“The Feminization of Medicine and Population Health…”

…is the title of a Commentary in the Feb 25, 2009 issue of JAMA, by Susan Philips and Emily Austin from Queens University in Kingston, Ontario, Canada.[1] It is a very interesting piece, citing data and also speculating on the implications of the increasing number of women in medicine. They note that there has been a steady increase in the number of women in medical school for the last 50 years, and that women now comprise 43.5% of graduates of US schools. They enter primary care specialties, however, at much higher rates than their male counterparts. In 2007, 33% of female Canadian graduates chose family medicine as opposed to 22% of males; while either of these numbers (and certainly the mean) would be great by US standards, it demonstrates a “gender gap” even in Canada, where many of the forces discouraging entry into primary care in the US are much lower (there is much less medical debt, income differentials between primary care doctors and specialists are less, there is a system built around primary care, and there is a national health insurance program), there is, apparently, great incentive for medical school graduates, especially men, to choose non-primary care specialties. “Women account for a minority of currently practicing Canadian physicians (37.9%) but a majority of that country’s family physicians (58.6%).” In the US, similarly, “fewer than half of medical school graduates are women (43.5%) but they account for the majority of residents in primary care programs”. The authors indicate that data from the UK, most of Europe, and Australia show similar trends.

It is a good thing that women physicians enter primary care at higher rates than men, because we need more primary care physicians, as demonstrated by much data that has been previously presented in this blog (Dec 11, 2008, and others). However, the authors also note that “Recent reports identifying lower productivity among female physicians have debated whether more women in medicine will exacerbate a shortage of physicians by limiting patient access to care….In general, women are less likely to work excessive hours or to work past the typical age of retirement. Female physicians see fewer patients per hour…”. The impact of this can be understood in terms of Robert Bowman’s Standard Primary Care (SPC) year (“10 myths regarding primary care in the future”, Jan 15, 2009), and suggests that the lower number of SPCs generated by NPs may also be in part associated with gender. In addition, a recent report from the Robert Graham Center cites data showing that while women enter primary care at higher rates than men, they are much less likely to enter rural practice, thus not helping to resolve the rural/urban health disparity problem.[2]

But Phillips and Austin go further. Finishing the sentence above, the write (and cite references for) “Female physicians…demonstrate better communication skills and include more preventive care than their male counterparts.” They note that the issue is health outcomes, not simply access to health care, a point I have repeatedly made. They note that:
“In developed countries, the number of physicians per capita, alone and separated from any analysis of the nature of care provided, has no association with mortality rates. In 1978, Cochrane examined how a number of proximate factors, including gross domestic product, physician density, sugar consumption, and cigarette smoking, were associated with mortality
rates in 1960 and 1970 across 18 developed countries.[3] He found no association between physician density and any of the standard mortality rates and concluded that health service factors were relatively unimportant in explaining differences in mortality among developed countries.”

They go on to note that in comparing physician densities and life expectancy, Canada has the lowest density among developed countries but life expectancies comparable with the Netherlands, the highest density, and better than the US, which is intermediate in density, and that the highest life expectancy is in Japan, with a physician density just higher than Canada’s. They note the work of Macinko and Starfield, which I have referenced before, showing that the strength of a country’s primary care system is most associated with good health outcomes[4], not physician density. Again, as I have discussed before, having more doctors working at the tertiary point of the health workforce pyramid does not improve health outcomes. It is important to note that this is referring to highly developed countries with large physician workforces; clearly there is a threshold (not reached in much of the world, and even in many rural areas of the US) below which there are too few physicians to meet health care needs. But there is also a density at which outcomes plateau, and even, at the highest levels, decreases because of the effect of unnecessary interventions.[5]

This does not, however, mean that having more women physicians will improve the quality of health care, not to mention compensate for their lower number of patients seen and SPC years. It clearly will not meet the needs of those people who live in rural areas, where women, even in primary care specialties, are less likely to practice. Phillips and Austin hope it will:
“Seeing more patients more frequently may not increase the life expectancy of those patients, but spending more time with each patient, hearing and listening more effectively, and including
more preventive measures (all characteristics identified in studies of female physicians) may result in fewer but more effective clinical encounters rather than a greater volume of encounters.”

Perhaps this will be true, I would certainly hope so. They are more certain that the benefit from women becoming primary care doctors will have benefit:
“Because women across time and place tend to become primary care physicians, the feminization of medicine may well have beneficial health outcomes possibly attributable to the nature of the care they provide irrespective of women’s lower volume output relative to that of men. Such an improvement in outcomes may occur because of the practice styles of women, who outnumber men in primary care, or because of the nature of generalism. Either way, as women increasingly enter medicine and become generalists, rather than being a liability by
not working excessively long hours or abandoning parenting, the quality of the care they provide may result in improved population health.”

So to the extent that having more women in medicine leads to more primary care physicians, and because of the convincing evidence that a system built upon primary care leads to improved health outcomes, this is a convincing argument (except for rural areas). The concept of “population health” is critical here, because it includes all people, not a sum of the individual people who have seen doctors.

The degree to which the practice “style” (more time spent with each patient, better communication) will enhance outcomes is more speculative. However, it certainly may, and there are few in the community who would not like to have their physicians spend more time with them and communicate with them better. I hope that this practice “style”, more characteristic of women, becomes more dominant, and I look forward to studies demonstrating that it does improve health outcomes.

[1] Phillips SP, Austin EB, “The feminization of medicine and population health”, JAMA Feb25,09;301(8)363-4
[2] Robert Graham Center, “Specialty and Geographic Distribution of the Physician Workforce: What Influences Medical Student & Resident Choices?”, pp. viii, 20. Accessed 3/4/09, http://www.graham-center.org/online/etc/medialib/graham/documents/publications/mongraphs-books/2009/rgcmo-specialty-geographic.Par.0001.File.tmp/Specialty-geography-compressed.pdf
[3] Cochrane AL, St Leger AS, Moore F. Health service “input” and mortality “output” in developed countries. J Epidemiol Community Health. 1978;32(3): 200-205.
[4] Macinko J, Starfield B, Shi L. The contribution of primary care systems to health outcomes within Organization for Economic Cooperation and Development (OECD) countries, 1970-1998. Health Serv Res. 2003;38(3):831-865.
[5] Goodman DC, Grumbach K. Does having more physicians lead to better health system performance? JAMA. 2008;299(3):335-337.