice insurance, but also identified the hospital’s rural
    location, and the area’s large Medicaid population and low Medicaid
    reimbursement rates as factors contributing to the physicians’ decisions to
    leave. Trauma services in Pennsylvania have also been affected in some
    localities. For example, a suburban Philadelphia trauma center closed for
    13 days beginning in December 2002 because its orthopedic surgeons and
    neurosurgeons reported they could not afford to renew their malpractice
    insurance. The situation was resolved when a new insurance company
    offered more affordable coverage to the surgeons and the governor
    introduced a plan to reduce physician payments to the state medical
    liability fund, according to a hospital official.
•   West Virginia: Access problems due to malpractice concerns in West
    Virginia involved ER specialty surgical services. One of the state’s major
    medical centers lost its Level I trauma designation for approximately 1
    month in the early fall of 2002 due to reductions in the number of
    orthopedic surgeons providing on-call coverage. During this time, patients
    who previously would have been treated at this facility had to be
    transferred to other facilities at least 50 miles away. The hospital’s Level I
    designation was restored when additional physicians agreed to provide on-
    call coverage after the state extended state-sponsored liability insurance
    coverage to physicians who provide a significant percentage of their
    services in a trauma setting. The state’s northern panhandle lost all
    neurosurgical services for about 2 years when three neurosurgeons who
    served the area either left or stopped providing these services in response
    to malpractice pressures, requiring that all patients needing neurosurgical
    care be transferred 60 miles or more, limiting patients’ access to urgent
    neurosurgical care. Full-time neurosurgical coverage was restored to the



    Page 15                   GAO-03-836 Medical Malpractice and Access to Health Care
                           area in early 2003 through an agreement with a group of neurosurgeons at
                           one of the state’s major academic medical centers. A hospital official from
                           this area reported that efforts to recruit a permanent full-time
                           neurosurgeon have been unsuccessful. Provider groups told us that
                           malpractice concerns have made efforts to recruit and retain physicians
                           more difficult; however, they also identified the rural location, low
                           Medicaid reimbursement rates, and the state’s provider tax on physicians
                           as factors that have made it difficult to attract and retain physicians.24


Some Reported Provider     Despite some confirmed reductions in ER on-call surgical coverage and
Actions Were Not           newborn delivery services that were related to physicians’ concerns about
Substantiated or Did Not   malpractice pressures and affected access to health care, we also
                           identified reports of provider actions taken in response to malpractice
Widely Affect Access to    pressures—such as reported physician departures and hospital unit
Health Care                closures—that were not substantiated or that did not widely affect access
                           to health care. Our contacts with 49 hospitals revealed that although 26
                           confirmed a reduction in surgeons available to provide on-call coverage
                           for the ER, 11 of these reported that the decreases had not prevented them
                           from maintaining the full range of ER services and 3 reported that the
                           surgeons had returned or replacements had been found. Hospital
                           association representatives reported that access to newborn delivery
                           services in Florida had been reduced due to the closures of five hospital
                           obstetrics units. However, we contacted each of these hospitals and
                           determined that these units were located in five separate urban counties,
                           and each hospital reported that demand for its now closed obstetrics
                           facility had been low and that nearby facilities provided obstetrics
                           services.25 In West Virginia, although access problems reportedly
                           developed because two hospital obstetrics units closed due to malpractice
                           pressures, officials at both of these hospitals told us that a variety of
                           factors, including low service volume and physician departures unrelated



                           24
                             West Virginia’s health care provider tax was imposed in 1993 as a 2 percent tax on
                           physicians’ gross revenues. The tax is gradually being phased out and will be eliminated in
                           2010. The tax rate is currently 1.4 percent. According to AMA, only one other state has a
                           similar tax on physicians.
                           25
                             Each of the five hospitals that closed its obstetrics unit told us that demand for obstetrical
                           services in its community was low. One hospital reported that there was a greater need in
                           the community for additional emergency room beds than obstetrics beds, and two hospitals
                           reported that their obstetrics units were originally opened based on managed care contract
                           requirements even though there was not a clear need for obstetrics services at these
                           facilities.




                           Page 16                         GAO-03-836 Medical Malpractice and Access to Health Care
    to malpractice, contributed to the decisions to close these units. One of
    the hospitals has recently reopened its obstetrics unit.

    Provider groups also asserted that some physicians in each of the five
    states are moving, retiring, or closing practices in response to malpractice
    pressures. In the absence of national data reporting physician movement
    among states related to malpractice concerns, we relied on state-level
    assertions of departures that were based on a variety of sources, including
    survey results, information compiled and quantified by provider groups,
    and unquantified anecdotal reports. (See table 1.)

    Table 1: Assertions of Numbers of Physicians Moving, Retiring, or Closing
    Practices in Response to Malpractice Pressures in Five States

                                                  Orthopedic       Other                        Other
                            Neurosurgeons          surgeons     surgeons OB/GYNs           physicians
                                              a             a            a            a              a
        Florida
        Mississippi                          5             3           11            5              50
        Nevada                               0             2            9           34              28
        Pennsylvania                        12            30           30           24              63
                                              a             a            a            a              a
        West Virginia

    Source: State provider organizations.

    Note: GAO summarized data from state provider organizations, generally for 2001 through 2003.
    a
    Provider organizations provided anecdotal reports that were not systematically collected or
    quantified.


    Although some reports have received extensive media coverage, in each of
    the five states we found that actual numbers of physician departures were
    sometimes inaccurate or involved relatively few physicians.

•   Reports of physician departures in Florida were anecdotal, not extensive,
    and in some cases we determined them to be inaccurate. For example,
    state medical society officials told us that Collier and Lee counties lost all
    of their neurosurgeons due to malpractice concerns; however, we found at
    least five neurosurgeons currently practicing in each county as of April
    2003. Provider groups also reported that malpractice pressures have
    recently made it difficult for Florida to recruit or retain physicians of any




    Page 17                                 GAO-03-836 Medical Malpractice and Access to Health Care
    type; however, over the past 2 years the number of new medical licenses
    issued has increased and physicians per capita has remained unchanged.26
•   In Mississippi, the reported physician departures attributed to recent
    malpractice pressures were scattered throughout the state and
    represented 1 percent of all physicians licensed in the state. Moreover, the
    number of physicians per capita has remained essentially unchanged since
    1997.27
•   In Nevada, 34 OB/GYNs reported leaving, closing practices, or retiring due
    to malpractice concerns; however, confirmatory surveys conducted by the
    Nevada State Board of Medical Examiners found nearly one-third of these
    reports were inaccurate—8 were still practicing and 3 stopped practicing
    due to reasons other than malpractice. Random calls we made to 30
    OB/GYN practices in Clark County found that 28 were accepting new
    patients with wait-times for an appointment of 3 weeks or less. Similarly,
    of the 11 surgeons reported to have moved or discontinued practicing, the
    board found 4 were still practicing.
•   In Pennsylvania, despite reports of physician departures, the number of
    physicians per capita in the state has increased slightly during the past 6
    years.28 The Pennsylvania Medical Society reported that between 2002 and
    2003, 24 OB/GYNs left the state due to malpractice concerns; however, the
    state’s population of women age 18 to 40 fell by 18,000 during the same
    time period. Departures of orthopedic surgeons comprise the largest single
    reported loss of specialists in Pennsylvania. Despite these reported
    departures, the rate of orthopedic surgeries among Medicare enrollees in
    Pennsylvania has increased steadily for the last 5 years, as it has
    nationally. (See fig. 1.)




    26
      The Florida Board of Medicine reported that 3,239 new licenses were issued in 2000, 3,577
    in 2001, and 3,858 in 2002. The number of physicians practicing in Florida per thousand in
    the population was 3.1 in both 2001 and 2002. Estimates of physicians per capita are based
    on counts of physicians practicing in the state reported by the Federation of State Medical
    Boards of the United States, Inc. (FSMB), and include osteopathic physicians.
    27
     Between 1997 and 2002 the number of physicians in Mississippi increased slightly, from
    1.9 to 2.0 per thousand in the population. Physician counts were reported by the Mississippi
    State Board of Medical Licensure and include osteopathic physicians and podiatrists.
    28
      Physicians practicing in Pennsylvania increased slightly between 1997 and 2001 from 2.6
    to 2.8 per thousand in the population and have remained essentially unchanged between
    2001 and 2002 at 2.8 per thousand in the population. Counts of physicians practicing in the
    state were reported by FSMB and include osteopathic physicians.




    Page 18                        GAO-03-836 Medical Malpractice and Access to Health Care
    Figure 1: Rates of Medicare-Covered Orthopedic Surgeries in Pennsylvania Have
    Increased

         100 Services per 1,000 part B fee-for-service beneficiaries


          90


          80


          70


          60


          50



           0




                                                                                                                                                      1
                              7
                7




                                                                             9
                                              8



                                                                8




                                                                                               9



                                                                                                              0



                                                                                                                           0



                                                                                                                                        1




                                                                                                                                                                      2
                                                                                                                                                     00
                             99
                99




                                                                             99
                                              99



                                                            99




                                                                                           99



                                                                                                            00



                                                                                                                       00



                                                                                                                                        00




                                                                                                                                                                      00
           e1



                             .1


                                         e1



                                                           .1


                                                                          e1



                                                                                          .1



                                                                                                         e2



                                                                                                                       .2



                                                                                                                                     e2



                                                                                                                                                     .2



                                                                                                                                                                 e2
                                                                                                                      ec
                                                           ec




                                                                                          ec
                         ec




                                                                                                                                                 ec
                                                                                                                                    n




                                                                                                                                                                  n
                                                                                                        n
                                          n




                                                                         n
            n




                                                                                                                                                               Ju
                                                                                                                                 Ju
                                                                                                     Ju
                                       Ju




                                                                      Ju
         Ju




                                                                                                                  –D
                                                      –D




                                                                                     –D
                       –D




                                                                                                                                               –D


                                                                                                                                                          n.–
                                                                                                                               n.–
                                                                                                   n.–
                                  n.–




                                                                    n.–
     n.–




                                                                                                                 ly
                                                      ly




                                                                                     ly
                        ly




                                                                                                                                                ly
                                                                                                              Ju
                                                   Ju




                                                                                  Ju
                     Ju




                                                                                                                                             Ju


                                                                                                                                                          Ja
                                                                                                                            Ja
                                                                                                Ja
                                  Ja




                                                                 Ja
    Ja




                              Pennsylvania
                              Four states without reported problems
                              United States
     Source: CMS.

    Notes: GAO analysis of Medicare part B claims data.

    Rates are based on Medicare part B allowed services per thousand Medicare part B fee-for-service
    beneficiaries and include all musculoskeletal surgeries provided by orthopedic surgeons.


•   In West Virginia, provider groups did not provide us with specific numbers
    of physician departures, but did offer anecdotal reports of physicians who
    have moved out of state or left practice. Despite these reports, the number
    of physicians per capita increased slightly between 1997 and 2002.29




    29
      From 1997 through 2002, the number of physicians practicing in West Virginia increased
    from 2.0 to 2.2 per thousand in the population. Counts of physicians practicing in the state
    were reported by FSMB and include osteopathic physicians.




    Page 19                                                         GAO-03-836 Medical Malpractice and Access to Health Care
Some Providers Report        Some providers in each of the five states also reported that physicians
Reducing Certain Services,   have recently cut back on certain services they believe to be high risk to
but Access to Care Not       reduce their malpractice insurance premiums or exposure to litigation.
                             Evidence was based on surveys conducted by state and national medical
Widely Affected              and specialty provider groups and anecdotal reports by state provider
                             groups, generally between 2001 and 2002. The most frequently cited
                             service reductions included spinal surgeries and joint revisions and repairs
                             (all five states), mammograms (Florida and Pennsylvania), and physician
                             services in a nursing home setting (Florida and Mississippi).

                             Survey data used to identify service cutbacks in response to physician
                             concerns about malpractice pressures are not likely representative of the
                             actions taken by all physicians. Most surveys had low response rates—
                             typically 20 percent or less.30 Moreover, surveys often did not identify any
                             one specific service as widely affected or identified service reductions in a
                             nonspecific manner. For example, in responding to one recent survey,
                             neurologists reported reducing 12 different types of services; however, the
                             most widely reported reduction for any one service type was reported by
                             fewer than 4 percent of respondents.31 AMA recently reported that about
                             24 percent of physicians in high-risk specialties responding to a national
                             survey have stopped providing certain services; however, the response
                             rate for this survey was low (10 percent overall), and AMA did not identify
                             the number of responses associated with any particular service.32

                             Our analysis of utilization rates among Medicare beneficiaries for three of
                             the specific services frequently cited as being reduced—spinal surgery,
                             joint revisions and repairs, and mammography—did not identify recent
                             reductions. For example, utilization of spinal surgeries among Medicare
                             beneficiaries in the five states generally increased from July 2000 through
                             June 2002, and is currently higher than the national average. (See fig. 2.)
                             Utilization of joint revision and repair services among Medicare
                             beneficiaries in the five states is slightly below, but has generally tracked


                             30
                              A survey of orthopedic surgeons in Mississippi yielded a response rate of 10 percent and
                             surveys of orthopedic surgeons in Florida and Pennsylvania and of neurologists nationally
                             all yielded response rates of about 20 percent.
                             31
                              Preliminary results as of January 23, 2003, of a joint AMA and American Academy of
                             Neurology survey.
                             32
                               AMA, National Physician Survey on Professional Medical Liability (Chicago, Ill.: April
                             2003). We attempted to obtain data from this survey specific to the nine states we
                             reviewed. However, AMA did not release the data out of concern that response rates for
                             these states were unacceptably low.




                             Page 20                       GAO-03-836 Medical Malpractice and Access to Health Care
the national average and has not recently declined.33 (See fig. 3.) Contrary
to reports of reductions in mammograms in Florida and Pennsylvania, our
analysis showed that utilization of these services among Medicare
beneficiaries is higher than the national average in both Florida, where
utilization rates have recently increased, and in Pennsylvania, where the
pattern of utilization has not recently changed. (See fig. 4.) We also
contacted selected hospitals and mammography facilities reported to have
had problems in these two states and found that the longer wait times
cited by provider organizations were more likely due to causes other than
malpractice pressures.34




33
 Joint revision and repairs reported by orthopedic surgeons as those reduced due to
malpractice concerns include certain hip, knee, and shoulder procedures.
34
  We contacted mammography facilities reported to have had problems in Pennsylvania
and Florida. Representatives from both Pennsylvania mammography facilities contacted
told us that increased demand for radiology services was the primary cause for longer wait
times. One facility in Florida indicated that long wait times were due to a shortage of
radiology technicians rather than radiologists. A representative of another Florida facility
told us that malpractice concerns were leading to wait times of 3 or more months and that
demand for these services was also increasing. We contacted six mammography facilities
near this Florida facility and found relatively short wait times. Wait times for screening
mammograms ranged from 0 to 20 days at four locations and 20 to 30 days at two locations,
while wait times for diagnostic mammograms among all six locations ranged from 30 to 40
days, but in all cases could be scheduled sooner if a physician deemed it necessary. We
recently reported on the nation’s overall capacity to provide mammography services. See
U.S. General Accounting Office, Mammography: Capacity Generally Exists to Deliver
Services, GAO-02-532 (Washington, D.C.: Apr. 19, 2002).




Page 21                        GAO-03-836 Medical Malpractice and Access to Health Care
Figure 2: Rates of Medicare-Covered Spinal Surgeries in Five States with Reported
Problems Have Recently Increased

     3.5 Services per 1,000 part B fee-for-service beneficiaries


     3.0


     2.5


     2.0


     1.5


  1.0


      .5


      0




                                                                                                                                                         1
                          7
            7




                                                                          9
                                          8



                                                            8




                                                                                                            0



                                                                                                                              0



                                                                                                                                           1




                                                                                                                                                                         2
                                                                                            9




                                                                                                                                                        00
                         99
            99




                                                                          99
                                          99



                                                        99




                                                                                                            00



                                                                                                                          00



                                                                                                                                           00




                                                                                                                                                                         00
                                                                                        99
       e1



                         .1



                                     e1



                                                       .1



                                                                     e1




                                                                                                       e2



                                                                                                                         .2



                                                                                                                                        e2



                                                                                                                                                        .2



                                                                                                                                                                    e2
                                                                                       .1




                                                                                                                         ec
                                                       ec
                     ec




                                                                                                                                                    ec
                                                                                       ec




                                                                                                                                                                     n
                                                                                                                                       n
                                                                                                        n
                                      n




                                                                      n
        n




                                                                                                                                                                  Ju
                                                                                                                                    Ju
                                                                                                     Ju
                                   Ju




                                                                   Ju
     Ju




                                                                                                                    –D
                                                  –D
                   –D




                                                                                                                                                  –D
                                                                                  –D




                                                                                                                                                             n.–
                                                                                                                                  n.–
                                                                                                n.–
                              n.–




                                                                n.–
 n.–




                                                                                                                    ly
                                                  ly
                    ly




                                                                                                                                                   ly
                                                                                  ly




                                                                                                                 Ju
                                               Ju
                 Ju




                                                                                                                                                Ju
                                                                               Ju




                                                                                                                                                             Ja
                                                                                                                               Ja
                                                                                                Ja
                              Ja




                                                                Ja
Ja




                          Five states with reported problems
                          Four states without reported problems
                          United States
Source: CMS.

Notes: GAO analysis of Medicare part B claims data.

Rates are based on Medicare part B allowed services per thousand Medicare part B fee-for-service
beneficiaries and include all musculoskeletal spine surgeries performed by orthopedic surgeons.




Page 22                                                         GAO-03-836 Medical Malpractice and Access to Health Care
Figure 3: Rates of Medicare-Covered Joint Revisions and Repairs in Five States
with Reported Problems Have Not Recently Declined

       8     Services per 1,000 part B fee-for-service beneficiaries


       7


       6


       5


       4


       3


       2


       1


       0




                                                                                                                                                           1
                          7
            7




                                                                             9
                                           8



                                                             8




                                                                                              9



                                                                                                              0



                                                                                                                            00



                                                                                                                                             1




                                                                                                                                                                           2
                                                                                                                                                          00
                         99
            99




                                                                             99
                                           99



                                                         99




                                                                                             99



                                                                                                              00




                                                                                                                                             00




                                                                                                                                                                           00
                                                                                                                           20
       e1



                         .1



                                      e1



                                                        .1



                                                                        e1



                                                                                             .1



                                                                                                         e2




                                                                                                                                        e2



                                                                                                                                                          .2



                                                                                                                                                                      e2
                                                                                                                           c.
                                                        ec




                                                                                       ec
                     ec




                                                                                                                                                      ec
                                                                                                                                         n




                                                                                                                                                                       n
                                                                                                          n
                                       n




                                                                         n
        n




                                                                                                                      De




                                                                                                                                                                    Ju
                                                                                                                                      Ju
                                                                                                       Ju
                                    Ju




                                                                      Ju
     Ju




                                                   –D




                                                                                         D
                   –D




                                                                                                                                                    –D
                                                                                                                     ly–
                                                                                     ly–




                                                                                                                                                               n.–
                                                                                                                                 n.–
                                                                                                  n.–
                                n.–




                                                                  n.–
 n.–




                                                   ly
                    ly




                                                                                                                                                     ly
                                                                                                                   Ju
                                                Ju




                                                                                  Ju
                 Ju




                                                                                                                                                  Ju



                                                                                                                                                               Ja
                                                                                                                                 Ja
                                                                                                  Ja
                               Ja




                                                                 Ja
Ja




                              Five states with reported problems
                              Four states without reported problems
                              United States
  Source: CMS.

Notes: GAO analysis of Medicare part B claims data.

Rates are based on Medicare part B allowed services per thousand Medicare part B fee-for-service
beneficiaries and include selected services (hip, knee, and shoulder repairs/revisions that were
identified as high risk) provided by orthopedic surgeons.




Page 23                                                          GAO-03-836 Medical Malpractice and Access to Health Care
Figure 4: Rates of Medicare-Covered Mammograms in Florida and Pennsylvania
Remain above the National Average

  240 Services per 1,000 female part B fee-for-service beneficiaries

  220

  200

  180

  160

  140

  120


  100



       0




                                                                                                                                                          1
                          7
            7




                                                                            9
                                          8



                                                            8




                                                                                              9



                                                                                                              0



                                                                                                                            00



                                                                                                                                            1




                                                                                                                                                                          2
                                                                                                                                                         00
                         99
            99




                                                                            99
                                          99



                                                        99




                                                                                          99



                                                                                                              00




                                                                                                                                            00




                                                                                                                                                                          00
                                                                                                                           20
       e1



                         .1



                                     e1



                                                       .1



                                                                       e1



                                                                                         .1



                                                                                                         e2




                                                                                                                                       e2



                                                                                                                                                         .2



                                                                                                                                                                     e2
                                                                                                                           c.
                                                       ec




                                                                                         ec
                     ec




                                                                                                                                                     ec



                                                                                                                                                                      n
                                                                                                                                        n
                                                                                                          n
                                      n




                                                                        n
        n




                                                                                                                      De




                                                                                                                                                                   Ju
                                                                                                                                     Ju
                                                                                                       Ju
                                   Ju




                                                                     Ju
     Ju




                                                  –D




                                                                                    –D
                   –D




                                                                                                                                                   –D
                                                                                                                     ly–




                                                                                                                                                              n.–
                                                                                                                                 n.–
                                                                                                  n.–
                              n.–




                                                                n.–
 n.–




                                                  ly




                                                                                    ly
                    ly




                                                                                                                                                    ly
                                                                                                                   Ju
                                               Ju




                                                                                 Ju
                 Ju




                                                                                                                                                 Ju


                                                                                                                                                              Ja
                                                                                                                                Ja
                                                                                                  Ja
                              Ja




                                                                Ja
Ja




                          Florida
                          Pennsylvania
                          Four states without reported problems
                          United States
 Source: CMS.

Notes: GAO analysis of Medicare part B claims data.

Rates are based on Medicare part B allowed services per thousand female Medicare part B fee-for-
service beneficiaries and include all mammograms performed by radiologists.


Although data limitations preclude an analysis of physician services in a
nursing home setting, interviews with industry representatives did not
reveal widespread reductions of services provided in these facilities.
Nursing home representatives in all five states reported that facilities are
facing increasing malpractice pressures due to higher premiums or
decreased availability of coverage and in two states reported that these
pressures are causing some physicians to stop providing services in these
facilities. However, they also told us that residents still receive needed
physician services.




Page 24                                                     GAO-03-836 Medical Malpractice and Access to Health Care
Health Care Providers   Some health care providers have taken certain actions to avoid access
Have Taken Actions to   problems in the face of malpractice-related pressures. Several hospital
Avoid Access Problems   officials we contacted reported they are assuming physicians’ liability
                        insurance costs to avoid any access problems related to malpractice
                        pressures. Officials in 9 of 49 hospitals contacted in the five states
                        reported that, in order to retain needed staff, they have either hired
                        physicians as direct employees, thereby covering their malpractice
                        insurance premiums in full, or provided them with partial premium
                        subsidies. An unpublished survey completed by The Hospital &
                        Healthsystem Association of Pennsylvania found that 5 of 89 hospitals or
                        health systems responding had taken these measures to maintain adequate
                        staffing. An official at a small hospital in a largely rural Mississippi county
                        told us that the hospital recently hired six family practitioners who
                        provide all of its obstetrics services in order to assume their liability
                        insurance costs and prevent loss of these services after the physicians’
                        premiums increased significantly. An official at a West Virginia hospital
                        reported that increasing numbers of newly recruited physicians are
                        coming to the area as direct employees of hospitals.

                        In addition, where allowed by state law, some providers are going without
                        malpractice insurance coverage. For example, a provider group in
                        Mississippi reported that increasing numbers of nursing homes are going
                        without coverage for some period of time because insurers are not
                        renewing their policies or are raising premiums to rates that are
                        unaffordable. According to an official from one insurer of Mississippi
                        nursing homes, more than 40 homes statewide were without coverage at
                        some point during 2002 as compared to fewer than 5 homes in 2001.
                        Similarly, while Florida law does not require that physicians carry
                        malpractice insurance, hospitals may impose such a requirement on
                        affiliated physicians.35 One hospital contacted in the state told us it has




                        35
                         Florida law imposes certain requirements on physicians who decide to go without
                        coverage. For example, physicians with hospital staff privileges who decide not to carry
                        commercial coverage must maintain assets or credit of at least $750,000 annually to cover
                        potential malpractice claims. Under certain circumstances, physicians may waive this
                        requirement but are required to inform all patients if they do.




                        Page 25                       GAO-03-836 Medical Malpractice and Access to Health Care
                           loosened this requirement in response to physicians’ concerns over
                           increasing malpractice premiums.36


                           Several recently published surveys report that physicians practice
Physicians Reportedly      defensive medicine in response to malpractice pressures.37 In addition,
Practice Defensive         most published studies designed to measure the prevalence of and costs
                           associated with such practices generally conclude that physicians practice
Medicine, but              defensive medicine in specified circumstances and that doing so raises
Prevalence and Costs       health care costs. However, because the surveys generally had low
                           response rates and were not precise in measuring the prevalence of these
of Such Practices Are      practices, and because the studies examined physician practice behavior
Not Reliably               in only narrowly specified clinical situations, the results cannot be used to
Measured                   reliably estimate the overall prevalence or costs of defensive medicine
                           practices.


Physicians Report          Physicians responding to surveys reported that they practice defensive
Practicing Defensive       medicine to varying extents, but low response rates and imprecise
Medicine, but Surveys      measurements of defensive medicine practices preclude generalizing these
                           responses to all physicians. For example, a 2003 AMA survey found that, of
Must Be Interpreted with   the 30 percent of responding physicians who reported recently referring
Caution                    more complex cases to specialists, almost all indicated that professional
                           liability pressures were important in their decision; and an April 2002
                           survey conducted by the American Academy of Orthopaedic Surgeons
                           found that, of the 48 percent of responding orthopedists who reported that
                           the costs of malpractice insurance caused them to alter their practice,
                           nearly two-thirds reported ordering more diagnostic tests.38 However, the


                           36
                             A March 2003 survey conducted by AHA reported that some hospitals are taking on more
                           risk in response to malpractice pressures. This includes not purchasing coverage, allowing
                           their physicians to practice without coverage, paying higher deductibles, reducing coverage
                           levels, and increasingly becoming self-insured. In addition to actions taken by health care
                           providers, some states have taken steps to make malpractice insurance more affordable or
                           easier to obtain.
                           37
                            Because of the potential for increased health care costs, we highlight the practice of
                           defensive medicine associated with the overutilization of certain diagnostic tests or
                           procedures to reduce exposure to malpractice liability. Such practices are sometimes
                           referred to as “positive defensive medicine.” Physicians may also reduce or eliminate
                           certain services they believe place them at risk of malpractice litigation. Such practices are
                           sometimes referred to as “negative defensive medicine.”
                           38
                            AMA, National Physician Survey on Professional Medical Liability (Chicago, Ill.: April
                           2003). American Academy of Orthopaedic Surgeons, Medical Malpractice Insurance
                           Concerns – Final Report (Rosemont, Ill.: April 2002).



                           Page 26                         GAO-03-836 Medical Malpractice and Access to Health Care
response rates for the AMA and AAOS surveys were about 10 and 15
percent, respectively, raising questions about how representative these
responses were of all physicians nationwide. Another 2002 survey of 300
physicians conducted by a polling firm found that, due to concerns about
medical malpractice liability, 79 percent of respondents reported ordering
more tests, 74 percent reported referring patients to specialists more
often, and 41 percent reported prescribing more medications than they
otherwise would based only on medical necessity.39 However, these survey
results do not indicate whether the respondents practice the cited
defensive behaviors on a daily basis or only rarely, or whether they
practice them with every patient or only with certain types of patients.

Officials from AMA and several medical, hospital, and nursing home
associations in the nine states we reviewed told us that defensive medicine
exists to some degree, but that it is difficult to measure; and officials cited
surveys and published research but could not provide additional data
demonstrating the extent and costs associated with defensive medicine.
Some officials pointed out that factors besides defensive medicine
concerns also explain differing utilization rates of diagnostic and other
procedures. For example, a Montana hospital association official said that
revenue-enhancing motives can encourage the utilization of certain types
of diagnostic tests, while officials from Minnesota and California medical
associations identified managed care as a factor that can mitigate
defensive practices. According to some research, managed care provides a
financial incentive not to offer treatments that are unlikely to have medical
benefit.40




39
 Harris Interactive, The Fear of Litigation Study – The Impact on Medicine, a special
report prepared at the request of Common Good (Rochester, N.Y.: April 2002),
http://ourcommongood.com/medicine/item?item_id=3396 (downloaded June 4, 2003).
40
 Daniel P. Kessler and Mark B. McClellan, “Medical Liability, Managed Care, and Defensive
Medicine,” working paper #7537, National Bureau of Economic Research (Cambridge,
Mass.: 2000).




Page 27                       GAO-03-836 Medical Malpractice and Access to Health Care
Some Research Identifies      Most research that has attempted to measure defensive practices has
Defensive Medicine in         examined physician practices under specific clinical situations.41 For
Certain Clinical Situations   example, based on clinical scenario surveys, records review, and a
                              synthesis of prior research, a 1994 study concluded that the percentage of
                              diagnostic procedures related to defensive medicine practices is higher in
                              specific clinical situations, such as the management of head injuries in ERs
                              and cesarean deliveries in childbirth, but lower when measured across
                              multiple procedures.42 The same study also surveyed physicians about nine
                              hypothetical clinical scenarios likely to encourage defensive medicine
                              practices and found the share of physicians reporting taking at least one
                              clinical action primarily out of concern about malpractice varied widely
                              depending on the situation—from 5 percent for back pain to 29 percent for
                              head trauma. A more recent 1999 study that used records review found
                              that reduced malpractice premiums for OB/GYNs were related to a
                              statistically significant but small decrease in the rate of cesarean sections
                              performed for some groups of mothers, a procedure researchers believe to
                              be influenced by physicians’ concerns about malpractice liability.43

                              Some studies have also concluded that certain tort reforms may reduce
                              defensive medicine as evidenced by slower growth in health care
                              expenditures; however, these studies have not fully considered the range
                              of factors that can influence medical spending.44 For example, a 1996 study
                              using records review found that for a population of elderly Medicare
                              patients treated for acute myocardial infarction or ischemic heart diseases,
                              certain tort reforms led to reductions of 5 to 9 percent in hospital



                              41
                               Researchers generally rely on two approaches to measure the extent of defensive
                              medicine practices. They (1) use surveys to present a clinical scenario, ask physicians to
                              choose a treatment and provide a rationale for their decision, and may also examine the
                              variation in survey responses across groups facing different amounts of malpractice
                              pressure, or (2) review clinical or other records to compare actual treatment approaches
                              and health care expenditures across groups of physicians facing different amounts of
                              malpractice pressure.
                              42
                               U.S. Congress, Office of Technology Assessment, Defensive Medicine and Medical
                              Malpractice, OTA-H-602 (Washington, D.C.: U.S. Government Printing Office, 1994).
                              43
                               Lisa Dubay, Robert Kaestner, and Timothy Waidmann, “The Impact of Malpractice Fears
                              on Cesarean Section Rates,” Journal of Health Economics, vol. 18, no. 4 (1999): 491-522.
                              44
                               Researchers have found that physician practice patterns and health care spending can
                              vary greatly across geographic regions for many reasons. See Jonathan Skinner and John E.
                              Wennberg, “How Much Is Enough? Efficiency and Medicare Spending in the Last Six
                              Months of Life,” working paper #6513, National Bureau of Economic Research (Cambridge,
                              Mass.: April 1998).




                              Page 28                        GAO-03-836 Medical Malpractice and Access to Health Care
                          expenditures.45 However, this study did not control for other factors that
                          can affect hospital costs, such as the extent of managed care penetration
                          in different areas. When controlling for managed care penetration in a
                          2000 follow-up study, the same researchers found that the reductions in
                          hospital expenditures attributable to direct tort reforms dropped to about
                          4 percent.46 Moreover, preliminary findings from a 2003 study that
                          replicated and expanded the scope of these studies to include Medicare
                          patients treated for a broader set of conditions failed to find any impact of
                          state tort laws on medical spending.47 Appendix III summarizes the
                          methods, findings, and limitations of published studies examining
                          defensive medicine.


Studies Cannot Be         Although available research suggests that defensive medicine may be
Generalized to Reliably   practiced in specific clinical situations, the findings are limited and cannot
Estimate Defensive        be generalized to estimate the prevalence and costs of defensive medicine
                          nationwide. Because the studies focused on specific clinical
Medicine Prevalence and   circumstances and populations, even slight changes in these scenarios
Costs                     could yield significant changes in the degree of defensive medicine
                          practices identified. Consequently, reports that use the results of these
                          studies to estimate defensive medicine practices and costs nationally are
                          not reliable. For example, recent reports by the U.S. Department of Health
                          and Human Services (HHS) applied the 5 to 9 percent hospital cost savings
                          estimate for Medicare heart patients to total national health care spending
                          to estimate the total defensive medicine savings that could result if federal




                          45
                           The researchers found that direct reforms (such as caps on damage awards, abolition of
                          punitive damages, and collateral-source rule reforms) were associated with reduced
                          medical expenditures, while indirect reforms (such as caps on contingency fees, mandatory
                          payment of damages through periodic installments, joint and several liability reform, and
                          patient compensation funds) were not. Daniel P. Kessler and Mark B. McClellan, “Do
                          Doctors Practice Defensive Medicine?” Quarterly Journal of Economics, vol. 111, no. 2
                          (1996): 353-90.
                          46
                           Kessler and McClellan, “Medical Liability, Managed Care, and Defensive Medicine.”
                          47
                           U.S. Congress, Congressional Budget Office (CBO), Cost Estimate: H.R. 5 – Help
                          Efficient, Accessible, Low-cost, Timely Healthcare (HEALTH) Act of 2003 (March 2003).
                          CBO characterizes results relating to its analysis of defensive medicine practices as
                          preliminary.




                          Page 29                       GAO-03-836 Medical Malpractice and Access to Health Care
                           tort reforms were enacted.48 Because the 5 to 9 percent savings only
                           applies to hospital costs for elderly patients treated for two types of heart
                           disease, the savings cannot be generalized across all services, populations,
                           and health conditions.


                           Premium rates reported for the physician specialties of general surgery,
States with Certain        internal medicine, and OB/GYN—the only specialties for which data were
Noneconomic                available—were relatively stable on average in most states from the mid-
                           to late 1990s and then began to rise, but more slowly among states with
Damage Caps Had            certain noneconomic damage caps.49 Malpractice claims payments against
Lower Recent Growth        all physicians between 1996 and 2002 also tended to be lower and grew
                           less rapidly on average in states with these damage caps than in states
in Malpractice             with limited reforms; however, these averages obscured wide variation
Premium Rates and          between states in any given year and for individual states from year to
Claims Payments            year. Like the premium rate data, these claims payment data do not depict
                           the experience of all providers; they exclude institutional providers such
                           as hospitals and nursing homes, for which comprehensive data were not
                           available. Moreover, differences in both premiums and claims payments
                           are also affected by multiple factors in addition to damage caps, and we
                           could not determine the extent to which differences among states were
                           attributable to the damage caps or to additional factors.


Premium Growth Was         The average medical malpractice premium rates across the three
Lower in States with       specialties reported by MLM (general surgery, internal medicine, and
Noneconomic Damage         OB/GYN) remained relatively stable during the mid- to late-1990s. From
                           1996 to 2000, average premium rates for all states changed little, as did
Caps Than in States with   average premium rates for states with certain caps on noneconomic
Limited Reforms            damages and states with limited reforms, increasing or decreasing




                           48
                             HHS, Office of the Assistant Secretary for Planning and Evaluation, Confronting the New
                           Health Care Crisis: Improving Health Care Quality and Lowering Costs By Fixing Our
                           Medical Liability System (Washington, D.C.: July 24, 2002),
                           http://aspe.hhs.gov/daltcp/reports/litrefm.htm (downloaded June 9, 2003); and Addressing
                           the New Health Care Crisis: Reforming the Medical Litigation System to Improve the
                           Quality of Health Care (Washington, D.C.: Mar. 3, 2003),
                           http://aspe.hhs.gov/daltcp/reports/medliab.htm (downloaded June 9, 2003).
                           49
                            Noneconomic damages compensate for harm that is not easily quantifiable, such as pain
                           and suffering.




                           Page 30                       GAO-03-836 Medical Malpractice and Access to Health Care
annually by no more than about 5 percentage points on average.50 After
2000, premium rates began to rise across most states on average, but more
slowly among the states with certain noneconomic damage caps. In
particular, from 2001 to 2002, the average rates of increase in the states
with noneconomic damage caps of $250,000 and $500,000 or less were 10
and 9 percent, respectively, compared to 29 percent in the states with
limited reforms. (See fig. 5.)




50
 We focused our analysis on those states with noneconomic damage caps as a key tort
reform because such caps are included in proposed federal tort reform legislation and
because published research generally finds these caps to have a greater impact on medical
malpractice premium rates and claims payments than some other tort reform measures.
See appendix II for details on our classification of states by tort reforms.




Page 31                       GAO-03-836 Medical Malpractice and Access to Health Care
Figure 5: Premium Rates for Three Physician Specialties Rose After 2000, but to a
Lesser Extent in States with Noneconomic Damage Caps

    35 Average percentage growth in premium rates


    30


    25


    20


    15


    10


      5


      0


     -5


    -10
          1996 –            1997 –               1998 –               1999 –   2000 –       2001 –
          1997              1998                 1999                 2000     2001         2002
                   States with caps of $250,000 (4 states)
                   States with caps of $500,000 or less (8 states)a
                   States with limited reforms (11 states)
                   All states
    Source: MLM.


Notes: GAO analysis of MLM base premium rates, excluding discounts, rebates, and surcharges,
reported for the specialties of general surgery, internal medicine, and OB/GYN.
Premiums are adjusted for inflation to 2002 dollars.
a
This category excludes states with caps of $250,000.


The recent increases in premium rates were also lower for each reported
physician specialty in the states with these noneconomic damage caps.
From 2001 to 2002, the average rates of premium growth for each specialty
in the states with these noneconomic damage caps were consistently
lower than the growth rates in the limited reform states. (See fig. 6.)




Page 32                                   GAO-03-836 Medical Malpractice and Access to Health Care
Figure 6: Recent Premium Growth Was Lower for Three Physician Specialties in
States with Noneconomic Damage Caps

    40 Average percentage growth in premium rates, 2001-2002


    35
                     33

    30                                                    29
                                                                            27
    25


    20

                             16                                16
    15                                                                           14
          12
                                           11    10
    10          9                                                   9
                                                                        8

     5


     0
           General surgery                 Internal medicine            OB/GYN

                States with caps of $250,000 (4 states)

                States with caps of $500,000 or less (8 states)a

                States with limited reforms (11 states)

                All states
Source: MLM.

Note: GAO analysis of MLM base premium rates, excluding discounts, rebates, and surcharges,
reported for the specialties of general surgery, internal medicine, and OB/GYN.

Premiums are adjusted for inflation to 2002 dollars.
a
This category excludes states with caps of $250,000.


In addition to including rates for only three specialties, premium rates
reported by MLM are subject to other limitations. First, because MLM
relies on a voluntary survey, its data do not include all insurers that
provide coverage in each state. Certain companies that may have a large
market share in a particular state may not be included. MLM estimates that
its 2002 survey may exclude about one-third of the total malpractice
insurance market nationwide. Second, insurers that do report rates have
not consistently done so across all the years, or have not consistently
reported premiums in different geographic areas within each state. We
generally excluded data from insurers that did not consistently report
premium rates across most of the years studied. Third, premium rates do
not reflect discounts, premium offsets, or rebates that may effectively
reduce the actual premium rate, or surcharges that are assessed in certain



Page 33                                 GAO-03-836 Medical Malpractice and Access to Health Care
                          states for physician participation in mandatory state-funded insurance
                          programs. These surcharges can range from a small amount to more than
                          the base premium rate.

                          Other studies have found a relationship between direct tort reforms that
                          include noneconomic damage caps and lower rates of growth in
                          premiums.51 For example, in a recent analysis of malpractice premiums in
                          states with and without certain medical malpractice tort limitations, the
                          Congressional Budget Office (CBO) estimated that certain caps on damage
                          awards in combination with other elements of proposed federal tort
                          reform legislation would effectively reduce malpractice premiums on
                          average by 25 to 30 percent over the 10-year period from 2004 through
                          2013.52 A 1997 study that assessed physician-reported malpractice
                          premiums from 1984 through 1993 found that direct reforms, including
                          caps on damage awards, lowered the growth in malpractice premiums
                          within 3 years of their enactment by approximately 8 percent.53


Average Claims Payments   Average per capita payments for claims against all physicians tended to be
and Growth Lower in       lower on average in states with noneconomic damage caps than in states
States with Noneconomic   with limited reforms.54 From 1996 through 2002, the average per capita
                          payments were $10 for states with these damage caps compared with $17
Damage Caps Than in       for states with limited reforms. Within these averages, however, were wide
States with Limited       variations among states. For example, in 2002 the per capita claims
Reforms                   payments among states with these caps ranged from $4 to $16, compared
                          with $3 to $33 among states with limited reforms. In addition, two states
                          among those with limited reforms had consistently higher average claims
                          payments, raising the overall average among this group of states.55 Absent


                          51
                            Direct reforms are limits on amounts that can be recovered in a malpractice action
                          including: caps on noneconomic or total damages, abolition of punitive damages, collateral
                          source rule reforms, and abolition of mandatory prejudgment interest.
                          52
                           CBO, Cost Estimate: H.R. 5.
                          53
                            Daniel P. Kessler and Mark B. McClellan, “The Effects of Malpractice Pressure and
                          Liability Reforms on Physicians’ Perceptions of Medical Care,” Law and Contemporary
                          Problems, vol. 670, no. 1 (1997): 81-106.
                          54
                           Per capita claims payments are the total claims payments in each state divided by the
                          state population.




                          Page 34                        GAO-03-836 Medical Malpractice and Access to Health Care
the claims experience of these two states, the average claims payment for
states with limited reforms from 1996 through 2002 would decrease to $11,
only slightly higher than the $10 in states with these damage caps.

Average growth in per capita claims payments for all physicians was also
lower among the states with caps on noneconomic damages than among
the states with limited reforms. From 1996 through 2002 average per capita
claims payments grew by 5 and 6 percent in the states with noneconomic
damage caps of $250,000 and $500,000 or less, respectively, compared to
10 percent in the states with limited reforms. However, the growth in these
payments also varied widely among states in any given year and within
individual states from year to year. For example, from 2001 to 2002, the
average growth in claims payments on an individual state basis ranged
from a 68 percent decrease in the District of Columbia to a 70 percent
increase in Wyoming. Within the same state, growth rates fluctuated
widely from year to year. For example, Mississippi experienced an 18
percent decrease in claims payments from 1999 to 2000, followed by a 61
percent increase in 2001, and a 5 percent decrease in 2002.

The claims payment data reported to NPDB that we analyzed contain
certain limitations. The data include malpractice claims against licensed
physicians, and not against other institutional providers such as hospitals
and nursing homes, thus limiting the overall completeness of the data
across all providers. In addition, as we have previously reported, certain
claims payments may be underreported to NPDB. When physicians are not
specifically named in a malpractice settlement, the related claims
payments may not be reported.56 Nevertheless, because insurers must
report payment of claims against physicians subject to federal law and not
varying state laws, NPDB data are useful in comparing trends across
states. Other sources of claims payment data are subject to limitations of




55
  Average per capita claims payments among states with limited tort reforms were highest
in the District of Columbia and Pennsylvania in each year from 1996 through 2002. For
example, in 2002, average claims payments were $27 and $33 for the District of Columbia
and Pennsylvania, respectively, compared to from $3 to $18 in the remaining states with
limited tort reforms.
56
 See U.S. General Accounting Office, National Practitioner Data Bank: Major
Improvements Needed to Enhance Data Bank’s Reliability, GAO-01-130 (Washington,
D.C.: Nov. 17, 2000).




Page 35                       GAO-03-836 Medical Malpractice and Access to Health Care
completeness or comparability.57 See appendix II for more information on
the limitations of NPDB and other claims data sources.

For states that have adopted certain tort reforms, especially caps on
noneconomic damages, other studies have also found associations with
lower claims payments. In its recent analysis of malpractice premiums and
claims payments in states with various medical malpractice tort
limitations, CBO found that caps on damage awards result in lower
malpractice costs.58 Another study based on claims data in 19 states
showed that direct reforms were associated with a smaller percentage of
claims resolved with some compensation to plaintiffs and reduced claim
frequency.59 In contrast, other researchers who have examined the effect
of indirect tort reforms on malpractice costs have found mixed results.60
One study found that indirect reforms did not reduce malpractice cost
indicators, while another found that a greater number of reforms (both
direct and indirect) were associated with lower malpractice costs.61 These
studies have also relied on claims data that have limitations in terms of
their completeness and comparability.




57
  For example, the National Association of Insurance Commissioners (NAIC) maintains
data on claims costs reported by malpractice insurers; however, NAIC officials told us that
reporting requirements are dictated by state law. As a result, certain types of insurers are
exempted from reporting in certain states (such as insurers operating in a single state,
certain physician mutual companies, or—in all states—self-insured groups), thus limiting
the usefulness of the data for making state-level comparisons.
58
 CBO, Cost Estimate: H.R. 5.
59
 See Daniel P. Kessler and Mark B. McClellan, “How Liability Law Affects Medical
Productivity,” Journal of Health Economics, vol. 21, no. 6 (2002): 931-55.
60
  Indirect reforms are changes in laws that do not directly specify limits on amounts that
can be recovered in a malpractice action; rather, they may indirectly affect recoverable
amounts, such as by limiting attorneys’ contingency fees or allowing periodic rather than
lump sum payments of awards.
61
 Kessler and McClellan, “The Effects of Malpractice Pressure and Liability Reforms on
Physicians’ Perceptions of Medical Care” and Stephen Foreman, Pennsylvania Medical
Society Health Services Research Institute, [Premium] Deceit: A Critique of a Center For
Justice and Democracy Study by J. Robert Hunter and Joanne Doroshow (Harrisburg,
Pa.: Jan. 8, 2003).




Page 36                        GAO-03-836 Medical Malpractice and Access to Health Care
Factors Other Than Caps    Differences in malpractice premiums and claims payments across states
on Noneconomic Damages     are influenced by several factors other than noneconomic damage caps.
Also Affect Premiums and   First, the manner in which damage caps are administered can influence
                           the ability of the cap to restrain claims and thus premium costs. Some
Claims Payments Trends     states permit injured parties to collect damages only up to the specified
                           level of the cap regardless of the number of defendants, while other states
                           permit injured parties to collect the full cap amount from each defendant
                           named in a suit. Malpractice insurers told us that imposing a separate cap
                           on amounts recovered from each of several defendants increases total
                           claims payouts, which can hinder the effectiveness of the cap in
                           constraining premium growth. Second, tort reforms unrelated to caps can
                           also affect premium and claims costs. For example, California tort reform
                           measures not only include a $250,000 cap but also allow other collateral
                           sources to be considered when determining how much an insurer must
                           pay in damages and allow periodic payment of damages rather than
                           requiring payment in a lump sum, among other measures. Malpractice
                           insurers told us that these provisions in addition to the cap have helped to
                           constrain premium growth in that state. In Minnesota, which has no caps
                           on damages but has relatively low growth in premium rates and claims
                           payments, trial attorneys maintain that prescreening requirements reduce
                           claim costs and premiums by preventing some meritless claims from going
                           to trial. Third, state laws and regulations unrelated to tort reform, such as
                           premium rate regulations, vary widely and can influence premium rates.
                           Some states such as Minnesota and Mississippi tend not to regulate rates,
                           while others, such as California, require state approval of the premium
                           rates charged by insurers.62 Finally, insurers’ premium pricing decisions
                           are affected by their losses on medical malpractice claims and income
                           from investments, and other market conditions such as the level of market
                           competition among insurers and their respective market shares.63 We could
                           not determine the extent to which differences in premium rates and claims
                           payments across states were attributed only to damage caps or also to
                           these additional factors.




                           62
                            In 1988, California passed Proposition 103, which in part required greater state oversight
                           and approval of premium rate increases.
                           63
                            For more information on the factors that influence malpractice premium rates, see
                           GAO-03-702.




                           Page 37                        GAO-03-836 Medical Malpractice and Access to Health Care
                           We received comments on a draft of this report from three independent
External Comments          health policy researchers and from AMA. Each of the researchers has
and Our Evaluation         expertise in malpractice-related issues and has conducted and published
                           research on the effects of malpractice pressures on the health care system,
                           and two of the three are physicians. The independent researchers
                           generally concurred with our findings and provided technical comments,
                           which we incorporated as appropriate.

                           In its written comments, AMA questioned our finding that rising
                           malpractice premiums have not contributed to widespread health care
                           access problems, expressing concern that the scope of our work limited
                           our ability to fully identify the extent to which malpractice-related
                           pressures are affecting consumers’ access to health care. We disagree with
                           AMA, as explained below. However, in response to AMA and the other
                           reviewers’ comments, we clarified the report’s discussion of the scope of
                           work and methods used to assess health care access issues. AMA’s
                           comments fell into four general areas: completeness of evidence
                           examined, measures used to assess access problems, time lags in available
                           data, and the cost and impact of defensive medicine.


Completeness of Evidence   AMA questioned our finding that access problems were not widespread
Examined                   based on our work in 5 states, whereas it has identified 18 states “in a full-
                           blown liability crisis.” It further cited results from its own recent physician
                           survey on professional liability as evidence that medical liability concerns
                           are causing physicians to limit their practices. The report clearly states the
                           scope of our work and does not attempt to generalize our findings beyond
                           the 5 states with reported problems that we reviewed. However, these 5
                           states were among the most visible and often-cited examples of “crisis”
                           states by AMA and other provider groups. We believe that our finding that
                           malpractice-related concerns contributed to localized but not widespread
                           access problems in these states provides relevant and important insight
                           into the overall problem. With respect to AMA’s reference to evidence
                           available from its own survey, our report notes that the low response rate
                           of 10 percent to its survey precludes the ability to reliably generalize the
                           survey results to all physicians.

                           AMA suggested that we withhold release of the report until we contacted
                           state and national medical and specialty associations to obtain more
                           complete and accurate information about access to care problems and it
                           provided contacts for associations in each of the five states with reported
                           problems and for four national specialty associations. We made these
                           contacts throughout the course of our work, and the information these


                           Page 38                   GAO-03-836 Medical Malpractice and Access to Health Care
                          associations provided formed the basis for many of our findings. As the
                          draft report noted, we contacted state medical, hospital, and nursing home
                          association representatives in each of the five states with reported
                          problems. We also contacted nine national medical and specialty
                          associations, including three of the four AMA cited, which were specified
                          in the draft report. In response to AMA’s comments, we added an
                          appendix to specify the names of each national and state provider
                          association we contacted during the course of our work.

                          AMA commented that we failed to account for the two clinical areas of
                          patient care in which impairment of access has been the most egregious:
                          obstetrical and ER services. It attributed its concern to our
                          acknowledgment in the report that we were unable to use Medicare claims
                          data to investigate reported concerns about these services. Because of the
                          recognized limitations of Medicare claims data for these and other
                          services, we used other methods to explore whether malpractice-related
                          pressures had affected access to ER on-call surgical services and newborn
                          deliveries and indeed found—and reported—evidence of access problems
                          for these services in localized areas. In response to AMA and technical
                          comments from the other reviewers, we clarified the report’s discussion of
                          our methodology for this issue.


Measures Used to Assess   AMA commented that using aggregated data on physician supply to draw
Access Problems           conclusions about access to care is problematic. It said that physicians
                          tend to hold multiple state licenses and typically retain their licenses when
                          they relocate their practices, thus potentially obscuring the supply of
                          practicing physicians, and overall counts of physicians can obscure the
                          impact of changes for different specialties and different jurisdictions. We
                          agree that measuring changes in physician supply—especially changes due
                          to malpractice-related issues—and the related effects on access to care is
                          problematic. Sharing AMA’s concerns, during the course of our work we
                          obtained available data reported by state medical licensing agencies for
                          newly licensed physicians and for physicians practicing in the state
                          whenever possible rather than for all licensed physicians and contrasted
                          those data with reports of departing physicians. As noted in the draft
                          report, although we reported physician supply and practice changes at the
                          state level, the number of recent departures attributed specifically to
                          malpractice concerns was relatively small and usually not concentrated in
                          particular locales. Also as noted in the draft report, we further explored
                          reports of specialty-specific problems, such as orthopedic surgeons in
                          Pennsylvania and OB/GYNs in Nevada. For example, we analyzed rates of
                          all procedures performed by orthopedic surgeons in Pennsylvania and


                          Page 39                   GAO-03-836 Medical Malpractice and Access to Health Care
                         found them to be growing, and called a random sample of OB/GYN
                         practices in Clark County, Nevada, and on that basis determined that
                         obstetrical care was readily available. Moreover, our Medicare claims
                         analysis of certain high-risk services was specialty-specific. For example,
                         to assess assertions by orthopedic surgeons that they have reduced the
                         provision of spinal surgeries and joint revisions and repairs, our analysis
                         was limited to only those services performed by orthopedic surgeons.


Time Lags in Available   AMA commented that our analysis of Medicare claims data as of June 2002
Data                     does not capture the current experience of physician decisions to curtail
                         certain services or to retire or relocate their practices, the impact of which
                         takes time to develop. We agree it is challenging to identify data that are
                         sufficiently current and reliable to describe the effects of reported
                         problems. However, we reported that premium increases began about
                         2000, and others have found that premiums began increasing as early as
                         the late 1990s. We therefore believe that analyzing Medicare claims data
                         through June 2002 provides important insights into at least 2 years of this
                         most recent period of rising premiums. Moreover, we augmented our
                         Medicare claims analysis with more recent qualitative data, such as
                         interviews in late 2002 and early 2003, with national and state provider
                         associations and local providers in areas where access problems were
                         reported to exist.


The Cost and Impact of   AMA commented that while specific estimates of defensive medicine costs
Defensive Medicine       have not been conclusive, the vast majority of peer-reviewed research
                         indicates that those costs are enormous, in the tens of billions of dollars
                         per year. To support this point, AMA cited three recent government
                         studies. As our report notes, the peer-reviewed literature attempts to
                         quantify the extent and sometimes the cost of defensive medicine under
                         narrowly defined clinical circumstances that cannot be generalized more
                         broadly. Two of the three government studies that AMA cited are
                         examples of what we believe to be overgeneralizations of prior study
                         results. We cite one of these by way of example in our report. The third
                         government study AMA cited does not address the cost of defensive
                         medicine but instead explicitly notes the difficulty of estimating such costs
                         and the speculative nature of existing estimates.

                         AMA also commented that our draft report ignored the impact of defensive
                         medicine costs in terms of patient access, expressing the view that these
                         costs are ultimately reflected in rising health insurance premiums that
                         contribute substantially to the number of uninsured. Our draft report


                         Page 40                   GAO-03-836 Medical Malpractice and Access to Health Care
noted that, because of the absence of data to reliably measure overall
malpractice-related costs—such as the combined cost of malpractice
insurance premiums, litigation, and defensive medicine practices—we did
not assess the indirect impact on access to care that may result from any
added costs that malpractice pressures impose on the health care system.
In response to AMA’s comment, we moved our discussion of this point to
the report’s Results in Brief.


As agreed with your offices, unless you publicly announce this report’s
contents earlier, we plan no further distribution until 30 days after its issue
date. At that time, we will send copies to other interested congressional
committees and Members of Congress. We will also make copies available
to others on request. In addition, this report is available at no charge at the
GAO Web site at http://www.gao.gov.

Please call me at (202) 512-7118 or Randy DiRosa at (312) 220-7671 if you
have any questions. Other major contributors are listed in appendix IV.




Kathryn G. Allen
Director, Health Care—Medicaid
 and Private Health Insurance Issues




Page 41                   GAO-03-836 Medical Malpractice and Access to Health Care
              Appendix I: National and State Provider
Appendix I: National and State Provider
              Associations Contacted



Associations Contacted

              During the course of our work, we contacted a number of national and
              state health care provider associations in order to identify the actions
              health care providers have taken in response to malpractice pressures and
              the localized effects of any reported actions on consumers’ access to
              health care.

              National Provider Associations

              American Academy of Neurology
              American Association of Neurological Surgeons
              American Association of Orthopaedic Surgeons
              American College of Emergency Physicians
              American College of Obstetricians and Gynecologists
              American College of Radiology
              American Health Care Association
              American Hospital Association
              American Medical Association

              State Provider Associations

              Table 2: State Provider Associations GAO Contacted

               State                  Provider association
               California             California Association of Health Facilities
                                      California Healthcare Association
                                      California Medical Association
               Coloradoa              Colorado Health and Hospital Association
               Florida                Florida Health Care Association
                                      Florida Hospital Association
                                      Florida Medical Association
               Minnesota              Minnesota Health and Housing Alliance
                                      Minnesota Hospital Association
                                      Minnesota Medical Association
               Mississippi            Mississippi Health Care Association
                                      Mississippi Hospital Association
                                      Mississippi State Medical Association
               Montana                Association of Montana Health Care Providers
                                      Montana Medical Association




              Page 42                       GAO-03-836 Medical Malpractice and Access to Health Care
Appendix I: National and State Provider
Associations Contacted




    State                 Provider association
    Nevada                Nevada Health Care Association
                          Nevada Hospital Association
                          Nevada State Medical Association
    Pennsylvania          The Hospital & Healthsystem Association of Pennsylvania
                          Pennsylvania Health Care Association
                          Pennsylvania Medical Society
    West Virginia         West Virginia Health Care Association
                          West Virginia Hospital Association
                          West Virginia State Medical Association

a
We also contacted officials from the Colorado Medical Society and the Colorado Health Care
Association, but they did not respond to our request for an interview.




Page 43                          GAO-03-836 Medical Malpractice and Access to Health Care
                       Appendix II: Scope and Methodology
Appendix II: Scope and Methodology


                       In response to concerns about rising malpractice premiums, we examined
                       how health care provider responses to rising premiums have affected
                       access to health care, what is known about how rising premiums and fear
                       of litigation cause health care providers to practice defensive medicine,
                       and how rates of growth in malpractice premiums and claims payments
                       compare across states with varying levels of tort reform laws.


Consumers’ Access to   To evaluate how actions taken by physicians in response to malpractice
Health Care            premium increases have affected consumers’ access to health care, we
                       focused our review at the state level because reliable national data
                       concerning physician responses to malpractice pressures were not
                       available. We selected nine states that encompass a range of premium
                       pricing and tort reform environments. Five of the states—Florida,
                       Mississippi, Nevada, Pennsylvania, and West Virginia—are among those
                       cited as “crisis” or “problem” states by the American Medical Association
                       (AMA) and other health care provider organizations based on such factors
                       as higher than average increases in malpractice insurance premium rates,
                       reported difficulties obtaining malpractice coverage, and reported actions
                       taken by providers in response to their concerns about rising premiums
                       and malpractice litigation. Four of the states—California, Colorado,
                       Minnesota, and Montana—are not cited by provider groups as
                       experiencing malpractice-related problems. (See table 3.)




                       Page 44                      GAO-03-836 Medical Malpractice and Access to Health Care
                                                                  Appendix II: Scope and Methodology




Table 3: Tort Reforms and Average Rates of Premium Increases in Nine States

                                                                                      Tort reforms in place as of 1995a
                                                                                                                                                     Average annual
                                                                                                                                                       premium rate
                                                                                         Noneconomic                                                       increase,
 Extent of                                                   Noneconomic                 damage cap of                                                    2001–2002
 malpractice                                                 damage cap of                $500,000 or         Other tort       Limited tort             (percentage
                                                                                                 b                                      c
 problems                           State                      $250,000                      less              reforms           reforms                    change)
                                              e                                                                        f
 States with reported               Florida                                                                        X                                                  23
          d
 problems                           Mississippi                                                                                        X                              45
                                    Nevada                                                                                             X                              28
                                    Pennsylvania                                                                                       X                              35
                                    West Virginia                                                                  Xg                                                 12
 States without                     California                           X                                         X                                                   6
 reported problems
                                    Colorado                             X                                         X                                                   8
                                                                                                                       f
                                    Minnesota                                                                      X                                                   5
                                    Montana                              X                                         X                                                  10

Sources: National Conference of State Legislatures (NCSL) and Medical Liability Monitor (MLM).

                                                                  Notes: GAO analysis of state tort reforms obtained from the NCSL “State Medical Liability Laws
                                                                  Table” (Oct. 16, 2002) and independently confirmed in selected instances.

                                                                  Premium increases are based on base rates reported by MLM for specialties of general surgery,
                                                                  internal medicine, and obstetrics/gynecology (OB/GYN). Premiums are in 2002 dollars.
                                                                  a
                                                                   States are categorized based on tort reforms enacted as of 1995 because research indicates any
                                                                  impact reforms may have on premium rates or claims payments would follow the implementation of
                                                                  tort reforms by at least 1 year. Mississippi, Nevada, and West Virginia have recently enacted varying
                                                                  tort reforms.
                                                                  b
                                                                      This category excludes states with caps of $250,000.
                                                                  c
                                                                      States had no damage caps or collateral source reform.
                                                                  d
                                                                  Problem status based on the American Medical Association (AMA) classification of “crisis” state as of
                                                                  April 2003.
                                                                  e
                                                                   Florida enacted a noneconomic damage cap of $250,000 in 1988, but the cap was limited to cases
                                                                  involving arbitration; noneconomic damage limits may increase if the plaintiff or defendant refuses to
                                                                  arbitrate.
                                                                  f
                                                                  Florida and Minnesota enacted mandatory collateral source offsets that directly reduced expected
                                                                  malpractice awards.
                                                                  g
                                                                      West Virginia enacted a $1 million cap on noneconomic damages.




                                                                  Page 45                              GAO-03-836 Medical Malpractice and Access to Health Care
Appendix II: Scope and Methodology




In each of the nine states we reviewed, we contacted or interviewed
officials from associations representing physicians, hospitals, and nursing
homes to more specifically identify the actions physicians have taken in
response to malpractice pressures and the localized effects of any
reported actions on access to services. (See app. I for a complete list of the
provider organizations we contacted at the state and national levels.) Such
actions were reported only in the five states with reported problems. In
these five states we obtained and reviewed the evidence upon which the
reports were based. Evidence of physician departures, retirements,
practice closures, and reduced availability of certain hospital-based
services consisted of survey results, information compiled and quantified
by provider groups, and unquantified anecdotal reports collected by
provider groups. Although we did not attempt to confirm each report cited
by state provider groups, we judgmentally targeted follow-up contacts
with local providers where the reports suggested potentially acute
consumer access problems or where multiple reports were concentrated
in a geographic area. With the local providers we contacted directly,
including representatives of physician practices, clinics, and hospitals, we
discussed the reports provided by the state provider groups and explored
the resulting implications for consumers’ access to health care. In total, we
contacted 49 hospitals and 61 clinics and physician practices in the five
states. From these contacts we identified examples of access problems
that were related to providers’ concerns about malpractice-related
pressures as well as examples of provider actions that did not appear to
affect consumer access or were not substantiated.

We separately examined evidence of specific high-risk services that
providers reportedly reduced in response to concerns about malpractice
pressures. Such evidence consisted of results from surveys conducted by
national and state-level medical, hospital, and specialty associations that
identified the high-risk procedures physicians reported reducing or
eliminating in response to malpractice pressures. High-risk services
commonly identified in these surveys included spinal surgeries, joint
revisions and repairs, mammograms, physician services in nursing homes,
emergency room services, and obstetrics. We analyzed Medicare
utilization data to assess whether reported reductions in three of these
high-risk services—spinal surgery, joint revisions and repairs, and
mammograms—-have had a measurable effect on consumers’ access to




Page 46                      GAO-03-836 Medical Malpractice and Access to Health Care
                     Appendix II: Scope and Methodology




                     these services.1 To calculate service utilization rates per thousand fee-for-
                     service Medicare beneficiaries enrolled in part B, we used Medicare part B
                     physician claims data from January 1997 through June 2002 and the
                     Medicare denominator files from 1997 through 2001.2 For 2002, we
                     estimated each state’s part B fee-for-service beneficiary count by adjusting
                     the 2001 count by the change in the 65 and older population between 2001
                     and 2002 and the change in Medicare beneficiaries enrolled in part B
                     managed care plans between January 1 and July 1, 2002.3


Defensive Medicine   To assess what is known about how rising premiums and fear of litigation
Practices            cause health care providers to practice defensive medicine, we reviewed
                     studies that examined the prevalence and costs of defensive medicine and
                     the potential impact of tort reform laws on mitigating these costs that
                     were published in 1994 or later, generally in peer-reviewed journals, or
                     were conducted by government research organizations. We identified
                     these studies by searching databases including MEDLINE, Econlit,
                     Expanded Academic ASAP, and ProQuest; and through contacts with
                     experts and affected parties. Several studies published prior to 1994 were
                     reviewed by the Office of Technology Assessment (OTA) in its
                     comprehensive 1994 report on defensive medicine, which we included in
                     our review. In addition, we also explored the issue with medical provider
                     organizations and examined the results of several recent surveys,
                     including those conducted by national health care provider organizations,


                     1
                      Limitations to Medicare data precluded an assessment of trends for physician services
                     provided in nursing homes, emergency room services, and obstetrics services. Utilization
                     rates of services provided in nursing homes per Medicare beneficiary could not be
                     completed because Medicare data do not identify the beneficiaries that reside in these
                     facilities. Emergency room services could not be analyzed because it is not possible to
                     accurately count emergency room services in the part B physician claims data. Obstetrics
                     services could not be analyzed because Medicare beneficiaries are mostly elderly, so the
                     counts of females of childbearing age are not representative of the general population.
                     2
                      Medicare part B claims for these specific services were identified by the five-digit
                     procedure codes specified in the Centers for Medicare & Medicaid Services’ (CMS) Health
                     Care Common Procedure Coding System (HCPCS).
                     3
                      Population data were obtained from the U.S. Bureau of the Census. Medicare enrollment
                     data were obtained from the Medicare Denominator File. The Medicare Denominator File
                     contains data on all Medicare beneficiaries entitled to benefits in a given year and includes
                     information on the programs in which they participate. The changes in Medicare
                     enrollment in managed care programs were reported in CMS’s MMCC Monthly Summary
                     Report on Medicare Managed Care Plans. See HHS, CMS, Medicare Managed Care
                     Contract (MMCC) Plans - Monthly Summary Report (Baltimore, Md.: Jan. 1, 2002 and July
                     1, 2002), http://www.cms.hhs.gov/healthplans/statistics/mmcc/ (downloaded Apr. 16, 2003).




                     Page 47                        GAO-03-836 Medical Malpractice and Access to Health Care
                           Appendix II: Scope and Methodology




                           in which providers were asked about their own defensive medicine
                           practices.


Malpractice Premium Rate   To assess the growth in medical malpractice premium rates and claims
and Claims Payments        payments across states, we compared trends in states with tort reforms
Growth                     that include noneconomic damage caps (4 states with a $250,000 cap and 8
                           states with a $500,000 or less cap4) to the 11 states (including the District
                           of Columbia) with limited reforms and the average for all states. We
                           focused our analysis on those states with noneconomic damage caps as a
                           key tort reform because such caps are included in proposed federal tort
                           reform legislation and because published research generally reports that
                           such caps have a greater impact on medical malpractice premium rates
                           and claims payments than some other types of tort reform measures. We
                           did not separately assess trends in the 28 states with various other tort
                           reforms because of the wide range of often dissimilar and incomparable
                           tort reforms that are included among these states. Because research
                           suggests that any impact of tort reforms on premiums or claims can be
                           expected to follow the implementation of the reforms by at least 1 year,
                           we grouped states into their respective categories based on reforms that
                           had been enacted no later than 1995 and reviewed premium rate and
                           claims payment data for the period 1996 through 2002. We relied upon a
                           summary of state tort reforms compiled by the National Conference of
                           State Legislatures (NCSL) to place states within the reform categories and
                           reviewed the information with respect to the 9 study states for accuracy in
                           February 2003. (See table 4.)




                           4
                            The eight states with a $500,000 or less cap do not include the four states with a $250,000
                           cap.




                           Page 48                        GAO-03-836 Medical Malpractice and Access to Health Care
                                         Appendix II: Scope and Methodology




Table 4: State Tort Reform Categories, Based on Reforms in Place as of 1995

 Noneconomic damage cap                Noneconomic damage cap
                                                          a
 of $250,000                           of $500,000 or less                            Other reformsa, b                Limited reformsc
 (4 states)                            (8 states)                                     (28 states)                      (11 states)
                                              d
 California                            Hawaii                                         Alabama                          Arkansas
                d                                      e
 Colorado                              Louisiana                                      Alaska                           District of Columbia
                                                           d
 Montana                               Massachusetts                                  Arizona                          Kentucky
                                                   d
 Utah                                  Michigan                                       Connecticut                      Mississippi
                                       Missourif                                      Delaware                         Nevada
                                                                                                  g
                                       North Dakota                                   Florida                          Ohio
                                       South Dakota                                   Georgia                          Oklahoma
                                       Wisconsin                                      Idaho                            Pennsylvania
                                                                                      Illinois                         South Carolina
                                                                                      Indiana                          Vermont
                                                                                      Iowa                             Wyoming
                                                                                                      h
                                                                                      Kansas
                                                                                              i
                                                                                      Maine
                                                                                                          j
                                                                                      Maryland
                                                                                      Minnesota
                                                                                      Nebraska
                                                                                      New Hampshirei
                                                                                      New Jersey
                                                                                      New Mexicoj
                                                                                      New York
                                                                                      North Carolina
                                                                                      Oregon
                                                                                      Rhode Island
                                                                                      Tennessee
                                                                                      Texas
                                                                                      Virginia
                                                                                      Washington
                                                                                      West Virginia

Source: NCSL.

                                         Notes: GAO analysis of summary data compiled by NCSL (Oct. 16, 2002). We independently
                                         reviewed selected sections for accuracy.
                                         a
                                          In states with patient compensation funds (PCF), the fund cap, rather than the per provider cap, is
                                         considered under these criteria. PCFs are either voluntary or mandatory state-sponsored funds that
                                         provide insurance coverage for health care providers beyond that guaranteed by the provider’s
                                         medical liability insurance policy.




                                         Page 49                           GAO-03-836 Medical Malpractice and Access to Health Care
Appendix II: Scope and Methodology




b
 States had a noneconomic or total damage cap above $500,000, any punitive damage cap, or
collateral source reform.
c
    States had no damage caps or collateral source reform.
d
    Caps may be increased or removed under special circumstances.
e
 Louisiana’s PCF cap is subject to a total cap of $500,000 for all claims of malpractice. Amounts
awarded for future medical expenses are paid from the state fund and not by individual providers, and
those amounts are not subject to the $500,000 limit.
f
    Missouri’s cap is indexed to inflation and was $500,000 in 1997, increasing to $547,000 by 2002.
g
 Florida enacted a noneconomic damage cap of $250,000 in 1988, but the cap was limited to cases
involving arbitration; noneconomic damage limits may increase if the plaintiff or defendant refuses to
arbitrate.
h
 Kansas enacted a noneconomic damage cap of $250,000 in 1988, but these damages are
recoverable by each party from all defendants.
i
    A noneconomic damage cap is limited to wrongful death cases.
j
    Damage cap increased beyond $500,000 during 1995.


To assess the growth in medical malpractice premiums, we analyzed state-
level malpractice premium rates for the specialties of general surgery,
internal medicine, and obstetrics/gynecology (OB/GYN) reported by
insurers to the Medical Liability Monitor (MLM) from 1996 to 2002.5 Our
analysis does not capture the experience of other physician specialties and
other types of medical providers such as hospitals and nursing homes.
MLM reports base premium rates that do not reflect discounts or rebates
that may effectively reduce the actual premium rates charged. We
generally excluded data from insurers that did not consistently report
premium rates across most of the years studied. We also excluded
surcharges for contributions to state patient compensation funds (PCF)
because these were inconsistently reported across states and years.6 We
adjusted rates for inflation using the urban consumer price index. We
calculated a composite average premium across all three specialties, as
well as specialty-specific average premiums, for each year. We then
analyzed growth rates in these average premiums from 1996 through 2002
across all states.



5
 MLM is a private research organization that annually surveys professional liability
insurance carriers in 50 states to obtain their base premium rates for the specialties of
internal medicine, general surgery, and OB/GYN.
6
 Where physicians participate in PCFs, they typically pay an annual surcharge for
participation in the fund, an assessment for payments made out of the fund, or both. These
surcharges can range from a small percentage of the base premium to nearly as much, and
in some instances, more than the base premium.




Page 50                              GAO-03-836 Medical Malpractice and Access to Health Care
Appendix II: Scope and Methodology




To assess the growth in medical malpractice claims payments, we
analyzed state level claims payment data from the National Practitioner
Data Bank (NPDB) from 1996 to 2002, which had been adjusted to 2002
dollars.7 We calculated average per capita claims payments and their
growth rates for each state across this time frame. Assuming a 1-year lag
to allow the reforms to affect these indicators, we calculated overall
averages of these indicators from 1996 to 2002, and used these averages to
compare average per capita payments and their rates of growth across the
reform categories.

The NPDB claims data we analyzed contain notable limitations. First, they
include malpractice claims against licensed physicians only, and not
against institutional providers such as hospitals and nursing homes.8
Secondly, as we have previously reported, NPDB claims may be
underreported. When physicians are not specifically named in a
malpractice judgment or settlement, the related claims are not reported to
the data bank, and certain self-insured and managed care plans may be
underreported as well.9 The extent to which this underreporting occurs is
not known. Finally, NPDB data do not capture legal and other
administrative costs associated with malpractice claims.

We examined other sources of information on claims payments, and found
none to be a comprehensive data source for each state that captures
malpractice claims costs from all segments of the malpractice insurance
market—commercial insurers, physician-mutual companies, and self-



7
 NPDB, established by the Health Care Quality Improvement Act of 1986, is maintained by
the Secretary of Health and Human Services and is a nationwide source of information on
physicians and other licensed health care practitioners who have been party to a medical
malpractice settlement or judgment. Insurers are required by law to report payments made
on behalf of these providers in settlement or satisfaction of a judgment in a malpractice
action, and are subject to civil penalties for noncompliance. Pub. L. No. 99-660, tit. IV, 100
Stat. 3743, 3784 (codified at 42 U.S.C. §§ 11101-11152 (2000))
8
 NPDB reports payments for claims against all licensed practitioners, including, physicians,
nurses, and dentists; however, we analyzed payments only for claims against physicians.
The consulting firm of Tillinghast-Towers Perrin estimates that total malpractice claims
costs (including payments and defense and administrative costs) in 2001 were
approximately $21 billion. See Tillinghast-Towers Perrin, U.S. Tort Costs: 2002 Update –
Trends and Findings on the Costs of the U.S. Tort System,
http://www.tillinghast.com/tillinghast/ (downloaded June 9, 2003). Payments reported for
physician claims in the NPDB database for the same year (excluding associated
defense/administrative costs) represent about 20 percent of these total costs.
9
See GAO-01-130.




Page 51                         GAO-03-836 Medical Malpractice and Access to Health Care
Appendix II: Scope and Methodology




insured and other groups. For example, data reported to the National
Association of Insurance Commissioners (NAIC) have been used in other
research; however, data are not reported consistently across states and
exclude payments from certain insurers. According to NAIC officials, the
laws that dictate reporting requirements differ by state, and not all
insurers are required to report in every state. They also stated that
exempted insurers can include those operating in a single state and certain
physician mutual companies.10 In all states, self-insured groups, which
represent a substantial proportion of the medical malpractice insurance
market, are exempted from reporting.11 Similarly, the Insurance Services
Office (ISO) is a private organization providing state-level price advisory
information to state insurance regulators. However, ISO does not operate
in all states, nor does it uniformly collect data on hospital claims, or claims
from physician mutual companies, and represents only 25 to 30 percent of
the total medical malpractice market. Physician Insurers Association of
America is an association of physician mutual companies; however, it does
not share proprietary state-level claims data. Jury Verdict Research is a
private research organization that collects data from several different
sources, including attorneys and media reports, among others. Some have
criticized the accuracy of this data set for several reasons, including a
varied and unsystematic data collection process and because large verdict
awards may be more likely to be included than smaller verdict awards.




10
  We found that exempted companies are disproportionately represented in states with
limited reforms.
11
 NAIC claims data represented slightly over a third of the total malpractice claim costs
reported by Tillinghast-Towers Perrin. See Tillinghast-Towers Perrin
http://www.tillinghast.com/tillinghast/.




Page 52                        GAO-03-836 Medical Malpractice and Access to Health Care
                                            Appendix III: Summary of Selected Research
Appendix III: Summary of Selected Research  Designed to Measure Defensive Medicine
                                            Prevalence and Costs


Designed to Measure Defensive Medicine
Prevalence and Costs
                                            Table 5 summarizes the scope, methods, results, and limitations of studies
                                            that examined the prevalence and costs of defensive medicine practices or
                                            the potential impact of tort reform laws on mitigating defensive medicine
                                            costs. Studies were published in 1994 or later, generally in peer-reviewed
                                            journals, or were conducted by government research organizations.

Table 5: Summary of Selected Research Designed to Measure Defensive Medicine Prevalence and Costs

Study              Scope                    Method                 Results                                Limitations
OTA, 1994a         Physicians from three    Physician clinical     Among other findings, defensive        Physician clinical scenario
                   national specialty       scenario surveys,      medicine causes less than 8            surveys were designed to elicit
                   societies (1993 data),   records reviews, and   percent of diagnostic procedures       defensive medicine practices
                   physicians from New      synthesis of prior     and varies significantly by clinical   among physicians; hence, they
                   Jersey (1993 data),      research.              situation.                             may overestimate the rate at
                   and cesarean                                                                           which defensive medicine is
                   deliveries in New                                                                      actually practiced.
                   York State (1984
                   data) and
                   Washington State
                   (1989 data).
Sloan and others, Births in Florida in      Survey of mothers      An increased threat of                 Results cannot be generalized,
    b          c
1995 and 1997     1987.                     and records reviews.   malpractice litigation is not          as study only assessed practice
                                                                   associated with improved birth         patterns in one state in 1 year.
                                                                   outcomes, and malpractice
                                                                   pressures generally had no
                                                                   impact on delivery method
                                                                   (cesarean vs. vaginal).
Kessler and        Medicare               Records reviews.         Direct tort reforms enacted by         Results cannot be generalized
               d
McClellan, 1996    beneficiaries treated                           states between 1985 and 1990           to all patients and procedures,
                   for a new heart attack                          reduced hospital expenditures for      and certain other factors that
                   or new ischemic                                 Medicare patients with a new           can influence practice patterns
                   heart disease (1984,                            heart attack or new ischemic           and health care expenditures
                   1987, and 1990                                  heart disease by 5 to 9 percent,       (such as the prevalence of
                   data).                                          respectively; indirect reforms had     managed care in an area) were
                                                                   no effect. Among states adopting       not controlled for.
                                                                   direct reforms prior to 1985, no
                                                                   consistent effect was found.
Dubay, Kaestner,   Births in the United     Records reviews.       A $10,000 reduction in                 Results are limited to only
and Waidmann,      States from 1990 to                             malpractice premiums could             certain socioeconomic groups
    e
1999               1992.                                           result in a 1.4 to 2.4 percent         of mothers.
                                                                   decline in the cesarean section
                                                                   rate for some mothers.
                                                                   Researchers concluded a total
                                                                   cap on damages would reduce
                                                                   the number of cesarean sections
                                                                   by 3 percent and total obstetrical
                                                                   charges by 0.27 percent.




                                            Page 53                        GAO-03-836 Medical Malpractice and Access to Health Care
                                                    Appendix III: Summary of Selected Research
                                                    Designed to Measure Defensive Medicine
                                                    Prevalence and Costs




 Study                     Scope                    Method                   Results                                  Limitations
 Kessler and               Medicare               Records reviews.           When controlling for the influence       Results cannot be generalized
 McClellan, 2000f          beneficiaries treated                             of managed care, direct tort             to all patients and procedures,
                           for a new heart attack                            reforms reduced hospital                 and certain other factors that
                           or new ischemic                                   expenditures for Medicare                can influence practice patterns
                           heart disease (1984-                              patients with a new heart attack         and health care expenditures
                           94 data).                                         or new ischemic heart disease by         (such as the supply of cardiac
                           Study attempted to                                about 4 percent.                         specialists in an area) were not
                           control for the                                                                            controlled for.
                           influence of managed
                           care.
 Kessler and               Medicare               Records reviews.           Direct tort reforms reduced              Findings cannot be generalized
                g
 McClellan, 2002           beneficiaries treated                             malpractice pressure and                 to all patients and procedure,
                           for a new heart attack                            hospital expenditures for                and certain other factors that
                           or new ischemic                                   Medicare patients with a new             can influence practice patterns
                           heart disease (1984-                              heart attack or new ischemic             and health care expenditures
                           94 data).                                         heart disease; indirect reforms          (such as the prevalence of
                           Study attempted to                                increased malpractice pressure           managed care in an area) were
                           identify the                                      in some cases.                           not controlled for.
                           mechanisms through
                           which reforms affect
                           the behavior of health
                           care providers.
                 h
 CBO, 2003                 Medicare                 Records reviews and No effect of tort controls on                 Results cannot be generalized
                           beneficiaries            expenditure analysis. medical expenditures or per                 to all patients and procedures.
                           diagnosed with a                               capita health spending.
                           broader set of
                           ailments than
                           considered in
                           previous research
                           (1989-99 data).

Sources: As noted below.

                                                    Note: Researchers generally rely on two approaches to measure the extent of defensive medicine
                                                    practices. They (1) use surveys to present a clinical scenario, ask physicians to choose a treatment
                                                    and provide a rationale for their decision, and may also examine the variation in survey responses
                                                    across groups facing different amounts of malpractice pressure, or (2) review clinical or other records
                                                    to compare actual treatment approaches and health care expenditures across groups of physicians
                                                    facing different amounts of malpractice pressure.
                                                    a
                                                    U.S. Congress, OTA, Defensive Medicine and Medical Malpractice, OTA-H-602 (Washington, D.C.:
                                                    U.S. Government Printing Office, 1994).
                                                    b
                                                     Frank A. Sloan and others, “Effects of the Threat of Medical Malpractice Litigation and Other Factors
                                                    on Birth Outcomes,” Medical Care, vol. 33, no. 7 (1995): 700-14.
                                                    c
                                                    Frank A. Sloan and others, “Tort Liability and Obstetricians’ Care Levels,” International Review of
                                                    Law and Economics, vol. 17, no. 2 (1997): 245-60.
                                                    d
                                                     Daniel P. Kessler and Mark B. McClellan, “Do Doctors Practice Defensive Medicine?” Quarterly
                                                    Journal of Economics, vol. 111, no. 2 (1996): 353-90.
                                                    e
                                                    Lisa Dubay, Robert Kaestner, and Timothy Waidmann, “The Impact of Malpractice Fears on
                                                    Cesarean Section Rates,” Journal of Health Economics, vol. 18, no. 4 (1999): 491-522.




                                                    Page 54                            GAO-03-836 Medical Malpractice and Access to Health Care
Appendix III: Summary of Selected Research
Designed to Measure Defensive Medicine
Prevalence and Costs




f
Daniel P. Kessler and Mark B. McClellan, “Medical Liability, Managed Care, and Defensive
Medicine,” working paper #7537, National Bureau of Economic Research (Cambridge, Mass.: 2000).
g
Daniel P. Kessler and Mark B. McClellan, “How Liability Law Affects Medical Productivity,” Journal of
Health Economics, vol. 21, no. 6 (2002): 931-55.
h
U.S. Congress, CBO, Cost Estimate: H.R. 5 – Help Efficient, Accessible, Low-cost, Timely
Healthcare (HEALTH) Act of 2003 (March 2003).




Page 55                           GAO-03-836 Medical Malpractice and Access to Health Care
                  Appendix IV: GAO Contacts and Staff
Appendix IV: GAO Contacts and Staff
                  Acknowledgments



Acknowledgments

                  Randy DiRosa, (312) 220-7671
GAO Contact
                  In addition to the person named above, key contributors to this report
Acknowledgments   were Gerardine Brennan, Iola D’Souza, Corey Houchins-Witt, and Margaret
                  Smith.




                  Page 56                      GAO-03-836 Medical Malpractice and Access to Health Care
             Related GAO Products
Related GAO Products


             Medical Malpractice Insurance: Multiple Factors Have Contributed to
             Increased Premium Rates. GAO-03-702. Washington, D.C.: June 27, 2003.

             National Practitioner Data Bank: Major Improvements Are Needed to
             Enhance Data Bank’s Reliability. GAO-01-130. Washington, D.C.:
             November 17, 2000.

             Medical Malpractice: Effects of Varying Laws in the District of
             Columbia, Maryland, and Virginia. GAO/HEHS-00-5. Washington, D.C.:
             October 15, 1999.

             Medical Liability: Impact on Hospital and Physician Costs Extends
             Beyond Insurance. GAO/AIMD-95-169. Washington, D.C.: September 29,
             1995.

             Medical Malpractice: Medicare/Medicaid Beneficiaries Account for a
             Relatively Small Percentage of Malpractice Losses. GAO/HRD-93-126.
             Washington, D.C.: August 11, 1993.

             Medical Malpractice: Experience with Efforts to Address Problems.
             GAO/T-HRD-93-24. Washington, D.C.: May 20, 1993.

             Medical Malpractice: A Continuing Problem with Far-Reaching
             Implications. GAO/T-HRD-90-24. Washington, D.C.: April 26, 1990.




(290230)
             Page 57                 GAO-03-836 Medical Malpractice and Access to Health Care
                         The General Accounting Office, the audit, evaluation and investigative arm of
GAO’s Mission            Congress, exists to support Congress in meeting its constitutional responsibilities
                         and to help improve the performance and accountability of the federal
                         government for the American people. GAO examines the use of public funds;
                         evaluates federal programs and policies; and provides analyses,
                         recommendations, and other assistance to help Congress make informed
                         oversight, policy, and funding decisions. GAO’s commitment to good government
                         is reflected in its core values of accountability, integrity, and reliability.


                         The fastest and easiest way to obtain copies of GAO documents at no cost is
Obtaining Copies of      through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full-
GAO Reports and          text files of current reports and testimony and an expanding archive of older
                         products. The Web site features a search engine to help you locate documents
Testimony                using key words and phrases. You can print these documents in their entirety,
                         including charts and other graphics.
                         Each day, GAO issues a list of newly released reports, testimony, and
                         correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site
                         daily. The list contains links to the full-text document files. To have GAO e-mail
                         this list to you every afternoon, go to www.gao.gov and select “Subscribe to e-mail
                         alerts” under the “Order GAO Products” heading.


Order by Mail or Phone   The first copy of each printed report is free. Additional copies are $2 each. A
                         check or money order should be made out to the Superintendent of Documents.
                         GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a
                         single address are discounted 25 percent. Orders should be sent to:
                         U.S. General Accounting Office
                         441 G Street NW, Room LM
                         Washington, D.C. 20548
                         To order by Phone:     Voice:    (202) 512-6000
                                                TDD:      (202) 512-2537
                                                Fax:      (202) 512-6061


                         Contact:
To Report Fraud,
                         Web site: www.gao.gov/fraudnet/fraudnet.htm
Waste, and Abuse in      E-mail: fraudnet@gao.gov
Federal Programs         Automated answering system: (800) 424-5454 or (202) 512-7470


                         Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800
Public Affairs           U.S. General Accounting Office, 441 G Street NW, Room 7149
                         Washington, D.C. 20548