Journal of the National Cancer Institute Advance Access originally published online on August 11, 2009
JNCI Journal of the National Cancer Institute 2009 101(17):1163-1165; doi:10.1093/jnci/djp263
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author 2009. Published by Oxford University Press.
EDITORIALS |
Numbers Needed to Decide
Affiliation of authors: Dartmouth Institute for Health Policy & Clinical Practice, Hanover, NH (SW, LMS); VA Outcomes Group, VA Medical Center, White River Junction, VT (SW, LMS)
Correspondence to: Lisa M. Schwartz, MD, MS, VA Outcomes Group (111B), VA Medical Center, White River Junction, VT (e-mail: lisa.schwartz{at}dartmouth.edu).
Selling screening can be easy. Induce fear by exaggerating risk. Offer hope by exaggerating the benefit of screening. And don’t mention harms. It is especially easy with cancer—no diagnosis is more dreaded. And we all the know the mantra: early detection is the best protection. Doubt it, and someone may suggest you need your head examined.
"If you are a woman over 35, be sure to schedule a mammogram. Unless you’re still not convinced of its importance. In which case, you may need more than your breasts examined." Old American Cancer Society Poster.
Messages selling screening are everywhere. The news regularly tells the story of celebrities asserting that their lives have been saved because of the early diagnosis of a cancer. It is very unusual to hear stories of those injured by overdiagnosis and overtreatment. Popular magazines report emotionally charged but wholly unrepresentative stories about young women with breast cancer and their fears of dying and leaving their young children (1). Medical centers use screening as a business strategy, offering free tests to attract patients (2). Public service announcements—like the American Cancer Society's slogan above—speak for themselves.
Many of the messages use the familiar tactics typified in Colin Powell's 2009 Prostate Cancer Awareness week poster:
Colin Powell says "Get checked. It could save your life." There are MORE CASES of prostate cancer than any other major cancer. Every THREE minutes an American man finds out he has prostate cancer. Nearly 30,000 men will die from prostate cancer this year. (http://www.prostateconditions.org/pcaw-media-kit)for example, highlighting big numbers (the count of 30 000 deaths) rather than the much smaller 10-year risk of dying from prostate cancer (less than 1% until age 70 years) (3).
Sadly, Jørgensen and Gøtzsche's (4) review of pamphlets disseminated by government-run mammography screening programs in seven European countries demonstrated that even nations sell screening. Three-quarters of the pamphlets failed to quantify the benefit of screening; none mentioned the most important harm of screening: overdiagnosis (the detection of cancers never destined to cause symptoms or death).
Consequently, the results of the Gigerenzer et al. (5) large survey in nine European Union countries reported in this issue of the Journal should come as no surprise. Most respondents either overestimated benefit or answered "don’t know" (although considered wrong in the study, this response was actually correct for prostate-specific antigen testing because at the time of the survey, the randomized trials had not been reported).
The findings of Gigerenzer et al. do, however, need to be interpreted with some caution. First, the method for assessing perceptions of the risk reduction from screening is biased toward overestimation. Women were asked to estimate "how many fewer women died from breast cancer in the group who participate in screening compared to women who did not participate"—men were asked the same question about prostate cancer screening. In both cases, responses were selected from a list of alternatives: "0, 1, 10, 50, 100, 200 (out of 1000)" or "I don’t know." The problem is asymmetry: Four of the six numeric choices are "overestimates." So someone guessing has a two in three chance of overestimating. This bias is compounded by the tendency for people to "guess" toward the middle of a set of responses. Finally, because most people have little experience quantifying the absolute risk reduction of any medical intervention, it is hard to know whether overestimation reflects a general problem—not one specific to screening.
Second, it is not clear if the study participants are truly representative of the European Union. The investigators do not report a true response rate, that is, participants divided by eligible participants contacted plus an estimate of eligible participants among noncontacts (6). Response rates to phone-based quota sampling (here potential participants were contacted until the various sampling strata of people of certain age, sex, region, or profession were filled) can be low—and to the extent that nonrespondents differ systematically from respondents—bias may be introduced. We suspect that nonrespondents are less enthusiastic than respondents about screening (that is why they were less interested in participating), so the findings may actually overestimate overestimation.
These cautions, however, do not diminish the importance of the study by Gigerenzer et al. Whether people overestimate the benefit of screening—or have no idea—the problem is the same. Without an accurate sense of how well screening works, people cannot begin to make informed decisions. We need to move from selling screening to helping people realize that screening is a genuine choice. That means routinely giving people the information needed to make these choices.
Unfortunately, promoting informed decision making is a lot harder than selling screening. Informed decision making requires credible information about the benefits and harms of screening. The information can be difficult to get (for many tests, it does not exist). And it can be hard to communicate since screening involves counterintuitive ideas, tradeoffs and numbers.
A big challenge is conveying the counterintuitive idea that screening does not always help—and can even be harmful (14). Surveys have shown that most people believe that cancer screening is almost always a good idea and few believe harm possible.
People need to understand that screening involves important tradeoffs: the chance of a future benefit (avoiding death from a disease) vs the larger chance of immediate harms. The harms can be serious. False-positive results cause anxiety and can lead to invasive and sometimes dangerous testing. Most importantly, screening leads to the overdiagnosis of some cancers never destined to cause harm. People who are overdiagnosed cannot experience any benefit from screening—they can only experience the anxiety of unnecessary diagnosis and the harms of unnecessary treatment.
To understand the true effect of screening, people need the numbers. What is my chance of dying from this cancer if I am not screened? And what are my chances if I am screened? And what about the harms? We attempt to provide this information for breast and prostate cancer screening (Tables 1 and 2). Despite the wealth of published literature on mammography screening, the numbers are still controversial. The database for prostate cancer screening is much smaller—results from the first two randomized trials (11,12), just published, came to different conclusions. The tables therefore are not meant to be the final word on mammography or prostate-specific antigen screening—any of the figures could be criticized—but to convey the order of magnitude of their effects. Furthermore, the data presented are based on averages, so the numbers would be different for men or women at high risk (such as those with a strong family history of early breast or prostate cancer).
|
|
Fortunately, there is growing evidence that people can understand the numbers if they are presented clearly. Structured tables like the screening facts boxes presented here work well, and we suggest using them routinely (15). In a national randomized trial, structured tables improved consumer knowledge and decision making about prescription drugs (16).
And people are starting to demand the numbers. Recently, a group of consumer health advocates and screening experts wrote an open letter to the Times of London criticizing the British National Health Service's mammography pamphlet for selling screening (17). The British government is reportedly rewriting the pamphlet to better present the magnitude of benefits and harms.
Screening can lead to important benefits, but it can also lead to important harms. And the net effect may be a very close call. Screening messages should reflect this complexity. We should not be selling screening. We should be giving people the numbers they need to decide for themselves.
NOTES
The authors cowrote this editorial. The order of their names is arbitrary.
The authors have no financial conflicts to disclose. The views expressed herein do not necessarily represent the views of the Department of Veterans Affairs or the United States Government.
The authors wish to thank Douglas Robertson, MD, MPH, Brenda Sirovich, MD, MPH, and William C. Black, MD, for helpful comments on earlier drafts.
REFERENCES
1. Burke W, Olsen AH, Pinsky LE, Reynold SE, Press NE. Misleading presentations of breast cancer in popular magazines. Eff Clin Pract (2001) 4:58–64.[Medline]
2. PSA screening test for prostate cancer: An interview with Otis Brawley, MD, by Maryann Napoli. (2009) Center for Medical Consumers. http://medicalconsumers.org/2003/05/01/psa-screening-test-for-prostate-cancer/. Accessed July 14.
3. Woloshin S, Schwartz L, Welch H. The risk of death by age, sex, and smoking status in the United States: putting health risks in context. J Natl Cancer Inst. (2008) 100(12):845–853. Erratum in: J Natl Cancer Inst. 2008;100(16):1133.
4. Jørgensen K, Gøtzsche P. Content of invitations for publicly funded screening mammography. BMJ (2006) 332(7540):538–541.
5. Gigerenzer G, Mata J, Frank R. Public knowledge of benefits of breast and prostate cancer screening in Europe. J Natl Cancer Inst (2009) 101(17):1216–1220.
6. The American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys (2006) 4th ed. Lenexa, KS: AAPOR.
7. Gøtzsche PC, Nielsen M. Screening for breast cancer with mammography. Cochrane Database Syst Rev (2006) (4):CD001877.
8. Elmore JG, Barton MB, Moceri VM, Polk S, Arena PJ, Fletcher SW. Ten-year risk of false positive screening mammograms and clinical breast examinations. N Engl J Med (1998) 338(16):1089–1096.
9. Solveig H, Vacek P, Skelly J, Weaver D, Geller B. Comparing screening mammography for early breast cancer detection in Vermont and Norway. J Natl Cancer Inst (2008) 100(15):1082–1091.
10. Zackrisson S, Andersson I, Janzon L, Manjer J, Garne P. Rate of over-diagnosis of breast cancer 15 years after end of Malmo mammographic screening trial: follow-up study. BMJ (2006) 332(7543):689–692.
11. Schroder F, Hugosson J, Roobol M, et al. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med (2009) 360(13):1320–1328.
12. Andriole G, Grubb R III, Buys S, et al. Mortality results from a randomized prostate-cancer trial. N Engl J Med (2009) 360(13):1310–1319.
13. Croswell J, Kramer B, Kreimer A, et al. Cumulative incidence of false-positive results in repeated, multimodal cancer screening. Ann Fam Med (2009) 7(3):212–222.
14. Welch HG. Should I be tested for cancer? (2003) Berkeley, CA: University of California Press.
15. Woloshin S, Schwartz LM, Welch HG. Know Your Chances: Understanding Health Statistics (2008) Berkeley, CA: University of California Press.
16. Schwartz L, Woloshin S, Welch H. Using a drug facts box to communicate drug benefits and harms. Ann Intern Med (2009) 150(8):516–527.
17. Baum M, Mccartney M, Thornton H, et al. Breast cancer screening peril: negative consequences of the breast screening programme. The Times (London) (2009) http://www.timesonline.co.uk/tol/comment/letters/article5761650.ece. Accessed July 9.
Related Article in JNCI
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2009 101: 1159.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||