The Demography of Medical Malpractice Suits against Radiologists

Introduction

The specter of malpractice action remains a matter of concern for the specialty of radiology in general and for individual practitioners in particular. An awareness of the basis for judgments has been made manifest by articles that describe the results of a series of suits on a unifying theme, be it common causes like diagnostic or procedural error/or the crafting of language in a radiology report. Focused surveys of radiologists have tended to rely on retrospective comments, which have emphasized the opinions of those who have at least once been sued.

A 2004 American College of Radiology Malpractice Survey attempted to “describe radiologists’ experiences and concerns” with respect to their perception of the medicolegal climate. This report found that 58% of respondents had been a defendant in a medical malpractice lawsuit or were the subject of a claim. Many of the findings in this study related to the subjective experiences and attitudes of the radiologists they surveyed. Yet of the 17 000 radiologists sent an e-mail link to the survey, only 1530 (9%) responded. The authors noted that those who did respond were likely to have a history of being a defendant in a lawsuit and to feel more strongly about malpractice issues than radiologists in general (1).

A tendency of such studies is to highlight concerns, especially among those previously sued, about the prospect of malpractice and perhaps based on recollection to exaggerate its expectancy (2–4). A recent study by Jena et al (5) compared malpractice rates across a wide spectrum of physician specialties. The study included 808 individuals representing 2.4% of all radiologists. In the study, the percentage of radiologists becoming the subject of a malpractice claim approached 7% per annum. Here too, a relatively small sample of the population of radiologists may not be representative of the experience of radiologists in general.

To obtain a more comprehensive assessment, one would need a larger sample to survey the demographic characteristics of those who have and those who have not been sued. This study, encompassing 8265 or 24.0% of all radiologists, enables us to evaluate malpractice information according to a range of parameters. Like the varying distribution of health care delivery throughout the United States, a difference in frequency of suits should be anticipated from state to state. Moreover, varying malpractice suit frequency should also be expected with respect to sex and age of radiologists as well. Such an investigation would require the recording of pertinent statistics independent of radiologists’ opinions.

This investigation seeks to delimit the demographic characteristics of malpractice claims against American radiologists by sex and location and to note the varying percentages of favorable outcomes and award amounts to plaintiffs by state, including lawsuits that were initiated and then withdrawn by plaintiffs or dismissed prior to seeking resolution, as well as all cases in which a settlement or verdict was rendered.

Materials and Methods

This study was approved by the Institutional Review Board of New Jersey Medical School and is compliant with the Health Insurance Portability and Accountability Act. A comprehensive data set composed of sex and age, state of residence, and board certification status of enrollees in a physician panel has become available from the credentialing files of One-Call Medical, a specialized preferred provider organization offering diagnostic imaging services for the workers’ compensation, group health, and auto insurance industries. As of August 2010, One-Call Medical had 8401 radiologists in its network, each of whom has undergone an extensive initial credentialing and/or a periodic re-credentialing process. The credentialing takes place on a 3-year cycle. The data for the cycle from 2007 to 2010 was evaluated. After initial credentialing, enrollees are re-credentialed every 3 years. The data accumulation occurred in the 3-year period between May 2007 and April 2010.

Histories of malpractice activities are derived from two sources. For claims in which a judgment was made against the radiologist, the record of the case was acquired from the files of the National Practitioners’ Data Bank (NPDB). Information available from the NPDB encompasses a short narrative listing the primary allegation, which typically includes the disease process, the specific reason for the claim (most often an error in diagnosis or a procedural complication), the resolution of the suit by either settlement or verdict, and the dollar amount of the award (6).

A radiologist seeking to become or remain a member of One-Call Medical’s panel of image interpreters is also obligated to disclose the circumstances of other malpractice claims in which he or she had been named as a defendant. This includes all claims starting from the beginning of and throughout his or her career, encompassing (a) those that were resolved in favor of the plaintiff, (b) those that were announced but were not pursued before any further action was taken, (c) those that proceeded to some extent and then were withdrawn, for which the circumstances were unclear, and (d) completed claims for which the radiologist was exonerated. Also to be revealed were awards to the plaintiff from direct payment by the radiologist (such payments are not necessarily reported to NPDB). For each of these claims, a short narrative was also required. Recredentialing in One-Call Medical is on a strict 3-year cycle. In preparation for renewal, each radiologist has to provide recent additions, if any, to his or her list of malpractice cases. Updates from the NPDB for each radiologist were also included. Taken together, these data accruals allowed compilation of a comprehensive career-long malpractice profile on every radiologist, which was up-to-date and delineated by number of cases, the specifics of each case (including their location), their clinical focus, and the sex and age of the defendant radiologist.

All statistical analyses were performed by incorporating weights inversely proportional to the probability of sampling to handle the sampling heterogeneity between states—some states have a larger proportion of radiologists enrolled. Standard descriptive statistics, means (with standard deviations), medians (with interquartile ranges [IQRs]), and proportions were calculated by using the svy command in Stata to include the sampling weights. Rates of malpractice claims per state were calculated by using a zero-inflated negative binomial regression model allowing for differences in years at risk. Person-years at risk were calculated by using the assumption that the average radiologist would begin practicing at the age of 27 years, then subtracting this from their listed age in 2010. Incidence rate ratios (IRRs) were calculated in comparison with the rate of malpractice claims for all 36 states. Confidence intervals (CIs) were calculated by using the delta method. Acknowledging that each state is also in the denominator for calculation of the standard error, we estimated the standard error without inclusion of each state as well and found negligible difference in the width of the CIs. Similarly, Poisson regression was used to calculate the IRR for any payment as a result of a malpractice claim compared with the average of all 36 states, adjusted for sex. To determine the association of age, sex, and state with amount of payment we used a general linear model assuming a gamma family distribution. Mean payment per state adjusted for age and sex was calculated from this model for a 50-year-old male radiologist. All analyses were performed with statistical software (Stata 12; Stata, College Station, Tex).

Results

Characteristics of Study Population

Our study covers the interval from 1955 to 2010 and includes the malpractice histories of 8401 American radiologists who were credentialed members of One-Call Medical as of August 2010. No physicians from Hawaii, North Dakota, South Dakota, or Vermont were enrolled with One-Call Medical during the study period. The District of Columbia and 10 states—Alaska, Delaware, Iowa, Idaho, Montana, New Hampshire, Rhode Island, Washington, West Virginia, and Wyoming—each contributed fewer than 50 radiologists to the pool of imaging interpreters. Because of their relatively meager number of enrollees, radiologists from these states were excluded from comparisons between individual states. By providing no state-specific detail about mean or median awards, sex of radiologist, or age of radiologist in states of residence with less than 50 radiologists in One-Call Medical’s panel, we further minimized the risk of reidentification, which could possibly have been discerned if only a few enrollees were present in any state and their specific cases were described. By so doing, we also minimized identifications through membership in a group practice in states in which there are but a few groups, but in which most of the radiologists from that group are enrolled with One-Call Medical. One hundred thirty-six radiologists were excluded due to this reason, leaving an overall total of 8265 radiologists in this study, or 24.0% of the 34 503 radiologists in the United States (7). The number of radiologists contracted with One-Call Medical at the time of the study in each of the remaining 36 states and the percentage of enrollees in One-Call’s panel in relation to the number of radiologists in each state is displayed in Table 1.

Table 1 Enrollees in Each State Compared with the Total Number of Radiologists Practicing in Each State

Note.—The states are ranked by declining percentage of radiologists who are One-Call Medical enrollees. The study encompasses the malpractice histories of greater than 20% of radiologists in 26 states and greater than 10% in 39 states. Overall, nearly 25% of all radiologists in the United States were enrolled with One-Call at the time of the study.

^*Data are from reference 7.

Of the 8265 radiologists enrolled, 6787 were men, 1209 were women, and for 269, a sex was not listed. Thus, among those subjects for whom a sex was specified or discerned from the name, 84.9% were men and 15.1% were women.

Physician age was specified for 7896 of our 8265 subjects. Their mean age was 50.7 years and ranged from 31 to 87 years. Arizona had the youngest average age (48 years), and New Mexico had the oldest (54 years). The mean age of enrolled male radiologists (51.2 years ± 0.2) was greater than the corresponding average for female radiologists (47.8 years ± 0.4). Characteristics of the study population are depicted in Figure 1and Table 2.

Figure 1:

Download as PowerPointOpen in Image Viewer

Table 2 Demographic Information of 8265 Radiologists

Note.— Unless otherwise indicated, data are numbers of radiologists or suits and numbers in parentheses are percentages. Of the 2600 (31.5%) of 8265 surveyed radiologists that indicated they had been sued, the average number of suits as well as the overall payment results are listed.

^*Numbers in parentheses are 95% CIs.

^† Mean age among 7896 radiologists (6593 men, 1176 women) for whom age was specified.

^‡ Numbers in parentheses are the IQR.

^§ Amount of money paid as a result of malpractice action could be determined for 3815 claims.

Frequency of Malpractice Claims in the United States

The study encompasses 4741 claims against 2600 radiologists. The percentage of radiologists surveyed with one or more claims filed against them was therefore 31.5%. Figure 2 shows a map charting the percentage of members of the One-Call Medical network in each state who have ever been the subject of a claim. In five states—Utah, Kansas, New York, New Jersey, and Missouri—more than 40% of radiologists have a history of being the defendant in at least one malpractice suit. In contrast, fewer than 20% of radiologists have ever been sued in six other states—Wisconsin, Alabama, North Carolina, Arkansas, Mississippi, and Nebraska.

Figure 2:

Download as PowerPointOpen in Image Viewer

The average number of suits per radiologist and the average number of suits per radiologist with a history of at least one suit in each state are shown in Table 3. In all, 1515 radiologists have been sued just once in their career, 599 have been sued twice, and 486 have been sued three or more times. Among radiologists who have been sued, the average number of claims filed against each is 1.81 (95% CI: 1.76, 1.86). In six states, this average is exceeded—New York, Indiana, New Jersey, Utah, Florida, and Pennsylvania. Hence in the aggregate, radiologists in these states are much more likely to have been sued multiple times than those in nearly every other state.

Table 3 Claims per Radiologist, Limited to States with 50 or More Enrollees

Note — Numbers in parentheses are 95% CIs. States with fewer than 50 One-Call Medical enrollees were excluded. The number of claims per radiologist was approximately double the national average of 0.569 in New York and Utah. In four states (New York, Utah, Indiana, and New Jersey), radiologists who had been sued at least once averaged approximately two or more suits, indicating a greater likelihood for radiologists practicing in those states to have been sued multiple times, which is also seen in the odds ratios.

^*The average number of claims for all radiologists in each state.

^† The average number of claims against all radiologists who have been sued at least once, which highlights the incidence of multiple claims against a single radiologist in each state.

^‡ The odds ratios for each state having an increased number of suits relative to North Carolina.

A radiologist’s sex and age play a role in the likelihood that he or she has been sued. The percentage of enrollees with a history of at least one suit is plotted with respect to the subject’s age and sex in Figure 3. As expected, as years in practice lengthen so too does the likelihood of being sued at least once. Though the percentage of men sued compared with their female counterparts of the same age is variable by age group, the overall odds of ever being a defendant in a case are greater for men than for women when controlling for age and state of residence (odds ratio = 1.42; 95% CI: 1.23, 1.65).

Figure 3:

Download as PowerPointOpen in Image Viewer

Outcomes of Malpractice Claims in the United States

The amount of money paid to the plaintiff as a result of a malpractice action could be determined for 3815 of the 4741 total claims. The remaining 926 claims outcome did not provide enough information to denote a payment amount or were still pending an outcome. Overall, the number of claims resulting in no payment was 1530 whereas the number of claims resulting in payment to the plaintiff was 2285—a percentage of 59.7%. The majority of the claims not resulting in a payment were cases in which the radiologist was dismissed as a defendant or the claim was withdrawn altogether.

Further information about whether the suit concluded with a verdict or a settlement was available for 2758 cases. Claims were settled before trial in 2293 cases, which vastly outnumbered the 99 listed claims that ended in a court-directed judgment. Although the majority of completed claims were settled in favor of the plaintiff, the defendant was ordered to make payment in 44.4% of cases in which a verdict was rendered (44 of 99). The average award paid on behalf of the radiologist in these cases was $411 112 (95% CI: $246 225, $576 000). The plaintiff received no award in 52 of these cases, while three cases did not have sufficient monetary information. There were 2109 of 2293 settled claims that resulted in payment made on behalf of the radiologist (cases settled out of court), with mean payment of $295 993 (95% CI: $276 911, $316 390) and median payment of $150 000 (IQR = $45 500, $350 000). Claims that were either noted to be pending or for which their narrative did not provide sufficient information about the outcome were excluded from this analysis.

The rates of suit for men and women radiologists per 100 practice-years varied considerably from state to state, as shown in Table 4. Nationwide, men averaged 2.76 suits for every 100 practice-years (95% CI: 2.63, 2.89) while women averaged 2.13 suits (95% CI: 1.87, 2.42). The IRR comparing sex is 1.37 (95% CI: 1.20, 1.56), indicating that males have a higher rate of being sued than their female counterparts.

Table 4 The Rates of Suits for Men and Women Radiologists and the IRR for Each State

Note.—Numbers in parentheses are 95% CIs. Radiologists from New York, Utah, Indiana, and New Jersey accrue, on average, a greater average number of suits per 100 practice-years, while those from Alabama, North Carolina, Virginia, and Mississippi average the fewest. Male radiologists have a higher rate of being sued compared with their female counterparts nationwide (IRR:1.37; 95% CI: 1.20, 1.56).

^*Tennessee was selected as the reference state, as it most closely approximates the nationwide average for both sexes.

Among the various states as listed in Table 5, radiologists from Oregon ranked first in mean payment at $715 707 (95% CI: $176 916, $2 895 365) followed by those from Illinois, New Jersey, Maryland, Alabama, and Ohio, each of whom were assessed average payments greater than $400 000. At the opposite end of the spectrum were radiologists in Nebraska, Utah, Colorado, New Mexico, and Indiana, with average awards ranging from $74 373 (95% CI: $38 692, $142 959) in Nebraska to $171 813 (95% CI: $101 165, $291 797) in Indiana.

Table 5 Adjusted Mean and Median Payments in Descending Order by Mean, with Corresponding 95% CIs or IQR

Note. — Means are given for a 50-year-old male radiologist. Mean payments are greater than median payments, reflecting the disproportionate effect of a few very large awards. Mean payments reveal a greater than ninefold difference between the states with highest and lowest average award, while median payments reflect a greater than 14-fold difference.

Discussion

The 46 states that have radiologic representatives in One-Call Medical’s interpretation panel encompass 94% of the U.S. population of 310 000 000 persons (8). The percentage of the enrollees among all radiologists was greatest in Maryland at 55.9%, but in 23 states more than one-quarter of all radiologists were credentialed. The graph of suits disaggregated by radiologist’s sex and age reveals a significant difference occurs between 50 and 69 years of age, with men at those ages more often sued than their female contemporaries. This discrepancy may be the result of sex-related dissimilarities in case mix or amount and requires further analysis for elucidation.

There is heterogeneity of the likelihood of ever being sued among the radiologists in the various states. The states in which more than 50% of enrollees have been at least once the subject of claims—Utah and New York—are dissimilar in location, population, and ethnicity. States having more than 40% of radiologists at least once the subject of a claim includes those in the East, Midwest, and West. States in which less than 20% of radiologists in our sample were sued include only representatives from the Midwest and South. The frequency of suits is also widely variable. In Utah and New York, the ratio of claims to enrollees is greater than 1, whereas in Alabama, Wisconsin, and North Carolina, the ratio is 0.23 or less. Among those radiologists who have been sued, the discrepancy is still wide. Overall nearly three-fifths of all claims result in a decision and payment for the plaintiff or his or her family.

A record of mean and median payments reveal great variations in the awards to plaintiffs among states in which there were 50 or more radiologists enrolled in One-Call Medical’s panel. Mean payments were uniformly higher than median payments, reflecting the disproportionate effect of several cases with very high payments on overall average designations. Such an inordinate influence is lessened when using the median designation. Although state-based rankings differed slightly with respect to median and mean awards, overall the states in which the payments were either highest or lowest remained so when accounted by either type of average index.

Some limitations in this study must be acknowledged. The location of the radiologist refers to present address. He or she may have moved one or several times, and the current state of residence may not be the same as it was when the claim of malpractice was made. Also, a credentialed radiologist may have been in a practice that had coverage facilities in adjacent states with the alleged incident occurring in the state in which the radiologist did not reside. Furthermore, radiologists participating in teleradiology organizations are apt to have many state licenses. Hence, a claim may be made against him or her about an imaging interpretation or procedural complication rendered for a patient in a remote state with the record of it referred to the physicians’ current address, not the location of the facility in which the patient was situated. Nearly all radiologists brought up for consideration for admission to One-Call’s panel of enrollees were approved except for those in which outstanding disciplinary issues served to disqualify them.

The NPDB records all transactions in which a payment has been made as a consequence of a malpractice judgment that favors the plaintiff, except for instances, presumably uncommon, in which the accused radiologist provides the payment directly to the patient or the patient’s family. The possibility of recall bias must be acknowledged in that that the extent of self-reported instances of initiated malpractice allegations in which no judgment was made against the defendant radiologist may be less than complete. It is presumable that some have either been forgotten by the radiologist or are not disclosed even if remembered, thereby reducing the recorded frequency of such charges. Nonetheless, the comprehensiveness of the available data, encompassing instances of malpractice claims from some of the older radiologists that occurred more than 30 years ago, suggest that widespread underreporting is unlikely.

Moreover, the possibility of selection bias cannot be eliminated, although distinctive evidence of differences between the practice characteristics of the radiologists in One-Call’s panel and all American radiologists were not readily apparent. Generally, when a group of radiologists sought to ally with One-Call Medical, all or nearly all the radiologists in that group submitted applications for enrollment. However, one discernible difference between this study population and radiologists in general is that the proportion of women in this study is less than the 21.3% of total board-certified radiologists who are women (7345 of 34 503) (7).

Temporal trends in suit frequency or award amount cannot be made from the evaluation of these data. This is because the interval between initiation of a suit and its resolution is very variable. In some cases a decision is reached within a year whereas in others the time between filing a claim and rendering a judgment may exceed 5 years or even longer. Thus, all cases settled in one particular year will include a range of durations from commencement of a suit to resolution, making it impossible to closely assess changing patterns of payment on an annual basis.

These limitations notwithstanding, state-by-state data reveal wide variation in the likelihood of a radiologist ever being the defendant in a malpractice suit and dissimilar but equally heterogeneous results for resolved suits with respect to judgment decisions and award amounts to the plaintiffs.

Disclosures of Conflicts of Interest: S.R.B. Financial activities related to the present article: author is the medical director of One-Call Medical. Financial activities not related to the present article: none to disclose. Other relationships: none to disclose. J.S.W. No relevant conflicts of interest to disclose. L.L. No relevant conflicts of interest to disclose. K.S.C. No relevant conflicts of interest to disclose. A.C. No relevant conflicts of interest to disclose. R.P. No relevant conflicts of interest to disclose.

Author Contributions

Author contributions: Guarantors of integrity of entire study, S.R.B., J.S.W., L.L., K.S.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, S.R.B., J.S.W., L.L., K.S.C., R.P.; statistical analysis, all authors; and manuscript editing, all authors