This study was limited to melanoma and internal cancers diagnosed in female orthopaedic surgeons within fifteen years of the survey date, in order to be comparable with the National Cancer Institute’s Surveillance Epidemiology and End Results (SEER 11) program’s fifteen-year limited-duration prevalence statistics1. On the basis of this criterion, we excluded four separate diagnoses of cancer (one each of the uterus, thyroid, and cervix, and a melanoma) reported in the returned surveys that were initially diagnosed during the years 1973-1991, as these were all diagnosed greater than 15 years prior to our survey date. Additionally, we excluded all reported diagnoses of basal-cell and squamous-cell skin cancer in the 499 study participants, as these were also excluded from the SEER data. No cancers were diagnosed during 1992 to 1993. We calculated the standardized prevalence ratio by dividing the observed number of cancers among female orthopaedic surgeons by the expected number, on the basis of the sex, age (ten-year groups), and race-specific prevalences in the general U.S. population. Confidence intervals and exact p values were calculated by assuming a Poisson distribution for the observed number of cases and using an approximation to the exact Poisson test2. While this method is more commonly applied to standardized mortality ratios, it also provides an accurate test of whether the standardized prevalence ratio departs from unity when more than ten cases are observed3,4.
One limitation of this study is that participants were not asked to provide information regarding their racial and ethnic background. However, the AAOS published 2005 to 2006 statistics5 regarding the racial distribution of (1) all orthopaedic surgeons (male and female combined) who were less than forty, forty to forty-nine, fifty to fifty-nine, sixty to sixty-nine, and seventy years old or more, and (2) female orthopaedic surgeons (all ages combined). We estimated the racial distribution in the study sample by assuming similar age-stratified racial distributions in female orthopaedists as in all orthopaedists combined. This assumption is conservative because a greater proportion of male orthopaedists were white in all age categories, and white women have higher rates of cancer and breast cancer than do minorities in all age categories1. Therefore, under this assumption, the expected number of cancers is slightly overestimated in the study population, leading to an underestimate of the standardized prevalence ratio.
We performed several sensitivity analyses to evaluate the robustness of our main results. First, to evaluate sensitivity to our assumption regarding race, we recalculated standardized prevalence ratios and p values assuming that the racial distributions in all age groups were similar to that reported for all female orthopaedists combined. Orthopaedists who were less than forty years old were less likely to be white than were those who were forty years of age or older5. Therefore, under this assumption, the expected number of cancers will be slightly overestimated in the youngest age group and slightly underestimated in the older groups; the magnitude of the bias is expected to be small overall, although the direction is difficult to predict. Second, to evaluate the influence of potential response bias, whereby women without cancer are less likely to participate, we recalculated standardized prevalence ratios and p values assuming that all nonrespondents had no history of cancer. Statistical analyses were conducted with use of R statistical software (www.r-project.org/).
The 499 practicing female orthopaedic surgeons who responded to our survey had an average age of 47.8 years and had been in practice for an average of 14.6 years (Table I).
Standardized Prevalence Ratio
There were twenty-nine cancers diagnosed within fifteen years of the survey, with breast cancer the most common cancer followed by melanoma (Table II). The respondents with cancer had been in postresidency practice an average of 16.8 years (range, six to forty years) at the time of their response to the survey and an average of 12.3 years (range, zero to forty years) at the time of their cancer diagnosis. Based on the SEER prevalence estimates1, only fifteen cancers overall and seven breast cancers were expected to occur among 499 U.S. women with a similar age and approximate racial distribution as the surgeons in this study (Table III). We estimated a significant 1.9-fold (95% confidence interval, 1.28 to 2.74; p = 0.0021) increased prevalence of cancer overall and 2.9-fold (95% confidence interval, 1.76 to 4.45; p < 0.0001) increased prevalence of breast cancer in practicing female orthopaedic surgeons compared with women in the general U.S. population.
These results were robust to assumptions regarding race and to potential response bias. When we assumed that the racial distribution in all age groups was similar to that reported for all female orthopaedists combined, the results were virtually unchanged (the standardized prevalence ratio was 1.95 [95% confidence interval, 1.30 to 2.80; p = 0.0004] for all cancers and 2.94 [95% confidence interval, 1.79 to 4.54; p < 0.0001] for breast cancer). When we assumed that all 116 nonrespondents had no history of cancer and included them in the analysis, our results, as expected, were attenuated, but still showed a significant excess prevalence (the standardized prevalence ratio was 1.55 [95% confidence interval, 1.04 to 2.22; p = 0.033] for all cancers and 2.34 [95% confidence interval, 1.42 to 3.61; p = 0.001] for breast cancer). Thus, the excess prevalence of cancer observed in this study cannot be explained by differential response rates among women with and without cancer.
The present study found a 1.9-fold increased prevalence of cancer and 2.9-fold increased prevalence of breast cancer in female orthopaedic surgeons compared with U.S. women of similar age and race. Of the fourteen cancers observed in excess of the expected number, thirteen were breast cancers. Thus, the increased prevalence of all cancer was almost entirely attributable to an increased prevalence of breast cancer.
The cause of an increased prevalence of breast cancer in female orthopaedic surgeons is unknown. This cohort may differ from the general population with respect to lifestyle factors related to breast cancer risk contributing to the excess prevalence. Female orthopaedists have high levels of education and income, which are associated with increased breast cancer risk6,7. High socioeconomic status is correlated with other risk factors for breast cancer, including older age at first birth, low parity, and nulliparity8. Other lifestyle factors that increase breast cancer risk include the use of exogenous hormones and alcohol consumption9. Further studies are needed to determine the extent to which the known risk factors for breast cancer can explain the excess prevalence of breast cancer found in female orthopaedic surgeons.
Greater than expected breast cancer prevalence has been reported for other female occupational cohorts with lifestyles and exposures similar to our cohort of female orthopaedic surgeons. For example, the risk of breast cancer for female physicians10,11 and female dentists10,12 is up to 1.7 and 1.6 times greater, respectively, compared with that for the general population. Finnish female physicians exposed to radiation have a 2.3 standardized incidence ratio of breast cancer relative to the entire Finnish population13. Also, female professionals have been found to have an elevated risk of breast cancer mortality11,14. In a prospective cohort study of California teachers, the breast cancer rate ratio was 1.51 (95% confidence interval, 1.48 to 1.54) compared with the general California population15. Our finding of a nearly threefold elevated prevalence of breast cancer was much greater than the elevated risk reported for these other female professional cohorts. Given that the breast cancer risk factors associated with reproductive factors and hormone use are likely to be similar among these cohorts, we suggest that the radiation-associated occupational exposures should be explored as a possible breast cancer risk factor among female orthopaedic surgeons.
This survey had limitations because a complete list of female orthopaedic surgeons is not available from the AAOS, any female orthopaedic surgery membership society, or the orthopaedic surgery subspecialty societies. One limitation is the possibility that all female orthopaedic surgeons in the United States were not captured in our survey study. Another limitation is that we could not include all retired female orthopaedic surgeons. This is because retired female orthopaedic surgeons who are no longer members of the AAOS are not listed in the directory and, therefore, could not be included in our survey.
Furthermore, because of the cross-sectional design of the study, there are other limitations. First, there may have been response bias, in which female orthopaedists with cancer were more likely to respond to the survey than were those without cancer, leading to an overrepresentation of cancer cases in the study sample. However, our sensitivity analysis including all nonrespondents, under the assumption that they did not have cancer, still showed a significant excess prevalence of cancer and breast cancer. Thus, these findings are unlikely to be explained by response bias. While it would be informative to sample nonrespondents to determine if there is any difference between respondents and nonrespondents, the surveys were conducted completely anonymously. We wanted to have a high response rate, and we did not want the personal information in the surveys to deter respondents from participating.
A second type of bias of concern in occupational studies is healthy worker bias16, whereby practicing orthopaedic surgeons may tend to be healthier than the general population because of the physical demands of the occupation. To the extent that female orthopaedists diagnosed with cancer are likely to stop working and discontinue their AAOS membership, the prevalence of cancer diagnosed within the previous fifteen years in this population and the corresponding standardized prevalence ratio may be underestimated in this study.
A third limitation is that there is no published report of breast cancer risk in female orthopaedic surgeons as far as we know. Without the benefit of previous studies, our survey questionnaire and findings were limited. Finally, participants in this survey were not asked to provide information regarding their racial and ethnic background. As explained in the Methods section, the racial distribution for all orthopaedists combined (published in the AAOS 2005 to 2006 statistics5) was used to estimate age-stratified racial distribution for female orthopaedic surgeons, as they were assumed to be similar.
To our knowledge, this study is the first to evaluate and describe increased prevalences of cancer and breast cancer in female orthopaedic surgeons. These findings have important implications for physician education and future research. Further research is needed to evaluate the risk factors for cancer in this group and other medical professionals with similar occupational exposures. We recommend that all female orthopaedic surgeons regularly perform breast self-examinations and adhere to mammography screening guidelines17. We speculate that the increased prevalence of breast cancer in female orthopaedic surgeons may be caused, in part, by chronic occupational exposure to low-dose radiation. Studies of female radiologic technologists have shown an increased breast cancer risk among those of young age exposed to radiation at the highest cumulative levels or before 1940, but not after 194018-20. However, to our knowledge, there is no reported study on breast cancer in female radiologists. Studies of medical radiation exposure have also described increased breast cancer risk in association with radiation therapy for cancer and benign disorders and with diagnostic radiation for tuberculosis, scoliosis, and pneumonia21-23.
NOTE: The authors thank Theresa Keegan, PhD, of the Northern California Cancer Center and Stanford University and Lesley Butler, PhD, of the Department of Environmental and Radiographic Health Sciences, Colorado State University, Fort Collins, Colorado, for their helpful comments. They also thank all female orthopaedic surgeons who responded to this study.
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity.
1. Ries LAGs Melbert D Krapcho M Stinchcomb DG Howlader N Horner MJ Mariotto A Miller BA Feuer EJ Altekruse SF Lewis DR Clegg L Eisner MP Reichman M Edwards BK, editors. SEER cancer statistics review, 1975-2005. Bethesda, MD: National Cancer Institute; 2008. http://seer.cancer.gov/csr/1975_2005/
2. Breslow NE Day NE. Statistical methods in cancer research. Volume II—the design and analysis of cohort studies. IARC Sci Publ. 1987;82:69.
3. Norvell DC Czerniecki JM Reiber GE Maynard C Pecoraro JA Weiss NS. The prevalence of knee pain and symptomatic knee osteoarthritis among veteran traumatic amputees and nonamputees. Arch Phys Med Rehabil. 2005;86:487–93.
4. van der Gulden JW Verbeek AL. Re: “A simple method to calculate the confidence interval of a standardized mortality ratio (SMR)”. Am J Epidemiol. 1992;136:1170–2.
5. Watkins-Castillo S. Orthopaedic practice in the US 2005-2006. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2006.
6. Laden F Spiegelman D Neas LM Colditz GA Hankinson SE Manson JE Byrne C Rosner BA Speizer FE Hunter DJ. Geographic variation in breast cancer incidence rates in a cohort of U.S. women. J Natl Cancer Inst. 1997;89:1373–8.
7. Willey SC Cocilovo C. Screening and follow-up of the patient at high risk for breast cancer. Obstet Gynecol. 2007;110:1404–16.
8. Keegan TH Chang ET John EM Horn-Ross PL Wrensch MR Glaser SL Clarke CA. Recent changes in breast cancer incidence and risk factor prevalence in San Francisco Bay area and California women: 1988 to 2004. Breast Cancer Res. 2007;9:R62.
9. MacMahon B. Epidemiology and the causes of breast cancer. Int J Cancer. 2006;118:2373–8.
10. Rix BA Lynge E. Cancer incidence in Danish health care workers. Scand J Soc Med. 1996;24:114–20.
11. Pepłońska B Szeszenia-Dabrowska N. [Occupational risk factors for breast cancer in the epidemiological studies]. Med Pr. 2001;52:483–95. Polish.
12. Simning A van Wijngaarden E. Literature review of cancer mortality and incidence among dentists. Occup Environ Med. 2007;64:432–8.
13. Jartti P Pukkala E Uitti J Auvinen A. Cancer incidence among physicians occupationally exposed to ionizing radiation in Finland. Scand J Work Environ Health. 2006;32:368–73.
14. MacArthur AC Le ND Abanto ZU Gallagher RP. Occupational female breast and reproductive cancer mortality in British Columbia, Canada, 1950-94. Occup Med (Lond). 2007;57:246–53.
15. Bernstein L Allen M Anton-Culver H Deapen D Horn-Ross PL Peel D Pinder R Reynolds P Sullivan-Halley J West D Wright W Ziogas A Ross RK. High breast cancer incidence rates among California teachers: results from the California Teachers Study (United States). Cancer Causes Control. 2002;13:625–35.
16. Koepsell TD Weiss NS. Epidemiologic methods: studying the occurrence of illness. New York: Oxford University Press; 2003. p 356-7.
17. Saslow D Boetes C Burke W Harms S Leach MO Lehman CD Morris E Pisano E Schnall M Sener S Smith RA Warner E Yaffe M Andrews KS Russell CA; American Cancer Society Breast Cancer Advisory Group. American Cancer Society guidelines for breast screening with MRI as an adjunct to mammography. CA Cancer J Clin. 2007;57:75–89. Erratum in: CA Cancer J Clin. 2007;57:185.
18. Doody MM Freedman DM Alexander BH Hauptmann M Miller JS Rao RS Mabuchi K Ron E Sigurdson AJ Linet MS. Breast cancer incidence in U.S. radiologic technologists. Cancer. 2006;106:2707–15.
19. Mohan AK Hauptmann M Linet MS Ron E Lubin JH Freedman DM Alexander BH Boice JD Jr Doody MM Matanoski GM. Breast cancer mortality among female radiologic technologists in the United States. J Natl Cancer Inst. 2002;94:943–8.
20. Doody MM Mandel JS Lubin JH Boice JD Jr. Mortality among United States radiologic technologists, 1926-90. Cancer Causes Control. 1998;9:67–75.
21. John EM Phipps AI Knight JA Milne RL Dite GS Hopper JL Andrulis IL Southey M Giles GG West DW Whittemore AS. Medical radiation exposure and breast cancer risk: findings from the Breast Cancer Family Registry. Int J Cancer. 2007;121:386–94.
22. Ronckers CM Erdmann CA Land CE. Radiation and breast cancer: a review of current evidence. Breast Cancer Res. 2005;7:21–32.
Copyright 2010 by The Journal of Bone and Joint Surgery, Incorporated
23. Preston DL Mattsson A Holmberg E Shore R Hildreth NG Boice JD Jr. Radiation effects on breast cancer risk: a pooled analysis of eight cohorts. Radiat Res. 2002;158:220–35. Erratum in: Radiat Res. 2002;158:666.