LAUSR

Surgical Residents and Medical Malpractice To the Editor Thiels et al1 explore an interesting and important area of medical liability—resident physicians and malpractice. We would like to raise 2 issues related to this article, one concerning the analysis and one concerning the data source. The article and the eTable1 report 1548 paid resident malpractice cases in the National Practitioner Data Bank from 2005 to 2014. These 1548 resident cases include both 808 paid malpractice claims and 740 adverse actions. The latter, which include actions such as the suspension of a physician’s medical license, are collected by the National Practitioner Data Bank along with paid malpractice claims. Therefore, in the eTable,1 the reason that “malpractice allegation group” and “outcome” are listed as missing for 740 cases is that these cases are not paid malpractice claims and so these attributes do not apply. The primary data were obtained by searching the legal database Westlaw. There has been a proliferation of articles in the medical literature relying on Westlaw. One of the articles Thiels et al1 cite itself cites an additional 17 articles concerning medical liability that used Westlaw.2 It is important to consider how Westlaw collects medical malpractice case information to understand this database’s limitations. The outcomes of state civil trials, which include medical malpractice trials, are not systematically publically reported by the courts. Rather, verdicts from these trials are collected by jury verdict reporters and are published in private publications that pull together the outcomes of cases. Different jury verdict reporters vary in their comprehensiveness and level of detail. Most jury verdict reporters function at the state level and so their quality also varies by state.3 Moreover, Westlaw contains data from only some jury verdict reporters. As a result of Westlaw’s system for collecting information on medical malpractice cases, data on malpractice cases from states with more comprehensive jury verdict reporters are overrepresented, as are large awards, which are more likely to be picked up by the jury verdict reporters.3 Given that the malpractice case samples obtained from Westlaw are both incomplete and unrepresentative, it is difficult to draw meaningful conclusions from case percentages based on these Westlaw data, as Thiels et al1 report in their article and eTable. For malpractice research in the medical literature using legal databases, we think the expectation should be that both Westlaw and LexisNexis, a Westlaw competitor with a different set of jury verdict resources, be queried, although this still only partially overcomes the methodological limitations we have detailed.