The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 61:844-850 (2006)
© 2006 The Gerontological Society of America
The Impact of Health Status on Physicians' Intentions to Offer Cancer Screening to Older Women
Mitchell T. Heflin,
Kathryn I. Pollak,
Maragatha N. Kuchibhatla,
Laurence G. Branch and
Eugene Z. Oddone
1 Center for the Study of Aging and Human Development and2
Cancer Detection, Prevention, and Control Research Program, Duke University Medical Center, Durham, North Carolina.
3 College of Public Health, University of South Florida, Tampa.
4 Durham Veterans Affairs Medical Center, Health Services Research & Development Center for Excellence in Primary Care, North Carolina.
Address correspondence to Mitchell T. Heflin, MD, MHS, 2511 Blue Zone Duke South, Box 3003, Duke University Medical Center, Durham, NC 27710. E-mail: hefli001{at}mc.duke.edu
 |
Abstract
|
|---|
Background. Screening for breast and cervical cancer reduces disease-specific mortality, but high rates of comorbidity and disability among elderly persons may alter the risks and benefits of screening.
Methods. We performed a mail survey of primary care physicians to estimate the impact of health status on physicians' intentions to offer cancer screening to older women. Respondents were asked to read a scenario about an older woman. Each scenario patient was one of three ages (70, 80, or 90) and had one of three levels of comorbidity and disability. Respondents were asked to estimate the likelihood of offering screening with mammography and Pap smear to these patients on a 5-point Likert scale. A logistic regression compared those physicians somewhat or very likely to offer screening with those less likely to do so. Further analyses examined the characteristics of physicians likely to "overscreen" the frailest older women (<5 years median life expectancy) or "underscreen" the healthiest (>10 years median life expectancy).
Results. Respondents returned 2003 completed surveys (37.4%). Controlling for age and prior screening, higher levels of comorbidity and disability were associated with a significantly lower likelihood of offering screening for both mammography and Pap smear. Nonetheless, a substantial percentage (30.7%) of physicians indicated a high likelihood of offering a frail 90-year-old woman a mammogram, and 13.4% would offer her a Pap smear. In general, overscreening was more common than underscreening. Female gender was associated with "overscreening" with mammography, whereas male gender and lack of board certification predicted "underscreening." Lack of board certification was associated with "overscreening" with Pap smear.
Conclusions. In addition to age, primary care physicians consider health status in deciding to offer cancer screening to older women. Education and guidelines for cancer screening should more explicitly address the risks of overscreening among frail older women.
SCREENING for breast and cervical cancer is a widely accepted part of primary care for women. Guidelines endorse screening women for these treatable cancers in their 50s and 60s (1–3). Beyond this age range, though, recommendations are less clear. Certainly, breast and cervical cancer remain important and preventable causes of morbidity and mortality in women older than 70 years (4–6). Many questions remain, however, about the utility of screening for malignancy in late life. Often, both the screening procedure and the available treatments for cancer have not been tested in older women. Additionally, medical illness and physical disability alter the risk/benefit ratio of screening (7). As well, individual patient preferences with respect to preventive care may change with age. These factors make the decision to offer cancer screening to the older patient more complex. Several studies have demonstrated that women rely on the recommendation of their primary care physician to guide their decisions on screening (8,9). Little evidence exists, though, to explain how physicians make decisions to offer screening to older women, and, specifically, how a patient's health status, including medical illness and physical disability, impact these decisions. We performed a nationwide survey of primary care physicians to examine the association of health status with physicians' self-reported intentions to offer mammography and Pap smear to older women.
|
METHODS
|
|---|
Sample
The survey sample was drawn from the American Medical Association's Physician Masterfile, a regularly updated database of providers nationwide that includes both members and nonmembers of the American Medical Association. A random sample of 6004 physicians currently in active primary care practice anywhere in the United States was selected for this survey, including 3002 internists and 3002 family practitioners. The goal sample size of approximately 6000 assumed a 50% response rate and a 90% power to detect a 1-point difference on a 5-point Likert scale of likelihood of offering cancer screening for each health status category with an 
= 0.05. The size also allowed for comparison of the two provider categories—internal medicine (IM) and family practice (FP).
Survey
In the fall and winter of 2002–2003, physicians were mailed a four-page survey inquiring about their intentions to offer mammography, Pap smear, and colorectal cancer screening to patients described in three different scenarios. This analysis examines responses for mammography and Pap smear. A preface to the survey instructed respondents to assume that all women described in scenarios were asymptomatic and average risk with respect to the cancer in question and were being seen by the physician at a routine visit in the outpatient clinic setting. Three factors were varied between surveys: age (70, 80, or 90 years), screening history (either screened in the past but due now or never screened), and health status (healthy, moderately ill, or frail). The health status variable described a specific level of comorbidity and function as follows: (a) Healthy = no medical conditions and independent in all activities of daily living (ADLs), (b) Moderately ill = osteoarthritis and hypertension and needing assistance with 2 instrumental ADLs (housework and laundry), and (c) Frail = diabetes, nephropathy, hypertension, and stroke and requiring assistance in 2 basic ADLs (bathing and dressing) (Table 1). Respondents were then asked to rate the likelihood of offering cancer screening to the patient described in the scenario on a 5-point Likert scale (1 = Very unlikely; 2 = Somewhat unlikely; 3 = Neither likely nor unlikely; 4 = Somewhat likely; 5 = Very likely). A final section of the survey asked them to describe the demographics of their clinic practice.
By systematically varying the combinations of age (3 options), health status (3 options), and screening history (2 options) for each cancer-specific scenario, we created 18 unique versions of the survey. Each respondent was presented with three scenarios, one for each type of cancer screening. The survey design and analysis were approved by the Duke University Medical Center Institutional Review Board.
Survey Validation
Scenarios were constructed to describe patients with a range of estimated life expectancies. Specifically, the three levels of health status were intended to portray women in the upper quartile (healthy), the middle 50% (moderately ill), and the lowest quartile (frail) of health and life expectancy. Estimates of the median life expectancy for patients in each scenario were derived from a model described by Walter and Covinsky (7). Table 2 presents the actual scenario-specific estimates. Scenarios were tested for logic and readability among 20 geriatricians and primary care physicians leading to minor revisions in descriptions of health status. Subsequently, in a test of the construct validity of the cases, 34 internists and family practitioners provided their own estimates of the life expectancies of the women described in each scenario. This study confirmed validity on two key points: (i) respondents significantly changed their estimates of life expectancy between levels of health status after controlling for age and (ii) respondent estimates of life expectancy appeared to correlate most highly with the intended estimates at the extremes, that is, those with low (<5 years) and high (>10 years) life expectancy.
Mailing for Main Sample
The sample was stratified by care specialty (FP or IM), and physicians were randomly assigned to receive 1 of the 18 versions of the survey. Surveys were mailed with self-addressed stamped envelopes. A reminder postcard followed the initial mailing at 2 weeks. A second mailing to nonresponders was sent approximately 4 weeks later. Due to a low initial response rate, we added a third round of mailings with a five-dollar cash incentive included with the survey (10). Data collection was then completed with a round of telephone calls to administer the survey. The final response rate was calculated using the number of respondents divided by the total sample less those returned as undeliverable. Mailings, follow-up calls, and data collection were conducted by NCS Pearson, Inc. (Edina, MN).
Analysis
Initial review of the frequencies of the likelihood of offering screening revealed a bimodal distribution of responses. Therefore, answers were dichotomized into "more likely to offer screening" (response 4 or 5) and "less likely to offer screening" (response 1, 2, or 3). Logistic regression was used to assess the relationship of health status, age, and screening history with the likelihood of offering screening. Additionally, tests were performed for interactions between the three independent variables.
To understand better the impact of physician characteristics on likelihood of offering screening, survey responses to scenarios portraying the healthiest and most frail patients were examined. Review of the literature and general consensus indicate that patients with <5 years of life expectancy are highly unlikely to benefit from screening, whereas those with >10 years to live are most likely to benefit (11–14). Subgroups of respondents receiving surveys depicting patients with <5 years of life expectancy were examined in bivariate and multivariate models to determine physician characteristics associated with overscreening—defined as being somewhat or very likely to offer a screening mammogram or Pap smear. Conversely, the subgroup of responders receiving scenarios with patients with >10 years of life expectancy were examined for physician characteristics associated with a low likelihood of offering screening (responses 1–3), designated as underscreening. Physician characteristics examined included gender, year of graduation (prior to 1983 vs later), board certification, type of practice (solo or two-person vs other), region of the country, and specialty type (FP or IM). The graduation variable was dichotomized at the observed median year of graduation in our sample.
|
RESULTS
|
|---|
Response
After three rounds of mailings and a round of telephone administration of the survey, 2384 surveys were returned, of which 217 (9.1%) were marked undeliverable. The total response rate of surveys was 37.4% (2167/5787). Of these, 164 (7.6%) were subsequently excluded from the analysis due to missing data on two or more items, leaving a total of 2003 surveys for final analysis. Most respondents were male (74%). Thirty-one percent worked in solo or two-person practice and 36% in group practice. Half graduated prior to 1983, and 83% were board certified in their specialty. Responders did not differ from nonresponders with respect to gender or geographic location. Responders were more likely to be family physicians, board certified in their designated specialty, and to have graduated prior to 1983.
Practice Demographics
Most responders reported that they were engaged in busy clinical practice, seeing a moderate percentage of older women. Most (89%) reported that they spent at least 60% of their time seeing patients, and 69% spent 80%–100% of their time doing so. Seventy-seven percent of responders estimated that at least 40% of their patients were older than 65 years. Most (87%) indicated that among their patients older than 65 years, at least 40% were women.
Mammography
Among all physicians responding to the survey, 67% were somewhat or very likely to offer the patient in the scenario a screening mammogram. Poorer health status was associated with a lower likelihood of offering screening mammography after controlling for patient age and screening history. Specifically, physicians were significantly less likely to offer screening mammography to a woman who was either moderately ill (odd ratio [OR] = 0.66, 95% confidence interval [CI] = 0.51–0.85) or in frail health (OR = 0.42, 95% CI = 0.33–0.55) compared to a woman in good health. Screening history was not significantly associated with physicians' intentions to offer mammography. Figure 1 shows the proportion of physicians who were somewhat or very likely to offer mammography to women in each age and health status category. Notably, a substantial number of physicians intended to offer screening to the 80- and 90-year-old women in frail health (60.0% and 30.7%, respectively). No significant interactions among patient age, health status, and screening history were found.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 1. Percentage of physicians somewhat or very likely to offer mammography by age and health status
|
|
Bivariate analysis of provider characteristics revealed that female physicians were more likely than male physicians to offer mammography to women with <5 years of estimated life expectancy (54.7% vs 40.5%, p =.001) (Table 3). Although the difference did not reach statistical significance, board-certified physicians were more likely than those lacking certification to offer mammography to these women (46.3% vs 36.5%, p =.063). In a multivariate analysis of physician characteristics associated with overscreening, female gender remained the sole significant factor associated with offering a woman with limited life expectancy a screening mammogram (OR = 1.75, 95% CI = 1.21–2.51).
View this table:
[in this window]
[in a new window]
|
Table 3. Overscreening for Breast Cancer: Somewhat or Very Likely to Offer Screening Mammography to a Woman With <5 Years of Estimated Life Expectancy.
|
|
Bivariate analysis of physician characteristics associated with underscreening, that is, indicating a low likelihood of offering a woman with >10 years of life expectancy a screening mammogram, revealed that male physicians were more likely than female physicians to underscreen (16.0% vs 7.8% , p =.0068) (Table 4). Nonboard-certified physicians were also significantly more likely than their certified colleagues to underscreen (22.9% vs 12.3%, p =.0061). Multivariate analysis confirmed that male and nonboard-certified physicians were significantly more likely to underscreen after controlling for other provider characteristics (OR = 2.00, 95% CI = 1.06–3.80 and OR = 1.86, 95% CI = 1.09–3.18, respectively).
View this table:
[in this window]
[in a new window]
|
Table 4. Underscreening for Breast Cancer: Less Likely to Offer Screening Mammography to a Woman With >10 Years of Estimated Life Expectancy.
|
|
Pap Smear
Among all physicians responding to the survey, 48.3% were somewhat or very likely to offer a screening Pap smear to the woman in the scenario. Poorer health status was also associated with a lower likelihood of offering screening Pap smear after controlling for age and history of screening. Physicians were significantly less likely to offer a screening Pap smear to a woman who was either moderately ill (OR = 0.62, 95% CI = 0.47–0.80) or in frail health (OR = 0.32, 95% CI = 0.24–0.43) compared to a woman in good health. A history of normal Pap smears was associated with a significantly lower likelihood of offering cervical cancer screening (OR = 0.66, 95% CI = 0.54–0.81). Figure 2 illustrates the proportion of providers who were somewhat or very likely to offer Pap smear to women in each age and health status category. Physicians' willingness to offer Pap smear dropped noticeably with older age and a decline in health status, with rates for the frail 80-year-old at 34.1% and for the frail 90-year-old at 13.4%. No significant interactions were found among age, health status, and screening history.
Bivariate analysis of physician characteristics associated with overscreening for cervical cancer revealed that board-certified physicians were less likely to offer Pap smear to women with <5 years of life expectancy than were those lacking certification (23.0% vs 33.9%, p =.017) (Table 5). No differences existed with respect to gender, specialty, year of graduation, region of the country, or practice type. In a multivariate analysis of physician characteristics associated with overscreening, board certification status remained significantly associated with a lower likelihood of offering a woman with limited life expectancy a Pap smear (OR = 0.57, 95% CI = 0.36–0.90). Bivariate and multivariate analyses of provider characteristics associated with underscreening for cervical cancer revealed no significant factors associated with this practice pattern.
View this table:
[in this window]
[in a new window]
|
Table 5. Overscreening for Cervical Cancer: Somewhat or Very Likely to Offer Screening Pap Smear to a Woman With <5 Years of Estimated Life Expectancy.
|
|
|
DISCUSSION
|
|---|
This survey was designed to measure the association of health status with primary care physicians' intent to offer screening for breast and cervical cancer. Health status, in this case, combines medical illness and physical function, two important determinants of life expectancy (15,16). The results of this study indicate that physicians consider health status in formulating their intent to offer screening mammography and Pap smear. After controlling for patient age and screening history, health status significantly influenced providers' self-report of intent to offer screening to patients in scenarios, with a 25%–60% lower odds of offering screening as health worsened. Age remained the dominant variable in the model with a 50%–80% lower likelihood of offering screening with each 10-year interval.
Although not true of mammography, a history of normal Pap smear was associated with a lower likelihood of offering screening. This may reflect clinicians' adherence to guidelines suggesting that women with a history of normal Pap smears and no new risk factors can stop cervical cancer screening in late life (13,17). In contrast, the survey also reveals that a significant percentage of physicians intend to offer cancer screening, particularly mammography, to the frailest patients in their practice. For example, 60% of providers were somewhat or very likely to offer mammography to an 80-year-old woman in frail health. Even more intriguing, nearly 31% of responders were somewhat or very likely to offer mammography to a frail 90-year-old woman, whose life expectancy is likely <2 years. The rates were somewhat lower for offering Pap smear (34% for frail 80-year-olds and 13% for frail 90-year-olds). Although these percentages appear higher than one might expect for frail elderly women, they approximate or even underestimate reported rates of receipt of screening found in a recent analysis of two cohorts of older adults in the United States (18,19).
In a study of nearly 5000 Medicare beneficiaries in California, Walter and colleagues examined results of a survey of older women about receipt of screening mammography and Pap smear (18). After controlling for age, they found that health status was not associated with patients' recall of receipt of either test. These results stand in direct contrast to our findings and may indicate an important disconnect between physicians' intentions and practice with respect to cancer screening (at least as reported by patients). Although physicians may intend to consider health status in determining a given patient's eligibility for screening, in actuality, sicker patients are being screened just as often as their healthier counterparts. As a result, chronically ill and functionally disabled elderly patients with limited life expectancy may be overscreened and, thereby, more susceptible to the potential physical and psychological harms of unnecessary testing (20). The difference may reflect a strong preference for screening on the part of many older patients, a phenomenon which has been documented in a recent nationwide survey of adults in the United States (21). Physicians themselves may also have difficulty estimating and communicating a specific patient's health status and candidacy for screening in a brief outpatient primary care visit. As a result, those with limited life expectancy and little potential for benefit may be overscreened, whereas the healthiest cohort may be underscreened if age is the only factor considered.
In an effort to understand physician characteristics associated with the phenomena of overscreening and underscreening, we conducted a subgroup analysis focusing on responses to scenarios presenting the healthiest and frailest women. The framework for this analysis divided scenarios based on the estimated median life expectancy of the fictional patient. For mammography, overscreening was a much more common phenomenon than was underscreening. Female gender and board certification were associated with intentions to offer a frail patient screening mammography, with female gender remaining significant in a multivariate model. Consistent with this finding, male gender and lack of board certification were associated with a higher likelihood of underscreening the healthiest older women. Prior research has documented higher rates of screening mammography among patients of female providers (22–24). It is interesting that one prior study indicated that female physicians may also be more likely to offer screening to younger women (<40 years old) for whom the benefit is less clearly defined (25). The results of this survey demonstrate a similar circumstance for frail, older women. Although female physicians' greater diligence for screening for breast cancer may lead to better mammography rates in their healthy older patients, it may also contribute to overscreening in certain populations with little potential for benefit. One could conclude that a similar phenomenon occurs in the practices of board-certified physicians given that prior studies have demonstrated that this group also performs better on measures of quality of care, including compliance with mammography (26).
The results for the subgroup analysis on cervical cancer screening revealed a very different pattern, with nonboard-certified physicians slightly, but significantly, more likely to offer Pap smears than their certified colleagues. Perhaps most practicing physicians, particularly those with more current knowledge or recent training, follow guidelines that offer relatively clear parameters on when to stop screening with Pap smear (13). In addition, most physicians associate cervical cancer risk with sexual activity in younger patients and understand the real but relatively low risk of dying from this disease in late life (7,13).
There are potential limitations to this study. We relied on patient scenarios to elicit physicians' intent to offer screening. Admittedly, it is impossible to incorporate in a hypothetical case all the factors that influence clinical decisions of this kind. Nonetheless, vignette-based survey research represents an important means of assessing medical decision making. Several recently published studies have used hypothetical cases to examine physicians' attitudes and decisions in response to patients of varying age, health, and ethnicity (27–29). We hoped to simulate the physician's mindset "at the door to the exam room," that is, their intent prior to encountering the patient. Understanding the limitations in measuring reality through hypothetical scenarios, we performed a careful pilot study that demonstrated the construct validity of the components of the scenarios.
Our final response rate did fall short of our initial goal. Indeed, surveying busy primary care clinicians presented major challenges. Using multiple methods to enhance the return of surveys, we were able to achieve an acceptable response rate. One may argue that the modest response rate may have influenced the reported prevalence rates for screening, in that those physicians more interested in screening may have been more likely to complete the survey. This possibility may have inflated our estimates of the rates of screening. As noted above, however, the rates of offering screening approximate or even underestimate those found in recent epidemiological studies of actual or self-reported rates of receipt of screening (18,19).
Summary
The results of this survey indicate that primary care physicians intend to consider health status, along with age, in determining the candidacy of older women for breast and cervical cancer screening. This finding contrasts recent reports of no association between health status and self-reported rates of receipt of screening. Further analysis indicates that female gender and possibly board certification are associated with an increased likelihood of offering screening mammograms to the frailest older women. If, indeed, overscreening for breast cancer reflects the enthusiasm and compulsion of these physicians for preventive care, then they should provide a ready and receptive target for additional education about the real benefits and risks of cancer screening in older adults through regional and national meetings, board review materials, and print and online publications. Additionally, the creation and dissemination of more explicit guidelines on breast and cervical cancer screening in older women may help reduce both overscreening and underscreening in this population. Of course, other important factors, including mandates of health systems and demands of patients, mediate this process as well. Primary care practices or health systems may consider modifying reminders or standards to include clear language on appropriateness of screening, specifically among the oldest, frailest patients. Additional research and education should also focus on how to discuss these difficult decisions with patients themselves.
|
Acknowledgments
|
|---|
We acknowledge the mentoring and support of Harvey Jay Cohen, MD in the completion of this project.
|
Footnotes
|
|---|
Decision Editor: Luigi Ferrucci, MD, PhD
Received October 29, 2005
Accepted March 1, 2006
|
References
|
|---|
- United States Preventive Services Task Force. Recommendations and Rationale for Cervical Cancer Screening. 2003; Available at: http://www.ahcpr.gov/clinic/3rduspstf/cervcan/cervcanrr.htm. Accessed February 12, 2006.
- Smith RA, Cokkinides V, Eyre HJ. American Cancer Society Guidelines for the early detection of cancer, 2005. CA Cancer J Clin. 2005;55:31-44.[Abstract/Free Full Text]
- U.S. Preventive Services Task Force. Screening for breast cancer: recommendations and rationale. Ann Intern Med. 2002;137:344-346.[Abstract/Free Full Text]
- SEER data on cervical cancer., Available at: http://seer.cancer.gov/csr/1975_2000/results_merged/sect_05_cervix_uteri.pdf. Accessed February 12, 2006.
- Armstrong LR, Hall HI, Wingo PA. Invasive cervical cancer among Hispanic and non-Hispanic women—United States, 1992–1999. MMWR. 2002;51:1067-1070.[Medline]
- SEER data on breast cancer., Available at: http://www.seer.cancer.gov/csr/1975_2000/results_merged/sect_04_breast.pdf. Accessed February 12, 2006.
- Walter LC, Covinsky KE. Cancer screening in elderly patients: a framework for individualized decision making. JAMA. 2001;285:2750-2756.[Abstract/Free Full Text]
- Fox SA, Roetzheim RG, Kington RS. Barriers to cancer prevention in the older person. Clin Ger Med. 1997;13:79-95.[Medline]
- Lane DS, Zapka J, Breen N, et al. A systems model of clinical preventive care: the case of breast cancer screening among older women. Prev Med. 2000;31:481-493.[Medline]
- VanGeest JB, Wynia MK, Cummins DS, Wilson IB. Effects of different monetary incentives of the return rate of a national mail survey of physicians. Med Care. 2001;39:197-201.[Medline]
- Walter LC, Lewis CL, Barton MB. Screening for colorectal, breast and cervical cancer in the elderly: a review of the evidence. Am J Med. 2005;118:1078-1086.[Medline]
- Fletcher SW, Elmore JG. Clinical practice. Mammographic screening for breast cancer. N Engl J Med. 2003;348:1672-1680.[Free Full Text]
- American Geriatrics Society Clinical Practice Committee. Screening for cervical cancer in older women. J Am Geriatr Soc. 2001;49:655-657 Available at: http://www.americangeriatrics.org/products/positionpapers/cer_carc_2000.shtml. Accessed February 12, 2006.[Medline]
- American Geriatrics Society Clinical Practice Committee. Breast Cancer Screening in Older Women. Available at: http://www.americangeriatrics.org/products/positionpapers/brstcncr.shtml. Accessed February 12, 2006.
- Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year mortality in older adults: the Cardiovascular Health Study. JAMA. 1998;279:585-592.[Abstract/Free Full Text]
- Inouye SK, Peduzzi PN, Robison JT, Hughes JS, Horwitz RI, Concato J. Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279:1187-1193.[Abstract/Free Full Text]
- Sawaya GF, McConnell KJ, Kulasingam SL, et al. Risk of cervical cancer associated with extending the interval between cervical-cancer screenings. N Engl J Med. 2003;349:1501-1509.[Abstract/Free Full Text]
- Walter LC, Lindquist K, Covinsky KE. Relationship between health status and use of screening mammography and Papanicolaou smears among women older than 70 years of age. Ann Intern Med. 2004;140:681-688.[Abstract/Free Full Text]
- Ostbye T, Greenberg GN, Taylor DH, et al. Screening mammography and Pap tests among American women 1996–2000: results from the Health and Retirement Study (HRS) and Asset and Health Dynamics Among the Oldest Old (AHEAD). Ann Fam Med. 2003;1:209-217.[Abstract/Free Full Text]
- Walter LC, Eng C, Covinsky KE. Screening mammography for frail older women: what are the burdens? J Gen Intern Med. 2001;16:779-784.[Medline]
- Schwartz LM, Woloshin S, Fowler FJ, Welch HG. Enthusiasm for cancer screening in the United States. JAMA. 2004;291:71-78.[Abstract/Free Full Text]
- Levy S, Dowling P, Boult L, et al. The effect of physician and patient gender on preventive medicine practices in patients older than fifty. Fam Med. 1992;24:58-61.[Medline]
- Franks P, Bertakis KD. Physician gender bias in clinical decision making: screening for cancer in primary care. Med Care. 1993;31:213-218.[Medline]
- Lurie N, Slater J, McGovern P, et al. Preventive care for women: does sex of physician matter? N Engl J Med. 1993;329:478-482.[Abstract/Free Full Text]
- Kreuter MW, Strecher VJ, Harris R, et al. Are patients of women physicians screened more aggressively? J Gen Intern Med. 1995;10:119-125.[Medline]
- Tamblyn R, Abrahamowicz M, Dauphinee WD, et al. Association between licensure examination scores and practice in primary care. JAMA. 2002;288:3019-3026.[Abstract/Free Full Text]
- Schigelone AS, Fitzgerald JT. Development and utilization of vignettes in assessing medical students' support of older and younger patients' medical decisions. Eval Health Prof. 2004;27:265-284.[Abstract/Free Full Text]
- Christakis NA, Asch DA. Physician character associated with decisions to withdraw life support. Am J Public Health. 1995;85:367-372.[Abstract/Free Full Text]
- Dries DL, Exner DV, Gersh BJ, et al. Racial differences in the outcome of left ventricular dysfunction. N Engl J Med. 1999;340:609-616.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S. L. McFall and M. Davila
Gender, Social Ties, and Cancer Screening Among Elderly Persons
J Aging Health,
December 1, 2008;
20(8):
997 - 1011.
[Abstract]
[PDF]
|
|
|
Top
Abstract
Methods
