A recent study published in February 2017 by Dr. Yusuke Tsugawa and colleagues from Harvard’s T.H. Chan School of Public Health has found that elderly hospitalized patients treated by female physicians are less likely to die within 30 days of admission, or to be readmitted within 30 days of discharge, than those cared for by male physicians (1).
The research team compiled data from over 1 million Medicare patients aged 65 years or older hospitalized with a medical condition and treated by general internists between 2011 and 2014.
Stringent data analysis accounted for patient and physician characteristics, and hospital indicators. This analysis revealed a small but significantly increased adjusted risk difference of mortality within 30 days of hospital admission between patients treated by male and female physicians. A similar result was observed for readmission to hospital within 30 days of discharge. The association was consistent across a variety of clinical conditions and illness severities. To overcome limitations inherent to an observational study, researchers capitalized on the quasi-random assignment of patients to hospitalists. These results confirm that healthier patients do not preferentially choose female doctors.
While the difference in risk seems modest, the authors calculate that 149 elderly patients on Medicare need to be treated by a female physician for one patient to benefit (number needed to treat (NNT) = 149). To put this number into context: a study of a popular statin estimated that 146 middle-aged men with hypercholesterolemia and no evidence of a previous cardiac event needed to be treated with pravastatin to prevent one cardiovascular death (2). Moreover, the researchers estimate that 32,000 annual deaths in the U.S. could be prevented if male doctors achieve outcomes equivalent to their female colleagues. This number would be even higher if the association holds for non-Medicare populations.
While the paper has received mass media attention, it has drawn some criticism. Firstly, the author’s claim a causal link between the physician’s gender and patient outcome. Many argue that this association is correlative and not causative because the study is observational and not a randomized controlled trial (RCT). However, as senior author Ashish Jha writes in his blog, RCTs are often not possible and sometimes inappropriate (3). For instance, there has never been a RCT of smoking yet its causal association with lung cancer is well accepted. Another major criticism points to authors’ assignment of outcome to one physician when truthfully patients are cared for by a team of physicians, nurses and allied healthcare professionals. Indeed, each hospitalization was assigned to an admitting physician based on the physician identification number (the National Provider Identifier in the Carrier File) that accounted for the largest amount of Medicare spending (specifically, Part B spending) during hospitalization. Assigned physicians were responsible for an average of 51.1% of total spending for a hospitalization, arguably representing the physician most influential to the patient’s care.
The results of this study are being taken seriously because of the authors’ robust study design, rigorous analysis, and mindful consideration of potentially confounding factors. The study is particularly impactful in discussing the wage gap between male and female physicians. Indeed, in American academic hospitals the annual salaries of female physicians, after adjusting for all potential confounders, were 8% lower than those of male physicians (4). Additionally, female physicians were less likely to be promoted to full professorships (5). The current study disproves arguments that have been used to justify the wage discrepancies (4).
The authors propose a few explanations from previous studies for the differences in outcomes of care including that female physicians are more likely to practice evidence-based medicine, perform as well or better on standardized examinations and provide more patient-centered care than their male counterparts. However, the purpose of this study was only to measure the difference between patient outcomes. Building on this work, future studies can determine what accounts for the observed differences and provide valuable insights into improving quality of care for all patients regardless of the gender of treating physicians.
- Tsugawa Y, Jena AB, Figueroa JF, Orav E, Blumenthal DM, Jha AK. Comparison of hospital mortality and readmission rates for medicare patients treated by male vs female physicians. JAMA Internal Medicine 2017 Feb 1; 177(2):206-213.
- Simes J, Furberg CD, Braunwald E, Davis BR, Ford I, Tonkin A, et al. Effects of pravastatin on mortality in patients with and without coronary heart disease across a broad range of cholesterol levels. The Prospective Pravastatin Pooling project. European Heart Journal. 2002; 23(3):207-15.
- Jha AK. An Ounce of Evidence | Health Policy [Internet]2016. [cited 2017]. Available from: https://blogs.sph.harvard.edu/ashish-jha/2016/12/22/correlation-causation-and-gender-differences-in-patient-outcomes/.
- Jena AB, Olenski AR, Blumenthal DM. Sex differences in physician salary in us public medical schools. JAMA Internal Medicine. 2016;176(9):1294-304.
- Jena AB, Khullar D, Ho O, Olenski AR, Blumenthal DM. Sex differences in academic rank in us medical schools in 2014. JAMA. 2015;314(11):1149-58.