Screening Mammography in Older Women: A Pilot Study of Residents’ Decisions Ross E. G. Upshur Department of Family and Community Medicine University of Toronto Toronto, Ontario, Canada Background: Screening mammography is a commonly employed preventive modality. The employment of mammography in older women is not supported by evidence from clinical trials, largely because elderly women were excluded from such trials. Most guidelines do not recommend routine screening in older women. How residents reason in this gray zone has been subject to little empirical study. Purposes: This study sought to answer two questions: How variable are residents’ decision responses to mammography screening scenarios in older women where there is no clear evidence? What reasons do residents give to justify these decisions? Methods: Residents were asked to respond to four scenarios and give their screening recommendations and the reasons justifying their decisions. Results: There was considerable variability in resident responses to the four scenarios. Only in one scenario was there near unanimity on the preferred screening decision. Resident perceptions of quality of life, longevity and understanding of the guidelines were cited as justification for their decisions. Conclusion: Clinical preceptors should be aware of how the variability of resident perceptions of such factors as quality of life and prognosis may influence decision-making. Teaching and Learning in Medicine, 14(1), 11–14
Copyright © 2002 by Lawrence Erlbaum Associates, Inc.
mammography until age 75 and biennially or at least every three years thereafter with no upper age limit for women with an expected life expectancy of 4 or more years.” Evidence-based medicine is defined as the “the conscientious, explicit and judicious use of best evidence in making health care decisions”7 When evidence is clear, its use in practice may be unproblematic. Proponents of evidence-based medicine recognize the importance of provider and patient values in clinical decision making. Naylor has written about the gray zones of EBM, but there is little empirical study concerning these gray zones.8 Specifically, how provider values, perceptions of quality of life and prognosis influence decision making require further analysis. There are no studies of how practitioners and patients face the gray zones of screening mammography in older women. One published report of two experts in breast cancer screening found considerable divergence of opinion for screening on scripted scenarios.3 The views of family medicine residents on this issue are largely unknown. Therefore, this pilot study sought to answer two questions:
Screening mammography is an established, widely employed technique for the prevention of breast cancer. Despite a recently published meta-analysis questioning the benefits of screening mammograms,1 most cancer control agencies and cancer prevention programs recommend screening mammography.2 The evidence justifying screening mammography derives from large randomized trials with variable age inclusions. The age at which screening should stop is an issue of ongoing controversy.3 Reasonable consensus for screening decisions after the age of 75 is lacking as is compelling empirical data. The U.S. Preventive Services Task Force recommends judgement in screening after the age of 75.4 The Canadian Periodic Task Force recommends annual screening until the age of 69.5 Recently the American Geriatrics Society Clinical Practice Committee published a position statement on breast cancer screening in older women.6 They note the absence of trial evidence supporting screening mammography in older women, but nonetheless recommend, on the grounds of life expectancy and prevalence of disease, that “Physicians should strongly consider recommending annual or at least biennial
This research was supported by a Research Scholar Award from the Department of Family and Community Medicine, University of Toronto and a New Investigator Award from The Canadian Institutes of Health Research Correspondence may be sent to Ross E. G. Upshur, Director, Primary Care Research Unit, Sunnybrook and Women’s College Health Sciences Centre, 2075 Bayview Avenue – Room E349B, Toronto, Ontario, Canada, M4N 3M5. E-mail:
[email protected]
11
UPSHUR
1.
2.
How variable are residents’ decision responses to mammography screening scenarios where there is no clear evidence? What reasons do residents give to justify these decisions?
Materials and Methods Residents in Family Medicine at the Department of Family and Community Medicine at Sunnybrook and Women’s College Health Sciences Centre were invited to participate in the study during a monthly day-long educational session. The study took place during a one-hour session on evidence-based medicine. Residents were given a page describing four scenarios and asked to decide whether to offer screening mammography or not and to give reasons to justify their decisions (see Appendix). Data were collected on categorical data and open-ended questions. Data were tabulated as counts for yes and no answers, and open-ended responses were collated and coded by theme. Illustrative quotations of the reasons given are provided. Residents gave consent to participate.
Results Table 1 lists the features of the participating residents. There were 14 residents in attendance at the seminar from a pool of 24. All 14 residents participated, 4 male and 10 female. They were equally distributed between PGY–1 and PGY–2 years. Table 2 describes the screening decisions. There was near unanimity to screen in the first scenario. However, there was considerable divergence in the other three scenarios. There were no differences between gender and post-graduate year to account for the differences. Table 3 lists the major reasons given to justify decisions. For Scenario 1, 13 residents would offer screen-
ing and only one would not. The existence of age based guidelines influenced decisions to treat. One resident wrote as a justification of her decision to offer screening that: “regardless of her prognosis for colon cancer, she is still in an age group for which some benefit can be had for detecting breast cancer earlier.” Another wrote: “a 65 year old fits into the age group requiring breast cancer screening.” Life expectancy issues featured as well: “She has satisfactory life expectancy and will benefit from early detection and treatment.” The dissenting resident was influenced by the recently published Lancet meta-analysis and the perception of the patient’s life expectancy: “Meta-analysis of large trials showed no decrease in mortality and she already has a poor prognosis from co-morbid illness.” In Scenario 2 the majority of the residents would not offer screening (10 of 14). Quality of life considerations featured in this decision: “Her quality of life may not improve even with early detection and treatment.” Guideline recommendations also featured prominently: “Recommended guidelines are to screen until age 70, one cannot cost-effectively continue to screen the entire population.” The residents offering screening had contrasting views of quality of life issues: “If she would like it and is happy with her quality of life, then screen.” Another wrote: “She is generally healthy and she won’t die of her arthritis.” For Scenario 3, a similar proportion to Scenario 2 of residents would decline to offer screening. Many residents simply stated that she does not fall within the guidelines for screening or that evidence for benefit was absent: “It wouldn’t decrease morbidity and mortality.” Those offering screening pointed out her potentially longer life span and her quality of life: “Breast cancer would dramatically decrease her quality of life. She may have another 15 years, therefore it is better to catch cancer early and treat.” Another stated: “The risk of breast cancer increases with age. She is otherwise Table 3. Major Reasonsa Given to Support Screening Decision Scenario
Table 1.
Demographic Information
Male Female Postgraduate Year 1 Postgraduate Year 2
Table 2. Scenario 1 2 3 4
12
4 10 7 7
Distribution of Decisions Screen
Do Not Screen
13 4 4 7
1 10 10 7
Reasons For
Reasons Against
1
Comorbidity Good prognosis Fits guidelines
Lancet meta-analysis Poor prognosis
2
If patient asks Good QOL Benefits > risks Fits guidelines
Poor QOL Lancet Meta-analysis Against guidelines
3
Good QOL High risk Reasonable life span: 1
Unclear benefit Lancet meta-analysis Against guidelines
4
Honor patient request
Poor prognosis Against guidelines
Note: QOL = quality or life. aMore than one response could be given.
SCREENING MAMMOGRAPHY IN OLDER WOMEN
healthy and may live healthily for many years and benefit from early detection and treatment.” In Scenario 4, the residents were equally divided concerning offering screening mammography. Those offering the intervention were influenced by the patient’s desire for the intervention: “The patient is requesting it.” “Offer it because she wanted it.” “If the patient requests it, [I] should comply.” Some residents would offer the intervention, but attempt to dissuade the woman: “I would not deny her the test but she would not be a surgical candidate, therefore I would discourage her strongly.” Those who would not offer screening uniformly pointed out the poor prognosis of an elderly woman with congestive heart failure: “[This patient has] poor life expectancy and may not benefit from early detection and treatment.” “This patient is more likely to die of heart failure.”
quality of life may be unarticulated to patients and be part of an unvoiced agenda in the clinical encounter that could lead to misunderstanding.9
Conclusion There is considerable variability among residents in a family medicine residency program concerning the use of screening mammography in older women. Perceived prognosis and quality of life, values, and guideline recommendations are important factors justifying decisions to offer screening in the absence of empirical evidence. Clinical teachers should elicit and discuss with residents how their values influence decision-making.
References Discussion This study, though small, indicates variability in decision making among residents. The participants were representative of the broader group of residents at the hospital. The small sample size limits the generalizability of the study results. However, expert opinion was greatly divergent on the same scenarios, so the results may reflect a more pervasive variability among physicians. A future study with a larger sample size is planned. What is striking is the divergence of justifications given for the scenarios. Values and subjective determination of quality of life and prognosis seem to be important determinants of decision making. The influence of guidelines is particularly important. Guidelines were used to justify decisions to screen or not to screen in each scenario. Perceptions of quality of life also played an important role justifying decisions. Patient preferences and values play an important role as half of the residents would screen an 84-year old woman because she had requested the intervention. The divergence of opinion is important from a pedagogical perspective. Preceptors may be unaware of residents’ views on such issues. Resident’s perceptions of
1. Gotzsche PC, Olson O. Is screening mammography for breast cancer justified? Lancet 2000;355:129–34. 2. Mai V, Meuser J. Ontario Breast Screening Program marks 10 years of service. Ontario Medical Review 2000;67;32–34. 3. Byers T, Costanza ME, Kattlove H. Screening mammography: When should it stop? Cancer Practice 1997;5:52–54. 4. Report of the US Preventive Services Task Force. Guide to Clinical Preventive Services; An Assessment of the Effectiveness of 169 Interventions. Screening for Breast Cancer. [Http://wonder.cdc.gov/wonder/PrevGuid/p0000109/p000010 9.ht] Accessed February 26, 2001. 5. Canadian Task Force on Preventive Health Care. Screening for Breast Cancer [http://www.ctfphc.org]. Accessed February 26, 2001. 6. AGS Position statement. Breast cancer screening in older women. JAGS 2000;48:842–844. 7. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn’t. BMJ 1996;312:71–72. 8. Naylor CD. Grey zones of clinical practice: some limits to evidence-based medicine. Lancet 1995;345:840–842. 9. Britten N, Stevenson FA, Barry C, Barber N, Bradley CP. Misunderstandings in prescribing decisions in general practice: Qualitative study. BMJ 2000;320:484–488.
Received 1 March 2001 Final revision received 11 July 2001
13
UPSHUR
Appendix: Four Scenarios For each of the following scenarios, indicate whether you would offer a screening mammogram to the woman and give your justification, why or why not. 1.
A 65-year-old woman who was treated for adenocarcinoma of the colon 6 months ago. Her prognosis for 5 year survival is 40% and for 10 years is 20%. Yes
2.
A 75-year-old woman who is wheelchair bound with arthritis but otherwise healthy. She is alert mentally but lives in a nursing home because of her arthritis. Yes
3.
No
An 80-year-old woman who drives a car and walks two miles each day. She has a few minor ailments but otherwise is healthy. Yes
4.
No
No
An 85-year-old woman with poorly controlled congestive heart failure. She recently read about the benefits of mammography in a magazine and is asking to be screened. Yes
No
Gender Status
Male Female Postgraduate Year-1
Postgraduate Year-2
Scenarios derived from “Screening mammography: When should it stop?” Cancer Practice 1997;5:52–4. Reprinted by permission of Blackwell Science, Inc.
14
