Fewer than one in five oncologic phase 3 randomized controlled trials (RCTs) had female corresponding authors, but the proportion of women in this authorship role appears to be gradually increasing, investigators report.
“Through identification of the factors associated with gender disparities in RCT leadership, we hope that the academic oncology community will work to better understand and address the underlying reasons for such imbalances,” wrote Ethan B. Ludmir, MD, of the University of Texas MD Anderson Cancer Center, Houston, and associates. The report is in JAMA Oncology.
The authors searched ClinicalTrials.gov in late 2017 for all oncologic phase 3 RCTs. Of the 1,239 they initially turned up, the authors narrowed them down to the 598 that used multiple arms to test a therapeutic intervention, underwent peer review, and published results of primary endpoints. Among the trials, all published between 2003 and 2018, 17.9% had female corresponding authors.
Industry-funded trials, which comprised 77.8% of the sample, had half as many female corresponding authors (14.4%) as those not funded by industry (30.1%) (P less than .001), “possibly reflecting gender biases that are enhanced in the context of industry relationships with academic medicine,” the authors wrote. The opposite trend appeared in cooperative group trials, a quarter of which (25.9%) had female corresponding authors, compared with 14.3% of noncooperative trials (P = .001).
Trials for breast cancer and head and neck cancer were most likely to have female corresponding authors, while the trials with the lowest rates were those for gastrointestinal, genitourinary, and hematologic cancers (P less than .001). The researchers also found gender disparities in the type of intervention tested: Radiotherapy and supportive care studies were more likely to have female corresponding authors, yet none of the surgical trials had any (P less than .001).
In addition, female corresponding authorship was more likely when the institutions were based in the United States (n = 329) than when they were overseas (P = .001). Women were corresponding authors in 22.5% of U.S. studies and 20% of Canadian studies but only 12% of European trials and 2.3% of Asian trials (P = .001).
Within the United States, more than a third of studies from institutions in the Southeast had female corresponding authors (34.1%), followed by those in the Midwest (27.5%) and West (25.9%). Southwestern institutions were least likely to have female corresponding authors (8.7%). Approximately twice as many studies came from Northeastern institutions as from other regions (n = 112), but only 18.8% of these had female corresponding authors (P = .03).
The frequency of female corresponding authors has been increasing, however: The authors calculated a 1.2% increase each year, “echoing data showing an approximate 1.0% annual increase in the number of female academic hematologist-oncologists,” they noted. “However, the absolute female corresponding author rate for these trials is still lower than the percentage of female academic oncologists in this general study period, ranging from 27% in 2000 to 39% in 2015.”
SOURCE: Ludmir EB et al. JAMA Oncology. 8 Aug 2019. doi: 10.1001/jamaoncol.2019.2196.
