Representation of Women As Authors, Reviewers, Editors in Chief, and Editorial Board Members at 6 General Medical Journals in 20

Letters

Analysis and interpretation of data: Thomas, Chang, Hess. Table 2. Multivariable Predictors of Sexual Activity Maintenance Drafting of the manuscript: Thomas. 4 Years After Baseline Critical revision of the manuscript for important intellectual content: All authors. Remaining Statistical analysis: Thomas, Chang. Sexually Active, Obtained funding: Hess. Variable OR (95% CI) P Valuea Administrative, technical, and material support: Hess. Age 1.04 (0.96-1.13) .29 Study supervision: Hess. White race (vs other) 3.09 (1.05-9.05) .04 Conflict of Interest Disclosures: Dr Thomas is a women’s health fellow funded through the Department of Veteran’s Affairs. No other disclosures are reported. Education level .60 Funding/Support: STRIDE was supported by grant AG024254 from the High school or less 1 [Reference] National Institute of Health’s National Institute on Aging. Some college 2.78 (0.66-11.74) .17 Role of the Sponsor: The funder had no role in the design and conduct of the Completed college or graduate degree 1.82 (0.50-6.55) .36 study; collection, management, analysis, and interpretation of the data; BMI 0.94 (0.89-0.99) .02 preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Menopausal status .053 Previous Presentation: An earlier version of the data in this study was Premenopausal 1 [Reference] presented orally at the Society of General Internal Medicine (SGIM) Annual Perimenopausal 0.91 (0.17-4.81) .91 Meeting; April 26, 2013; Denver, Colorado. Postmenopausal 0.23 (0.05-1.02) .053 Additional Contributions: We gratefully acknowledge the women who Hysterectomy 0.19 (0.04-0.92) .04 participated in STRIDE. Vaginal dryness (yes) 1.50 (0.62-3.65) .37 1. Morokoff P, Gilliland R. Stress, sexual functioning, and marital satisfaction. JSexRes. 1993;30(1):43-53. Married or in committed relationship (yes) 1.52 (0.56-4.12) .41 2. Ventegodt S. Sex and the quality of life in Denmark. Arch Sex Behav. Importance of sex 3.21 (1.30-7.93) .01 (moderately/quite/extremely) 1998;27(3):295-307. Use of SSRI or SNRI (yes) 0.75 (0.31-1.82) .53 3. Rosen R, Brown C, Heiman J, et al. The Female Sexual Function Index (FSFI): a multidimensional self-report instrument for the assessment of female sexual Abbreviations: BMI, body mass index; OR, odds ratio; SNRI, selective serotonin function. J Sex Marital Ther. 2000;26(2):191-208. reuptake inhibitor; SSRI, serotonin norepinephrine reuptake inhibitor. 4. Winterich JA. Sex, menopause, and culture: sexual orientation and the a P values obtained by multivariable logistic regression Wald test; overall P value meaning of menopause for women's sex lives. Gend Soc. 2003;17(4):627. for multilevel categorical variables obtained by likelihood ratio test. 5. Ginsberg TB, Pomerantz SC, Kramer-Feeley V. Sexuality in older adults: behaviours and preferences. Age Ageing. 2005;34(5):475-480. not accurately reflect what constitutes satisfying sex in this population, yielding falsely low scores. Women who reported greater importance of sex had higher Representation of Women as Authors, Reviewers, maintenance of sexual activity. In contrast, we found that Editors in Chief, and Editorial Board Members sexual function, as measured by the FSFI, is not associated with at 6 General Medical Journals in 2010 and 2011 maintenance of sexual activity. This suggests that the “qual- Although more women continue to enter the medical profes- ity” of sex does not affect whether a woman will continue to sion, disparities between the sexes persist in academic medi- have sex over time. Midlife women have many reasons for en- cine. This gender gap has implications for peer recognition and 1 gaging in sex that go beyond “quality.” academic advancement. In 2006, Jagsi and colleagues re- These findings challenge prior assumptions about female ported that the proportion of women as the first and the se- sexual function in midlife. As we study and care for these Invited Commentary page 635 women, a more nuanced understanding of female sexuality is nior (last listed) physician au- essential. thors of original research significantly increased between 1970 and 2004. Women, however, still represented a minority of the Holly N. Thomas, MD authors of original research and editorials in 6 prominent medi- 2 Chung-Chou H. Chang, PhD cal journals. A related study found a substantial increase in Stacey Dillon, MS the representation of women on editorial boards and as edi- Rachel Hess, MD, MS tors in chief of prominent journals. Using data from January 2010 to December 2011, we de- Author Affiliations: Division of General Internal Medicine, Center for Research termined the proportion of women who were authors of origi- on Healthcare, University of Pittsburgh, Pittsburgh, Pennsylvania (Thomas, nal research or editorials, reviewers, editors in chief, or edi- Chang, Dillon, Hess); VA Pittsburgh Healthcare System, Pittsburgh, Pennsylvania (Thomas). torial board members at 6 general medical journals: Annals of Corresponding Author: Holly N. Thomas, MD, Division of General Internal Internal Medicine (AnnInternMed), BMJ, JAMA, JAMA Inter- Medicine, Center for Research on Healthcare, University of Pittsburgh, 230 nal Medicine (JAMA Intern Med), The Lancet (Lancet), and The McKee Pl, Ste 600, Pittsburgh, PA 15213 ([email protected]). New England Journal of Medicine (NEJM). Published Online: February 10, 2014. doi:10.1001/jamainternmed.2013.14402. Methods | For original research and editorials, we categorized ar- Author Contributions: Drs Thomas and Hess had full access to all of the data in ticles according to the sex of the first and the senior author. Origi- the study and take responsibility for the integrity of the data and the accuracy nal research included original articles (AnnInternMedand NEJM), of the data analysis. Study concept and design: Thomas, Hess. research articles (BMJ), original contributions (JAMA), original Acquisition of data: Dillon, Hess. investigations (JAMA Intern Med), and articles (Lancet). Editori-

jamainternalmedicine.com JAMA Internal Medicine April 2014 Volume 174, Number 4 633

Downloaded From: https://jamanetwork.com/ on 09/29/2021 Letters

Figure. Representation of Women as First or Last Authors of Original Research or Editorials, Reviewers, and Editorial Board Members

Original Research 50 First author female Last author female 40

20 Original Research, % Original Research,

0 Ann Intern Med BMJ JAMA JAMA Intern Med Lancet NEJM

First author female 60/164 204/437 121/308 146/320 98/339 99/418 Last author female 42/164 126/437 80/308 88/320 62/339 92/418

Editorials 40 First author female Last author female 30

Editorials, % Editorials, 10

0 Ann Intern Med BMJ JAMA JAMA Intern Med Lancet NEJM

First author female 24/99 132/482 36/160 34/132 163/629 35/194 Last author female 32/99 139/482 35/160 30/132 137/629 38/194

Female Reviewers 40

10 Female Reviewers, % Reviewers, Female 0 Ann Intern Med BMJ JAMA JAMA Intern Med Lancet NEJM

1027/4423 776/2699 1837/6893 43/159 33/137 284/1710

Female Editorial Board Members 50

30 The findings are based on the 20 number of articles included in the

Members, % analysis. Data were gathered for 10 2010 and 2011 for 6 major medical Female Editorial Board Editorial Female journals. Ann Intern Med indicates 0 Ann Intern Med BMJ JAMA JAMA Intern Med Lancet NEJM Annals of Internal Medicine; JAMA Intern Med, JAMA Internal Medicine; 6/22 8/32 7/24 10/24 9/24 4/18 Lancet, The Lancet; and NEJM, The New England Journal of Medicine.

als included editorials (Ann Intern Med, BMJ, JAMA, and NEJM), Intern Med and NEJM, those who reviewed 2 or more times; and invited commentaries (JAMA Intern Med), and comments (Lan- for Lancet, those who reviewed 5 or more times. We determined cet). We excluded editorials written by editors in chief. We used the sex of the individuals using various methods, including in- published lists to identify reviewers, and, in accordance with Jagsi specting the author’s name, consulting institutional webpages and et al,2 we obtained information about editors in chief and editorial social networking websites, and via Internet searches. board members from editorial mastheads. For BMJ, information on reviewers was available only for 2010. For Ann Intern Med, BMJ, Results | We identified 1999 original research articles, 1867 edi- and JAMA, information was available on all reviewers; for JAMA torials, 16 242 reviewers, 7 editors in chief (2 for JAMA in 2010-

634 JAMA Internal Medicine April 2014 Volume 174, Number 4 jamainternalmedicine.com

Downloaded From: https://jamanetwork.com/ on 09/29/2021 Letters

2011), and 145 editorial board members who met the inclu- Invited Commentary sion criteria. We classified the sex of 98% of the eligible Shattering the Glass Ceiling individuals: 6511 first and senior authors of original research In 1999, I became editor in chief of The New England Journal and editorials, 16 021 reviewers, 7 editors in chief, and 144 edi- of Medicine (NEJM)—the first woman to head a major medical torial board members. journal. At the time, the professional staff consisted of 5 full- Our findings are shown in the Figure. The percentage of time deputy editors, 6 part-time associate editors, 3 statisti- women who were first author of original research ranged from cal consultants, and 1 consultant in molecular medicine. All 23.7% (NEJM)to46.7%(BMJ); for last author, the range was 18.3% but the last consultant were men. Of the 24 editorial board (Lancet)to28.8%(BMJ). The percentage of women who were members, just 3 were women. A few months later, Catherine the first author of editorials ranged from 18.0% (NEJM)to27.4% DeAngelis, MD, became editor in chief of JAMA, a happen- (BMJ); for last authors, the range was 19.6% (NEJM) to 32.3% (Ann stance so surprising that it warranted a front-page article in the Intern Med). The percentage of female reviewers ranged from Boston Globe (where I was quoted as saying, tongue-in- 16.6% (NEJM)to28.8%(BMJ). Four of the editors in chief were cheek, “There goes the neighborhood.”). women and 3 were men. The percentage of women on editorial But despite the dearth of women in the upper reaches of boards ranged from 22.2% (NEJM)to41.7%(JAMA Intern Med). academic medicine and medical journals, there were already changes under way, as documented by Erren et al1 and by Jagsi Discussion | In 2010 and 2011, we found continued increases in et al.2,3 These authors looked at the representation of women the proportion of women among first and senior authors of as authors of original research articles and editorials, as edi- original research in leading general medical journals com- torial board members and reviewers, and as members of medi- pared with the findings by Jagsi et al1 through 2004. The pro- cal school faculty. Together, they showed how the status of portion of women who were authors of editorials, editors in women has changed over the past several decades. chief, or editorial board members also increased. Nonethe- In 1980, when I first joined the staff of the NEJM as assis- less, most research articles and editorials continue to be writ- tant deputy editor, only 12% of first authors of research ar- ten by men. Women accounted for less than 30% of the re- ticles were women2 and there were no women on the edi- viewers at each of the 6 journals. Despite these increases, all torial board.3 At JAMA, there were also no women on the of the leading general medical journals can further improve editorial board, and only 4.6% of first authors were women. the representation of women in many capacities. By 2010-2011, according to Erren et al,1 23.7% of first authors at the NEJM were women, as were 39.3% at JAMA, and women Thomas Christoph Erren, MD, MPH constituted 22.2% of the editorial board at the NEJM and 29.2% Juliane Valérie Groß, MD, MPH at JAMA. Erren and colleagues also looked at 4 other general David Martin Shaw, PhD, MSc, MA, MML, PGCE medical journals—Annals of Internal Medicine, BMJ, JAMA In- Barbara Selle ternal Medicine, and The Lancet—and found considerable varia- tion among the 6. In general, BMJ was at the top in its representation of women (although JAMA Internal Medicine led in Author Affiliations: Institute and Policlinic for Occupational Medicine, Environmental Medicine, and Prevention Research, University Hospital of the percentage of women on the editorial board and had a per- Cologne, University of Cologne, Cologne, Germany (Erren, Groß); Institute for centage of women as first authors of original research compa- Biomedical Ethics, University of Basel, Basel, Switzerland (Shaw); currently a rable to that of BMJ), and the NEJM was at the bottom. medical student, University of Cologne, Cologne, Germany (Selle). To interpret these data requires looking at medical schools, Corresponding Author: Thomas Christoph Erren, MD, MPH, Institute and because authors, editorial board members, and peer review- Policlinic for Occupational Medicine, Environmental Medicine, and Prevention Research, University Hospital of Cologne, University of Cologne, Kerpener ers come from the upper rungs of the academic ladder. What Strasse, 61, D-50938 Cologne, Germany ([email protected]). is the composition of the pool from which the journals draw? Published Online: February 24, 2014. That depends on the level of seniority. In the United States, ap- doi:10.1001/jamainternmed.2013.14760. proximately half of all medical students are now women, but Author Contributions: Drs Erren and Groß had full access to all the data in the men still outnumber women among faculty members, par- study and take responsibility for the integrity of the data and the accuracy of ticularly at the top. (It may be different in the United King- the data analysis. Study concept and design: Erren, Groß, Selle. dom, which might explain the BMJ advantage.) In 2004, only Acquisition of data: All authors. 19% of associate and full professors on the clinical faculties of Analysis and interpretation of data: All authors. US medical schools were women, and women composed 38% Drafting of the manuscript: Erren, Shaw. 3 Critical revision of the manuscript for important intellectual content: All authors. of assistant professors. Calls for journals to invite more women Statistical analysis: Erren. to write nonresearch articles are all to the good4 because they Administrative, technical, or material support: Erren, Groß, Shaw, Selle. give women more visibility, but until women move up the aca- Study supervision: Erren. demic ladder, they will still be underrepresented as authors Conflict of Interest Disclosures: None reported. of research articles. The problem is not so much at the jour- 1. Jagsi R, Guancial EA, Worobey CC, et al. The “gender gap” in authorship of nals as it is at the medical schools. academic medical literature—a 35-year perspective. N Engl J Med. 2006;355(3):281-287. Have women advanced as fast as they should given the fact 2. Jagsi R, Tarbell NJ, Henault LE, Chang Y, Hylek EM. The representation of that they began to enter academic medicine in large numbers 2 women on the editorial boards of major medical journals: a 35-year perspective. in the 1970s? According to Jagsi et al, women made up 14% Arch Intern Med. 2008;168(5):544-548. of all US medical students in 1972 and the same percentage of

jamainternalmedicine.com JAMA Internal Medicine April 2014 Volume 174, Number 4 635

Downloaded From: https://jamanetwork.com/ on 09/29/2021 Letters

full professors among the clinical faculty in 2004. That lag of redefined. I do not advocate a lowered glass ceiling, but rather, 32 years for women to climb from medical student to full pro- placing the ceiling over a different edifice. Research productiv- fessor is approximately 10 years longer than for men. More- ity should no longer be considered the primary measure of aca- over, other data5 confirm that women are disproportionately demic success. If teaching and mentoring are rewarded com- represented at the lower rungs of the academic ladder, then mensurately with research, women will do very well. In fact, they stall and become associate or full professors, if at all, at men might well have to work harder than they are now to catch an older age than their male counterparts. up with women in these areas. In any case, I have no doubt that Why the lag? One factor is probably the reward system in physicians would be better educated and that the medical lit- academic medicine. As Jagsi et al2(p281) point out, “advance- erature would be less voluminous but of higher quality. And the ment is largely driven by peer-reviewed original research,”par- glass ceiling would shatter that much faster. ticularly that published in prestigious journals. Publication is the coin of the realm. I believe men are more likely than women Marcia Angell, MD to devote themselves single-mindedly to research, partly because women are disproportionately tapped for various aca- Author Affiliation: Department of Global Health and Social Medicine, Harvard Medical School, Cambridge, Massachusetts (Angell). demic citizenship duties (every committee needs at least 1 Corresponding Author: Marcia Angell, MD, Department of Global Health and woman), and because the child-bearing years coincide with the Social Medicine, Harvard Medical School, 13 Ellery Square, Cambridge, MA time of applying for first research grants. Research grants im- 02138. pose a fairly rigid schedule that can conflict with the flexibil- Published Online: February 24, 2014. ity needed in those early years. Greater flexibility in the tim- doi:10.1001/jamainternmed.2013.13918. ing of first research grants would help women pursue research Conflict of Interest Disclosures: None reported. careers. Young male physicians are also beginning to value flex- 1. Erren TC, Groß JV, Shaw DM, Selle B. Representation of women as authors, ibility when their children are young because, unlike their older reviewers, editors in chief, and editorial board members at 6 general medical male colleagues, very few of them have spouses that are house- journals in 2010 and 2011 [published online February 24, 2014]. JAMA Intern Med. doi:10.1001/jamainternmed.2013.14760. wives; however, the difficulties are not equal. A final possible 2. Jagsi R, Guancial EA, Worobey CC, et al. The “gender gap” in authorship of explanation for the slow advancement of women—one that will academic medical literature—a 35-year perspective. N Engl J Med. surprise very few women—is good old-fashioned sexism. This 2006;355(3):281-287. is demonstrated by Lawrence Summers, PhD, the former presi- 3. Jagsi R, Tarbell NJ, Henault LE, Chang Y, Hylek EM. The representation of dent of Harvard University, who said the most likely explana- women on the editorial boards of major medical journals: a 35-year perspective. tion for the relatively low numbers of women scientists is Arch Intern Med. 2008;168(5):544-548. that, compared with men, their brains just aren’t up to the 4. Gender progress (?). Nature. 2013;504(7479):188. doi:10.1038/504188a. job.6 (Summers did not explain how their brains evolved fast 5. Leadley J, Sloane RA. Women in US Academic Medicine and Science: Statistics and Medical School Benchmarking Report, 2009-2010. Washington, enough to account for the recent dramatic influx of women in DC: Association of American Medical Colleges; March 2011. science.) 6. Summers LH. Remarks at NBER Conference on diversifying the science & Despite the difficulties, women are pressing hard against the engineering workforce. Office of the President, Harvard University. glass ceiling, and it will inevitably shatter. But progress is too slow. http://www.harvard.edu/president/speeches/summers_2005/nber.php. One reform that should be instituted—not just because it would January 14, 2005. Accessed January 17, 2014. further equality between men and women, but because it would be of great benefit to academic medicine—is to change the re- COMMENT & RESPONSE ward system. Anecdotal evidence suggests that women do more than their share of teaching and mentoring. Those activities Optimizing the Impact of Drugs on Symptom Burden should be a basis for promotion, at least as much as publica- in Older People With Multimorbidity at the End of Life tions are. The primary mission of medical schools is, after all, To the Editor We commend Chaudhry et al1 on their excellent to educate the next generation of physicians. Clinical research article that reported high prevalence of symptoms that nega- and medical practice are important parts of that mission, but sec- tively affect functioning and quality of life in a cohort of older ondary. Faculty researchers often do little or no teaching, yet ad- community-dwelling people with multimorbidity and a life ex- vance rapidly on the basis of their publications, while excellent pectancy of less than 1 year. teachers languish at the lower rungs of the academic ladder. Too In older people, mulitimorbidity often coexists with poly- many men (and women) are doing pedestrian research simply pharmacy, commonly defined as the use of 5 or more drugs. to be promoted. Doubtless, one reason institutions reward re- In her Invited Commentary, Ritchie2 highlights that pharma- search over teaching is that research grants bring in more money cological treatment of 1 symptom may exacerbate another or than does tuition. Yet, medical schools have an obligation to do a coexisting condition, which may in part explain the in- better by their students and the physicians who teach them. In crease in symptoms in this population. recent years, there has been a certain amount of hand wringing To minimize drug-related symptoms in older people at the about the reward system, with some increased recognition of end of life, pharmacological treatments should be prioritized teaching, but there is still nothing close to parity. and rationalized. Symptomatic relief should take preference This is not a matter of suggesting that standards for pro- over preventive treatments, and drug therapies deemed no lon- motion be lowered so that women have an easier time getting ger necessary should be stopped to minimize cumulative drug- to the top. On the contrary, I am recommending that the top be related adverse effects. For instance, among patients in their

636 JAMA Internal Medicine April 2014 Volume 174, Number 4 jamainternalmedicine.com

Downloaded From: https://jamanetwork.com/ on 09/29/2021