Abstract
Objective To describe physician practices with regard to opportunistic screening for breast cancer in women aged 35 to 49 years and 70 years of age and older, and to identify the determinants associated with the practice of prescribing screening mammography.
Design Postal survey.
Setting Quebec.
Participants Simple random sample of 1400 general practitioners practising in Quebec in 2009.
Main outcome measures Five cancer screening practices among 4 types of female clientele and the factors influencing physicians in their practice of prescribing screening mammography.
Results The response rate was 36%. For women aged 35 to 49 years, more than 80% of physicians reported using practices judged adequate, except for the teaching of breast self-examination and referrals to genetic counseling (60% and 54%). For women 70 years of age and older with good life expectancy, only 50% of general practitioners prescribed screening mammography. For the 70 years of age and older age group without good life expectancy, for whom screening is not indicated, nearly half of physicians continued to do the clinical breast examination and more than one-third reviewed family history. The main determinants for the practice of prescribing mammography are a favourable attitude to screening, screening skills, peer support, belief in the efficacy of mammography, and sufficient knowledge of the issue and of recommendations.
Conclusion Improvements are needed in the practice of teaching breast self-examination to women aged 35 to 49 years and referring them to genetic counseling, as well as in prescribing mammography for women 70 years of age and older who are in good health. Public health actions to improve these practices should focus on physician attitudes and skills and on communicating clearer recommendations.
Aside from skin cancer, breast cancer is the most common form of cancer and the second most common cause of mortality in Canadian women.1 In 2010, there were an estimated 23 200 new cases of breast cancer in Canadian women of all ages, 4450 of which were in women younger than 50 years of age and 6600 in women 70 years of age and older. While breast cancer mostly affects women aged 50 to 69 years, nearly half of new cases are in women who are not a part of the group targeted by routine screening.1
Breast cancer screening programs in Canada targeting women aged 50 to 69 years are creating confusion with respect to opportunistic screening of non-targeted groups. Several studies show that women aged 70 years and older tend to be diagnosed with later stages of breast cancer compared with younger women.2,3 Early detection of breast cancer in women not only improves the chances of survival,4 but also leads to less invasive treatment.5 Breast cancer is less common but more aggressive in women aged 50 years and younger; it is often diagnosed at an advanced stage6 and associated with genetic mutations.
Studies on breast cancer screening practices for women aged 40 to 49 years are mostly American, because the 2002 US Preventive Services Task Force7 recommends mammography screening beginning at age 40 years.8–13 According to these studies, physicians do not follow the recommendations for women aged 40 to 49 years as closely as the recommendations for older women.13,14 In Quebec, all15–18 but one study14 was on routine screening (ages 50 to 69 years).
Recommendations for age limits for breast cancer screening (Box 1)19–23 are not consistent, which can and does result in confusion.
Several studies have examined the determinants of medical practice as it relates to breast cancer screening. Determinants with respect to physician characteristics that favourably influence the practice of prescribing mammograms are younger age,24,25 female sex,9,11,12,15,26 a favourable attitude toward mammography and belief in its efficacy,24,27 a gynecologic practice,8,24,27,28 and thorough knowledge of prevention.10,15,27 Practice setting characteristics that support screening include having a group practice,9 an urban25 and university setting,29 and a low patient load.15 Last, certain patient characteristics favourably influence physician practice; these include the woman’s motivation for having a mammogram, her compliance with the recommendation, and a positive family history of breast cancer.10
Recommendations for age limits for breast cancer screening
-
The 2009 US Preventive Services Task Force recommends mammography for women aged 50 to 74 years,19 whereas the Canadian recommendation applies to women aged 50 to 69 years.20
-
The 2005 American Geriatrics Society Clinical Practice Committee recommends that screening mammography continue until age 85 years, based on state of health.21
-
The 2009 National Comprehensive Cancer Network does not set an upper age limit, but advocates that screening begin at age 40 years.22
-
At present, there are no standardized criteria for genetic screening.23
Because a physician’s recommendation is the most influential factor in a woman’s decision to have a mammogram,30 it is important to understand the practices of front-line physicians.
Thus, the objective of our study is to describe physician practices with regard to opportunistic screening for breast cancer in women aged 35 to 49 years and aged 70 years and older, and to identify the determinants associated with the practice of prescribing screening mammography.
METHODS
A postal survey was sent to a simple random sample of 1400 general practitioners practising in Quebec. The sample was drawn from the list of the Fédération des médecins omnipraticiens du Québec (FMOQ), representing the 8452 general practitioners who practised in Quebec in 2009. Physicians in active practice whose clients included women aged 35 years and older were included.
An anonymous, self-administered questionnaire was developed in French based on questionnaires used in similar studies.14,31–33 The questionnaire was validated by experts and pretested with 10 general practitioners. The variables that were addressed were inspired by the Walsh and McPhee model (1992).34 The descriptive variables concerned breast cancer screening practices (review of family history of breast cancer, instruction in breast self-examination [BSE], performing clinical breast examination [CBE], prescribing mammography, and referral to genetic counseling) among 4 types of patients (women aged 35 to 49 years with or without risk factors [RFs] and women 70 years of age and older with or without good life expectancy [GLE] of 5 years or longer). For referral to genetic counseling, the 35- to 49-year-old group with RFs was divided into 2 groups: with family RFs and without family RFs. The dependent variables corresponded to the practice of prescribing mammography and the other variables related to the characteristics of the physicians and of the patients, as well as organizational or situational factors (Table 1).
List of independent and control variables
The questionnaire was sent out in November 2009, accompanied by a letter of support from the president of the FMOQ. A scientific incentive (summary of the recommendations for breast cancer screening) was offered to all the participants. Two reminders were sent out in accordance with the 2007 Dillman method (adapted).35
This study was approved by the research ethics board at Hôpital Charles LeMoyne in Longueuil, Que.
Statistical analyses
The analyses were performed using SAS and SPSS (version 17). Descriptive analyses were used to describe the physicians’ profiles and practices. Bivariate analyses were performed between the dependent variables and the various independent variables. Spearman ρ, Kendall τ-b, Wilcoxon, Mann-Whitney, and Kruskal-Wallis tests were performed because of the nonnormal distribution.
The dependent variable, ie, the frequency of mammography prescription, was represented by responses in the form of a Likert scale (never, sometimes, often, always) and was considered a continuous variable in order to generate a score from 1 to 4. A score of 1 corresponded to the undesirable practice and a score of 4 corresponded to the optimal practice, based on the existing recommendations and expert opinions (Table 2).19,21,22,36,37 This score was presented for the 4 types of patients. An overall score for mammography prescription (ranging from 4 to 16) for all of the women, regardless of age or RFs, was also developed by adding up all the scores. Screening practices deemed adequate (Table 2) were defined as follows: for young women with no RFs, instruction in BSE and mammography are not indicated. For women 70 years of age and older with GLE, only CBE and mammography are indicated. For women aged 35 to 49 years with RFs, all the breast cancer screening practices are recommended, whereas for women 70 years of age and older without GLE, no practice is indicated. Certain independent variables were also converted into scores: attitude, belief in the efficacy of mammography, peer support, screening skills, barriers, and knowledge of the recommendations and of the issue. These scores were supported by factorial analyses.
Criteria used to determine adequate breast cancer screening practices, for certain risk categories
Finally, multivariate analyses were performed between the dependent and the bivariately significant independent variables for each type of patient. Two confounding variables were integrated into the models, ie, the age and sex of the physician. Three multiple linear regression models (respecting the conditions of homoscedasticity of normality and linearity) and 2 logistic regression models (given that the assumptions of linear regression were not respected) were developed. The α threshold was .05.
RESULTS
The response rate was 36% (460 of 1279); 121 were excluded owing to retirement or incompatible practice. Most participants were 40 years of age and older (78%) and had been practising for at least 10 years (80%), and a little more than half were women (Table 3). Participants were similar to the starting sample according to sex and years in practice. Participants did not differ from the FMOQ population, but the latter differed from the sample in age and years of practice (Table 4).
Characteristics of participants
Comparison of characteristics of participants who were eligible for the starting sample and the FMOQ population
For women aged 35 to 49 years, more than 80% of physicians reported using practices deemed adequate, except for instruction in BSE and referral for genetic counseling (60% and 54%). For women 70 years of age and older with GLE, only 50% of general practitioners prescribed screening mammography. For the 70 years and older age group without GLE, for whom screening is not indicated, nearly half of physicians continued to perform CBE and more than one-third continued to review family history (Tables 5 and 6).
Frequency of breast cancer screening practices self-reported by physicians according to the risk categories of asymptomatic women of different age groups
Frequency of referral to genetic counseling for breast cancer self-reported by physicians according to the risk categories of asymptomatic women of different age groups
The determinants of a mammography prescribing practice that is deemed adequate are a favourable attitude to screening, screening skills, peer support, belief in the efficacy of mammography, sufficient knowledge of the issue and of recommendations, a high number of patients seen in the practice setting, and a high number of women aged 35 to 49 years seen for periodic health examinations (Tables 7 and 8).
Determinants for the practice of prescribing mammography for women 35 to 49 years of age with no risk factors, women 70 years of age and older with a good life expectancy, and women of all ages (correlation and multiple linear regression)
Determinants for the practice of prescribing mammography for women 35 to 49 years of age with risk factors and women 70 years of age and older without good life expectancy (correlation and logistic regression)
DISCUSSION
To our knowledge, this is the first Canadian study to explore the practices and determinants of breast cancer screening for women aged 35 to 49 years with and without RFs and for women 70 years of age and older with and without GLE. These are the groups of women for whom the recommendations are not always clear and for whom physicians might be uncertain which practices to adopt.
With respect to screening practices for women aged 35 to 49 years, more than 80% of physicians reported using practices deemed adequate. However, instruction in BSE presents a problem: while it is not recommended for women with no RFs, 38% of physicians continue to teach BSE,19,36,37 and for women with RFs, nearly 40% of physicians do not teach the BSE even though the experts recommend it. These results are not surprising because some physicians report being uncomfortable with the contradictory information circulating about instruction in BSE.18 In addition, only 54% of physicians refer women with familial RFs for genetic counseling, even though it is recommended.23
Among women 70 years of age and older with GLE, only 50% of physicians prescribe screening mammography. This result is not surprising because the recommendations with respect to the upper age limit contradict one another. However, there is a consensus on limiting screening to women with life expectancies of longer than 5 years,5 and this is why screening is not indicated for women 70 years of age and older without GLE. In spite of this, in our study, several physicians continued to perform CBE and to review the RFs in this group.
While most studies evaluate the determinants for breast cancer screening practices of physicians for women aged 50 to 69 years, our results on the determinants for these practices for younger women and older women are in line with the literature. Thus, a favourable attitude to screening,27 peer support,34 belief in the efficacy of mammography,24,27 and knowledge of the recommendations and of the issue of breast cancer screening10,15,27 are predictors of mammography prescribing. Our results show that screening skills are also a predictor; physicians who believe that they have the skills to perform screening prescribe mammography more often, even when it is not indicated.
Limitations
The main limitation of our study is its response rate; physicians are part of a population that does not tend to respond to surveys. The response rates in the literature for investigations into the same subject are similar, ranging from 20% to 55%.8,9,25–27,29 A non-response selection bias is therefore possible, despite the relative comparability of the participants to the FMOQ population and to the starting sample. A second limitation is that our study involves self-reported practices, which can result in an overestimation of actual practices.15,25 This overestimation could also affect certain determinants such as attitude, skills, and knowledge. A desirability bias is therefore possible, though anonymity might have decreased the effect of this bias. To minimize memory bias, we used short time frames in the questionnaire. Finally, our review of the literature made it possible to identify the confounding factors for which the multivariate analyses were adjusted.
The simple random sample means that our results can be generalized to all general practitioners in Quebec, and the relative comparability of the study participants, both in the sample and in the FMOQ population is a reassuring element.
Conclusion
Breast cancer mortality rates have decreased substantially over the past 25 years thanks to improvements in treatment and screening programs, but much remains to be done. We must pay more attention to women 50 years of age and younger and women 70 years of age and older in good health, because breast cancer in these populations represents a substantial burden and concrete solutions are available to physicians. For young women who are at risk, we must provide instruction in BSE, perform CBE, prescribe screening mammography, and, where appropriate, refer for genetic counseling. For older women in good health, CBE and prescribing mammography are appropriate. To improve these practices, we must act upon physicians’ attitudes and skills and issue clearer recommendations.
Acknowledgments
We would like to thank the Montérégie PQDCS (Quebec Breast Cancer Screening Program) team and Nathalie Bernier for their cooperation in carrying out this study. We would also like to thank the Fédération des médecins omnipraticiens du Québec for its support, the general practitioners for their participation, and the Montérégie Public Health Department for funding this project.
Notes
EDITOR’S KEY POINTS
-
There has been an emphasis on routine breast cancer screening for women 50 to 69 years old, but opportunistic screening for certain groups of women who are not in this age group is also important and needs improvement.
-
The messages and recommendations with respect to breast cancer screening must be clear.
-
We must encourage physicians to provide instruction in breast self-examination and referrals for genetic counseling for young women with risk factors and to prescribe mammography for women 70 years of age and older with good life expectancy.
Footnotes
-
This article has been peer reviewed.
-
La version en français de cet article se trouve à www.cfp.ca dans la table des matières du numéro de janvier 2012 à la page e39.
-
Contributors
Dr Kadaoui contributed to the study design; to data acquisition, analysis, and interpretation; and to writing the article. Dr Guay and Dr Baron contributed to the study design, to data analysis and interpretation, and to writing the article. Dr St-Cerny contributed to the data acquisition. Dr Lemaire contributed to the data analysis and interpretation. All authors reviewed and edited the article and approved the final version for publication.
-
Competing interests
None declared
- Copyright© the College of Family Physicians of Canada