Robinson, Pamela MBA; Xu, Xiao PhD; Keeton, Kristie MD, MPH; Fenner, Dee MD; Johnson, Timothy R. B. MD; Ransom, Scott DO, MBA
Medical liability reform is an urgent concern of physician organizations such as the American Medical Association (AMA) and the American College of Obstetricians and Gynecologists (ACOG). Expenses incurred by health care providers, including malpractice insurance, risk management expenses, court costs, attorney fees, and claims paid to victims, are contributing to the rising costs of health care.1 The findings of Localio et al2 and Ransom et al,3 however, suggest that expenditures arising from the medical malpractice litigation are rarely attributable to medical negligence. New factors, such as the spread of managed care and various tools of cost controls, and the increasing interest in patient safety are the real drivers of the current medical liability crisis.4 Managed care restricts the ability of hospitals and physicians to increase rates to share the burden of increased liability expenses. In addition, the current emphasis on patient safety draws public attention to the evaluation of medical care quality and could possibly instigate more litigation. Together, these factors put providers in a very difficult situation, which may impact the financial viability of many obstetrical practices.
The AMA believes that the United States legal system is forcing physicians away from the practice of medicine and therefore limiting patients’ access to health care. In an effort to encourage legal reform in the most affected areas of the country, the AMA has declared that 19 states are in a condition of full medical liability “Crisis,” 25 states are showing signs of developing a liability “Crisis,” while only 6 states are relatively safe from liability “Crisis” due to long-standing liability reform laws that include caps on damages.5 Similarly, ACOG has asserted that the professional liability system in the United States is severely jeopardizing patient care. The College has declared that women's health care is at risk in 22 “Red Alert” states6 and specifically indicates that liability insurance for obstetrician–gynecologists (ob-gyns) has become prohibitively expensive and is forcing ob-gyns to restrict or abandon their obstetrical practices.
Malpractice premiums are doubling and tripling in “Red Alert” states, and some ob-gyns are unable to obtain any malpractice insurance.6 Mello et al7 suggest that, although the medical malpractice crisis is not a new phenomenon, this crisis appears to be a professional issue of availability as well as affordability of medical malpractice insurance. Yet there has been little evidence of its potential impact on the exodus of physicians. This study seeks to examine current trends in this effect on a national basis by examining the relationship between state malpractice insurance premiums and obstetrician-gynecologist physician supply.
MATERIALS AND METHODS
To identify states with different levels of medical legal risk, we consulted the ACOG and AMA Web sites.5,6 The College divided the 50 states and the District of Columbia into “Red Alert” states and Safe states (Box: “States Categorized by ACOG “Red Alert” or Safe Liability Risk Status”). American Medical Association categorized all the states into “Crisis” states, Warning states, and Safe states (Box: “States Categorized by AMA Liability Risk Status”). For the purpose of this study, we grouped the AMA Warning states together with the AMA Safe states to make the data consistent with the ACOG classification style. Hereinafter, we call them the AMA “Remaining” states.
We obtained annual, state-specific, medical malpractice insurance rates from the Medical Liability Monitor.8 The rates were derived from a nationwide survey of medical liability insurance companies. Quotes were presented for either an entire state, a region within a state, a particular county, or a city metropolitan area. We developed a state average premium based on an average of the different premiums quoted per state each year. Separate fees or premiums paid into a state Patient Compensation Fund were excluded from our analysis because these charges were not consistently paid by physicians. We ranked the 50 states and the District of Columbia by 2002 average premiums and separated the 10 states with the highest average premiums and the 10 states with the lowest average premiums from the other states (hereinafter called the High-Premium states and the Low-Premium states, respectively). A list of these states and their average premiums in 2002 were presented in Table 1.
We used the 1995–2003 ACOG Membership Record for the number of practicing ob-gyns by state (ACOG fellowship statistics reports. Washington, DC: The American College of Obstetricians and Gynecologists; 1995-2003). To account for the size of different states and their demand for ob-gyns, we calculated the ratio of births to ob-gyn for each state. Data on birth rates were obtained from the National Center for Health Statistics’ National Vital Statistics Reports.9 Note that this represents the total number of births in each state and year. These births could be delivered by any providers who practiced obstetrical care. Therefore, the births per ob-gyn rate is a measure for the overall condition of obstetrical care in that state. Year 2002 was the most recent year for which the birth rates were available. It should be a sufficient estimate for the birth rates in 2003 to serve the research purpose here. The key variable examined in this study is the percentage change in the ratios between 1995 and 2003. We adopted the ACOG membership classification of Fellows and Junior Fellows and derived the percentage change for each category to distinguish the influence of medical legal risk on physicians of different experience. In this study, Junior Fellows include only those ob-gyns who were actually in practice.
We chose to compare data between 1995 and 2003 for 2 reasons. First, starting in December 1995, ACOG Membership Record categories were consistent and comparable to the following years, and 1995 was the earliest year in which separate data on practicing Junior Fellows became available. Second, we hypothesized that the influence of medical liability risk on ob-gyn supply does not show up instantaneously; rather its impact is on future supply. To evaluate the overall impact of medical legal risk on ob-gyns’ choice of practice locations, we sought to use the earliest year possible as the base period for comparison.
We examined the above data for differences in ob-gyn supply between states of different medical liability risk levels: ACOG “Red Alert” versus ACOG Safe; AMA “Crisis” versus AMA Remaining; and High Premium versus Low Premium. All analyses were conducted using SAS 9.1 statistical software (SAS Institute, Cary, NC). We calculated the mean, median, standard deviation, and interquartile range for the distribution of percentage changes. Shapiro-Wilke tests were conducted to evaluate whether the data were normally distributed. The percentage changes between high-risk states and low-risk states were compared using Student t tests if the data were normally distributed and Mann-Whitney tests if the data were not normally distributed. P values less than .05 were considered to be statistically significant. This study was approved by the University of Michigan Institutional Review Board.
To allow for regional variations within states, we conducted a similar analysis for the change in births per ob-gyn in all metropolitan regions with 2002 premiums over $100,000 (Medical Liability Monitor). Specifically, they include Miami (Dade County), Florida, at $210, 576; Cleveland (Cuyahoga County), Ohio, at $152, 496; Las Vegas (Clark County), Nevada, at $141,760; Detroit (Wayne County), Michigan, at $140,917; Nassau County, New York, at $115,431; and Chicago (Cook County), Illinois, at $110,091. Because the ACOG membership directory categorizes ob-gyn locations by chapters, which may have memberships that span several different cities and counties, we used a different data source of ob-gyn supply by county derived from the Area Resource File.10 From this data file, we extracted information on the total number of ob-gyns for patient care in each county and compared the percentage changes in births per ob-gyn between 1995 and 2001 (the latest year in which data were available) in these 6 high-premium counties with all other counties in the nation.
Tables 2 and 3 report the average number of births/ob-gyn ratio in states of different categories of medical liability risk and their percentage changes between 1995 and 2003, respectively. In most states, the number of births per Fellow decreased from 1995 to 2003, whereas the number of births per Junior Fellow increased during the same period. This indicates an improvement in the supply of Fellows and a worsening in the supply of Junior Fellows. Such patterns, however, are fairly equal across all states without regard to any particular state liability risk status. The percentage changes were similar between the high-risk and low-risk states by ACOG designation. Therefore, there is no significant difference in the supply of practicing ob-gyns (both Fellows and Junior Fellows) in ACOG “Red Alert” states versus the Safe states.
The tables also display the results for our analysis on the AMA risk categories. Once again, there is no statistically significant difference in the supply of practicing ob-gyns between the “Crisis” states and the Remaining states. The percentage changes in births per Fellow and births per Junior Fellow for AMA “Crisis” and Remaining states were similar. These results suggest that, like ACOG designation, the AMA designation of “Crisis” states, compared with the Remaining states, does not predict ob-gyn practice location.
Likewise, our analysis for premium-based medical liability risk categories shows that the percentage changes in births per Fellow were comparable between the 10 High-Premium states and the 10 Low-Premium states. There was no significant difference in their average percentage changes (–15.1% versus –16.4%, P = .67). However, we did observe a significant difference in the percentage changes of births per Junior Fellow between the High- and Low-Premium states. The median percentage increase in the High-Premium states was more than 5 times larger than the change in Low-Premium states (28.5% versus 5.0%, P = .03).
Our results from the county-level analysis also show a trend in the supply of ob-gyns (results not shown in the table). In general, most counties in the United States experienced decreases in the number of births per ob-gyn, indicating an improvement in the supply of obstetric care. However, on average, the 6 highest-premium counties had a lower rate of decrease than other counties in the nation (–3.4% versus –5.6%, P = .92). Even though the result lacked statistical significance, this may hint at a worse supply situation in meeting the population need for obstetric services in these high-premium counties relative to other areas of the country.
Finally, we examined the malpractice premium levels in ACOG “Red Alert” states and Safe states, as well as AMA “Crisis” states and Remaining states. On average, from 1995 to 2002, the premiums in ACOG “Red Alert” states increased by 32.9%, while the premiums in the Safe states increased by 30.9% (P = .19). Similarly, there is no significant difference in the change of premiums between AMA “Crisis” states and Remaining states (36.4% versus 28.8%, P = .17) over the same period of time.
In the United States, obstetric service providers have been struggling with climbing medical malpractice insurance premiums for many years and lack of affordable malpractice insurance coverage has affected access to care in some areas.11 Our analysis shows that, from 1995 to 2003, while the number of births per Fellow decreased in 48 of the 50 states and the District of Columbia, the number of births per Junior Fellow increased in 36 of the 50 states and the District of Columbia. In particular, we found that the increase in births per Junior Fellow was significantly greater for the 10 highest-premium states than the 10 lowest-premium states. Further, our data suggests that, on average, birth rates actually increased in the Low-Premium states while they decreased in the High-Premium states during the study period. Taken together, this supports our hypothesis that high malpractice premiums may have had an influence on the decision of Junior Fellows to practice in these states. This finding is consistent with our expectation that this group of ob-gyns is more sensitive to the impact of liability risk. Analysis of the top 6 high-premium counties also shows a tendency of negative impact of high malpractice premiums on ob-gyn supply. Level of malpractice premiums appears to be an importance factor influencing the supply of ob-gyns, especially for Junior Fellows.
Both ACOG and AMA designations of high-risk and low-risk states were based on a number of factors. The College designated its “Red Alert” states according to the availability of professional liability coverage (as measured by the number of insurance carriers issuing policies or abandoning the medical liability market, and the absolute cost and rate of increase in insurance premiums), a combination of geographic, economic, and other conditions that affect ob-gyn supply, such as the history of state tort reform (James Lumalcuri, ACOG Department of Professional Liability/Risk Management, July 2004, personal communication). The AMA's criteria included surveys and polls of medical students and residents’ and physicians’ perceptions of medical legal risk, as well as their concerns about medical liability coverage costs and access.12
As shown in our statistical analysis, there was no significant difference in the premium levels between ACOG “Red Alert” states and the Safe states, and between AMA “Crisis” states and the Remaining states. Both designations lacked a statistically significant relationship to malpractice premiums for ob-gyns. Indeed the ACOG “Red Alert” category includes several Low-Premium states such as Virginia (average premium $36,282), Georgia (average premium $37,616), Alabama (average premium $37,690), and Washington (average premium $38,597). Massachusetts (average premium $84,566) and Michigan (average premium $88,945), where average liability premiums are high, are not listed among the “Red Alert” states. Similarly, the AMA “Crisis” category includes several states where the medical liability premiums are low, eg, Arkansas (average premium $28,484), or modest, eg, Georgia (average premium $37,616), Washington (average premium $38,597), and Oregon (average premium $40,160). The few states that are declared as safe from liability crisis do not necessarily coincide with the states with low premiums either.
This deviation from the premium costs concerning ob-gyns might be an important reason for our failure to detect any significant difference in the supply of ob-gyns when using the ACOG and AMA designations. Our significant findings on the impact of medical liability premiums on the supply of Junior Fellows warrants more attention from both organizations toward premium management in their future role of guiding obstetrical care.
The sharp increase in the number of births per Junior Fellow in the High-Premium states suggests that the supply of new ob-gyns is not keeping pace with the population need for obstetric services in states where the medical malpractice insurance premiums are the highest. This raises concern about women's future access to obstetric care. Recent data show that medical students’ interest in the obstetrics-gynecology specialty has also declined.13–15 The number of U.S. medical doctors electing an ob-gyn residency training program has significantly declined. In 2004, the percentage of graduating U.S. senior medical students filling the Postgraduate Year One positions in obstetrics-gynecology decreased to 65%,16 compared with 86% approximately 10 years ago.17 When choosing a residency, the medical liability issue influenced the choice of state for 39% of medical students, and it affected the choice of specialty for 50% of them.18
It is likely that rising premiums are causing the pipeline for new ob-gyns to dry up. This may reduce patient access to and use of prenatal care in the future. A widely held belief is that higher medical liability risk may affect health care use through positive defensive medicine and overuse. Our results here suggest that increased malpractice premiums can influence obstetrical care in another way, ie, by reducing the entry of new ob-gyns to this specialty. Unless effective actions are taken, we may expect a shortage of ob-gyns in the coming years. This may greatly limit the capacity of our health care system to care for high-risk babies and deliver obstetric services in a timely manner. It can also restrict the ability of providers to spend quality time with obstetric patients. In the long run, high medical liability premiums may come at a cost of reduced patient access to high-quality obstetrical care.
Several factors limit the accuracy and depth of this study. First, the average liability costs data were derived from the Medical Liability Monitor survey of medical malpractice premiums. The number of participating insurance firms and their service areas were not consistent for each state across study years. Our estimate of state average premiums might be biased by the survey participants. Moreover, because the survey data did not have detailed information on the market share of each insurance firm in the state, we were not able to derive a weighted average premium, which should be a more appropriate measure of state malpractice liability risk.
Second, our study used the number of ob-gyns in practice who performed at least some obstetric services, but we did not account for the fact that these physicians might alter their practices to reduce high-risk and other obstetric services. The new Medical Liability Survey of ACOG members reveals that 9.2% of the members who were surveyed stopped practicing obstetrics because of liability insurance costs and availability, and 12.2% decreased the number of deliveries.19 Therefore, our results likely underestimated the real impact of medical liability risk.
Third, the present study focused on state-level effects. Nevertheless, the market for malpractice insurance varies within each state. Different areas of the states may have different insurance firms, and the premiums written may vary as well. Data used for this study restricted our ability to assess the influence of malpractice liability risk at a more detailed geographic level.
Finally, our analysis attempted to capture the association between high malpractice premiums and changes in the number of physicians in practice but was limited in its ability to account for other potential attributors for physician supply due to data availability. For example, we lacked information on physician reimbursement rates. High malpractice premiums can be offset in areas where physician reimbursement is high. That might contribute to a more stable supply in some states; yet our analysis did not account for local variations in obstetrician supply that may be influenced by reimbursement. Wealthy suburbs may be more likely to have an adequate supply of obstetricians, whereas a nearby poor urban population with poor insurance coverage is likely to have problems retaining obstetric providers. Further research should be completed to better understand these local economic influences on physician supply.
Clearly, we need more research on the interaction of malpractice premiums and obstetrician–gynecologist supply. As a useful first step, our study provided evidence of declining supply in new practitioners because of increasing medical legal premiums. There is legitimate reason for us to be concerned about patients’ future access to obstetric care and critical prenatal care. Both AMA and ACOG will be at the forefront of the liability reform movement to reverse the declining trend of new entries into obstetrics-gynecology. Research findings in this field will assist their decision making regarding malpractice premium management and will provide empirical evidence to inform policy changes and tort reform that could alleviate this situation.
1.Ransom SB, Domrowski MP, Shephard R, Leonardi M. The economic cost of the medical-legal tort system. Am J Obstet Gynecol 1996;174:1903–7.
2.Localio AR, Lawthers AG, Brennan TA, Laird NM, Herbert LE, Peterson LM, et al. Relation between malpractice claims and adverse events due to negligence: results of the Harvard medical practice study III. N Engl J Med 1991;325:245–51.
3.Ransom SB, Studdert DM, Dombrowski MP, Mello MM, Brennan TA. Reduced medicolegal risk by compliance with obstetric clinical pathways: a case-control study. Obstet Gynecol 2003;101:751–5.
4.Studdert DM, Mellow MM, Brennan TA. Medical malpractice. N Engl J Med 2004;350:283–92.
7.Mello MM, Studdert DM, Brennan TA. The new medical malpractice crisis. N Engl J Med 2003;348:2281–4.
8.Medical Liability Monitor. Trends in 2002 rates for physicians’ medical professional liability insurance. Chicago (IL): Medical Liability Monitor; 2002.
9.Sutton PD, Mathews TJ. Trends in characteristics of births by state: United States, 1990, 1995, and 2000–2002. Table 1: Births by race and Hispanic origin of mother: United States and each state, 1990, 1995, 2000, 2001, and 2002. Natl Vital Stat Rep 2004;52(19):1–152. Available at: http://www.cdc.gov/nchs/data/nvsr/nvsr52/nvsr52_19.pdf
. Retrieved February 1, 2005.
10.Area Resource File (ARF). Rockville (MD): U.S. Department of Health and Human Services, Health Resources and Services Administration, Bureau of Health Professions; 2003.
11.Donlen J, Puro JS. The impact of the medical malpractice crisis on OB-GYNs and patients in southern New Jersey. N J Med 2003;100:12–9.
13.Newton DA, Grayson MS. Trends in career choice by U.S. medical school graduates. JAMA 2003;290:1179–82.
14.Andriole DA, Whelan AJ, Schechtman KB. Recent trends in match process outcomes for U.S. senior medical students. Acad Med 2003;78(suppl):S6–9.
15.Nuthalapaty FS, Jackson JR, Owen J. The influence of quality-of-life, academic, and workplace factors on residence program selection. Acad Med 2004;79:417–25.
17.Queenan JT. The future of obstetrics and gynecology. Obstet Gynecol 2003;102(3):441–2.