Older Adults and Forgoing Cancer Screening: “I Think It Would Be Strange”

Importance Although there is a growing recognition that older adults and those with extensive comorbid conditions undergo cancer screening too frequently, there is little information about patients' perceptions regarding cessation of cancer screening. Information on older adults' views of screening cessation would be helpful both for clinicians and for those designing interventions to reduce overscreening.

Objective To obtain a deeper understanding of older adults' perspectives on screening cessation and their experiences communicating with clinicians about this topic.

Design Semistructured interview study.

Setting Senior health center affiliated with an urban hospital.

Participants We interviewed 33 older adults presenting to a senior health center. Their median age was 76 years (range, 63-91 years). Of the 33 participants, 27 were women; 15 were African American, 16 were white, 1 was Asian, and 1 was American Indian.

Main Outcome Measures We transcribed audio recordings of interviews and analyzed them using methods of grounded theory to identify themes and illustrative quotes.

Results Undergoing screening tests was perceived by participants as morally obligatory. Although many saw continued screening as a habit or custom not involving any decision, cessation of screening would require a major decision. Many asserted that they had never discussed screening cessation with their physicians or considered stopping on their own; some reported being upset when their physician recommended stopping. Although some would accept a physician's strong recommendation to stop, others thought that such a physician's recommendation would threaten trust or lead them to get another opinion. Participants were skeptical about the role of statistics and the recommendations of government panels in screening decisions but were more favorable toward stopping because of the balance of risks and benefits, complications, or test burdens.

Conclusions and Relevance For many older adults, stopping screening is a major decision, but continuing screening is not. A physician's recommendation to stop may threaten patient trust. Effective strategies to reduce nonbeneficial screening may include discussion of the balance of risks and benefits, complications, or burdens.

Screening for cancer is part of standard medical care, and educational and advocacy efforts for clinicians and the public aim at increasing cancer screening rates.¹ However, the risks and benefits of screening are altered by comorbid illness, poor functional status, or advanced age.^2-5 Positive results from screening tests lead to a cascade of diagnostic and treatment interventions that carry risk.² The risks may be amplified by conditions such as dementia, which make compliance with testing and treatment regimens more difficult.⁶ Recent studies suggest that cancer screening is conducted in many patients who are unlikely to benefit from such testing because of either advanced age or serious illness.^7-10

Drawing on these studies, some experts have called for efforts to reduce cancer screening in populations where it is either nonbeneficial or potentially harmful. For example, the US Preventive Services Task Force³ has begun to issue recommendations for age-based stopping points for some disease screening, such as stopping routine screening for colon cancer at age 75 years or cervical cancer at age 65 years.¹¹ Organizations such as the American Geriatrics Society¹² have recommended an individualized approach to screening decisions for older adults.

Despite the growing consensus that we need to curb overscreening, changing patient and physician behavior will be difficult in light of older adults' highly favorable views of screening. One study¹³ found that most residents of a retirement community planned to continue screening throughout their lives, and 43% would continue screening even against a physician's recommendation. A national telephone survey of adults aged 50 years or older found that only 9.8% had plans to stop screening. These plans were unrelated to self-reported health status or age, with individuals aged 70 years or older no more likely to stop than those aged 50 to 69 years.¹⁴

These attitudes are similar to those seen in Americans more generally; most Americans surveyed see screening as an undisputed good and fail to recognize how screening tests can be harmful or nonbeneficial.^15,16 Positive attitudes may help motivate individuals to undergo screening when their health status or age makes screening tests beneficial, but when they are older or ill, these same attitudes and limitations in understanding may make it hard for them to accept recommendations to stop screening.

Despite the data on older adults' highly positive views of screening, there is little information about how they respond to recommendations to stop screening or what considerations factor into their decisions. We conducted this study to examine older adults' beliefs and opinions about cancer screening and their experiences discussing screening cessation with their physicians. Our goal is to provide information for clinicians having such discussions and to guide the design and implementation of interventions aimed at reducing overscreening. We conducted semistructured interviews about cancer screening decisions with older adults seeking care at a senior health clinic. We chose this open-ended approach because little is known about how older adults will react to recommendations to stop screening or what variables may be most important to their decisions.

This study was conducted in a senior health center affiliated with an urban public hospital and was approved by the Indiana University Institutional Review Board. Potential participants were adults aged 60 years or older attending regularly scheduled primary care or specialty memory practice clinic visits. Participants could have normal cognition or mild cognitive impairment, as documented after comprehensive evaluation in a memory clinic. These persons were included if their Mini-Mental State Examination (MMSE)¹⁷ score was 20 or higher. We included some persons with cognitive impairment because mild cognitive impairment and dementia affect the risk-benefit ratio of screening.^6,18 All persons were judged by their treating physician to have the capacity to give informed consent for study participation and had been making decisions for their own health care, including cancer screening. Participants were ineligible if they had an MMSE score below 20 or did not speak English.

During the recruitment period, research assistants attended clinic sessions and considered all patients attending the session for enrollment. Medical records were reviewed to determine eligibility and we requested permission to enroll from each participant's physician.

Data collection included demographic variables, information on chronic health conditions, and recent experiences with cancer screenings and vaccinations (Table 1). Semistructured face-to-face interviews were conducted in private clinic rooms by 1 of 2 trained research assistants (seeeAppendix ). The first section included open-ended questions about patient perceptions and recent experiences with screening decisions. The second section probed a list of potential decision-making factors, such as physician recommendations, statistical data about the test, and hypothetical conditions such as living in a nursing home. Each person was asked to respond to the potential impact of the factor on his or her decision-making process. Audio recordings of the interviews were transcribed for analysis. Transcribed interviews and codes were entered into NVivo software (version 8; QSR International) to allow for sorting of the coded segments of text.

We conducted a thematic analysis, informed by methods of grounded theory.¹⁹ This approach relies on careful reading of the text to identify major themes in the data. The first 5 interviews were coded by all investigators using open and axial coding.^19,20 During the initial reading, phrases, sentences, or longer segments of text were identified by the investigator and labeled (open coding). Codes were based on the important concepts that emerged from the interviews rather than on the interview questions or other predefined categories. Team meetings were used to discuss the emerging codes. After the open coding process, the codes were organized into larger categories to reflect the major themes (axial coding). Subsequent interviews were read by all investigators but coded in detail by 1 of 3 investigators (A.M.T., P.H.S., or G.A.S.). At each team meeting, we discussed emerging themes and selected representative quotes. Subsequent interviews led to revision and refinement of the themes. Minutes from these meeting were used by one of us (A.M.T.) to write descriptive memos for each theme.

We used several strategies to establish trustworthiness of the data. Interviews were read by all investigators; disagreements about the meaning of themes or codes were discussed at team meetings and resolved by consensus. We included investigators with varied backgrounds, including a physician with a doctorate in philosophy (P.H.S.), a geriatrician (G.A.S.), and a general internist (A.M.T.) and nonphysician research staff with backgrounds in education (K.M.) and health care compliance (L.R.H.). Conducting interviews until theme saturation was reached also improved the trustworthiness of the data by ensuring that key themes had been successfully identified.

Of 42 potential participants, 33 patients enrolled, 2 were not approached because the physician did not approve participation, 3 did not demonstrate understanding of the study during the consent process, and 4 refused to participate. The participants' median age was 76 years (range, 63-91 years). Twenty-seven of 33 were women; 15 were African American, 16 were white, 1 was Asian, and 1 was American Indian (Table 1). The mean length of the interviews was 25 minutes (range, 13-40 minutes). We found that patients considered screening a moral obligation and did not consider continuing screening as a decision. Few had discussed screening cessation with a physician.

Patients' highly favorable views of screening were reflected in the many reasons to screen that they spontaneously expressed (Table 2), but overriding these specific reasons was a sense that screening was a moral obligation. For instance, an 84-year-old woman described the choice between stopping and continuing screening:

I think I should, because [stopping] would be the same as me taking my life. And that's a sin.

Screening was also equated directly with health and life:

I know that I need that done because I love my life, you know. I want to live.

The continuation of screening typically was characterized not as a decision but as something done automatically. A 91-year-old woman discussing her prior screening experiences stated,

Whenever we were told we were supposed to do something, I just, you know, did it.

For some patients, continued screening occurred as a matter of routine or habit, sometimes in response to a physician prompt, a form letter, or other brief reminder. Some expressed an expectation for an annual test. One 75-year-old woman said that she decided to have a mammogram “because I was getting them every year.”

Others simply went along with the plan for screening with little discussion or understanding of the purpose of the test:

Interviewer: And what went into the decision to have the mammogram done?

Participant: I’m not sure, other than it was a good thing to have on record.

When physicians recommended continuing screening, patients generally accepted the recommendation without further questioning or discussion:

Doc said I needed it, and I . . . told them to go ahead and do it.

Many participants stated that they had never discussed stopping with a physician or considered stopping on their own. Patients described discussions about both continuing and stopping screening as being very brief and containing little information. Many did not express any discomfort with this, suggesting that they did not have an expectation of receiving more information:

Interviewer: Can you tell me about what the doctor said to you the last time you made a decision about cancer screening?

Participant: Just recommended a mammogram.

Interviewer: [D]id you discuss the benefits or the harms of the test?

Participant: No.

Interviewer: Was there anything that the doctor said or did that you found especially helpful during that conversation?

Participant: No, not really. . . . Just seemed unconcerned and just a . . . typical test.

An 83-year-old woman described accepting her physician's recommendation to stop screening mammography after being given very little information:

She just looked at my paper and said she didn't think I needed them anymore.

Only 1 patient whose physician recommended against screening because of the patient's age was upset by the recommendation.

Some patients spontaneously mentioned factors that would influence a decision to forgo screening, such as poor health or the burden on others (Table 3). During the interview, we also asked participants to consider some factors that may influence their screening decisions; we describe several that elicited especially strong reactions (see Table 4 for all factors).

When asked how they would respond if their physician recommended against a cancer screening test, many participants said they would question the physician's recommendation or seek a second opinion:

I think it’d be strange for a doctor to advise not having that done, but I would take his word for it too, and ask why.

Some participants expressed disbelief that a physician would ever recommend stopping cancer screening:

You’ve been told something, this cancer can be prevented if you have these tests and then all of a sudden he tells me that I wouldn't benefit from that anymore. I couldn't see why.

In contrast, a few participants voiced unquestioning acceptance of the physician's recommendation either to screen or to forgo screening.

There were many negative responses to recommendations from experts or government panels. Common themes included the fact that guidelines for cancer screening and other healthful behaviors had changed in the past:

I have heard so much of the government changing their minds on this . . . just like for example, coffee's not good for you, then coffee's right for you, you know . . . I don't have too much faith in some of them.

Others raised concerns about financial incentives for the government to recommend against screening.

Several participants expressed the opinion that statistical information is not relevant to individual decision making:

I would not consider myself part of the statistics, and I think each individual person is different.

Others expressed optimism that even if the chance of benefit was low, they might be the one to benefit from the test.

Many participants expressed that they would be convinced to stop screening owing to the burdens of the test or if the burdens outweighed the benefits. Several brought up their experiences with colonoscopy as an example of a highly burdensome procedure:

I don't want to, mainly because I don't like that stuff you have to drink and then the fact that you have to spend so many hours on the toilet.

Our study of older adults attending an urban senior health center demonstrated that these adults felt a strong moral obligation to continue cancer screening. Although participants viewed undergoing screening as not a decision—because it is automatic, recommended, or morally obligatory—many did see cessation of screening as requiring a major decision. These findings emphasize how public health campaigns are working well to communicate the health benefits of screening, so much so that patients do not even consider undergoing screening as something to decide. Patients expect clinicians, like the rest of the health care establishment, to encourage and endorse screening. Given the steady drumbeat of physicians and public health campaigns in favor of screening, patients may find a recommendation to stop, once they reach a certain age, to be jarring. Our finding of a positive attitude toward screening is consistent with previous research finding widespread support for screening, among US adults,¹⁵ older adults in the United States,^13,14 and physicians.²¹

Public health education and physician endorsement may have created a high degree of “momentum” for continuing screening, even in situations in which the benefits may no longer outweigh the risks or burdens. In this case, an active force may be needed to slow down or stop the behavior. Adults considering cancer screening tests report that clinicians are their most important source of information.²² Some of our findings may guide the crafting of messages that are acceptable to patients and effective in reducing unnecessary screening. For example, framing the conversation in terms of increasing burdens in relation to benefits seems most acceptable to patients and may maintain or promote trust. This is likely to be more effective than citing statistics about the patient's likelihood of benefiting from the test or national guidelines about when to stop screening. In our study, when asked how they would respond to a physician's recommendation to stop screening, many participants said that they would question the recommendation or seek a second opinion, and some expressed the belief that a physician would never make such a recommendation. Although some participants stated that a hypothetical recommendation to stop would threaten their trust, such recommendations might be more effective in the context of an established trusting relationship with a primary care provider.

We found that patients were skeptical and suspicious about the recommendations of experts and government panels to stop screening, expressing the concern, for example, that such recommendations were formulated to save money. Patients also expressed limited trust in statistics. A panel's recommendation against screening may be based on complex statistics on estimated remaining life expectancy, risks, and rates of adverse effects, but our findings demonstrate the danger of assuming that patients will consider these statistical data helpful. Although some experts recommend that physicians incorporate information about individual patients' prognoses into their screening recommendations,²³ the same distrust and lack of understanding of statistics may keep such approaches from being persuasive. Patients may well believe that the statistics simply do not apply to them, as one of our participants said.

In contrast, our participants seemed to respond well to the idea that screening does not make sense if the burdens—including pain, time requirements, or stress—are expected to outweigh the benefits. This way of explaining why the physician is no longer recommending screening seemed to hold some promise for both patient understanding and acceptance. Of note, the burdens of colonoscopy were repeatedly cited as reasons not to continue with this screening test as the patient aged. Some participants also reported that they would not wish to keep undergoing screening if they lived in a nursing home, perhaps because such residence was seen by many of them as a proxy for imminent death and a low quality of life. In this setting, the benefits of life extension might seem limited, justifying cessation of screening.

The emphasis on decreasing benefits and increasing risks of screening may help patients to understand that at certain ages, or with certain comorbid conditions, other goals of care may have a higher payoff for improving quality of life and even extending life. Just as the public health and medical message to patients is that screening is a wise and admirable choice when they are middle aged and otherwise healthy, so must messages explain why not screening, and focusing on other aspects of care, may be wise and admirable when patients are older or ill.

This study has several limitations. It was conducted in a single public hospital setting and cannot be generalized to other settings. A high proportion of participants were female. Most participants were considering the cessation of screening tests as a hypothetical rather than real choice. Patient response may have been different if in fact each patient's personal physician had proposed stopping screening tests. We deliberately included some participants with known mild cognitive impairment who may have had diminished understanding of these issues. However, such individuals were still judged to have decision-making capacity by their physicians and would probably be making such decisions in routine clinical practice.

In conclusion, messages about the benefits of cancer screening have been so effective that patients continue screening out of a moral obligation or even a custom. When crafting messages to promote reconsideration of cancer screening, clinicians should be aware that discussions about advanced age, statistics, and the recommendations of government panels may at best be unconvincing to older adults and at worst diminish trust. Messages about increasing risks and burdens of screening are likely to be more effective, but because of deeply held beliefs about the benefits of screening, messages delivered in a physician-patient encounter may not be enough to change patient practice. Cultural and policy changes are also needed to modify the widely accepted beliefs that favor screening at all times.

Correspondence: Alexia M. Torke, MD, MS, Indiana University Center for Aging Research, 410 W 10th St, Ste 2000, Indianapolis, IN 46202 (atorke@iupui.edu).

Accepted for Publication: October 23, 2012.

Published Online: March 11, 2013. doi:10.1001/jamainternmed.2013.2903

Author Contributions:Study concept and design: Torke, Schwartz, and Sachs. Acquisition of data: Torke, Holtz, and Montz. Analysis and interpretation of data: Torke, Schwartz, Holtz, Montz, and Sachs. Drafting of the manuscript: Torke, Holtz, and Montz. Critical revision of the manuscript for important intellectual content: Torke, Schwartz, Holtz, Montz, and Sachs. Statistical analysis: Torke. Obtained funding: Sachs. Administrative, technical, and material support: Schwartz, Holtz, Montz, and Sachs. Study supervision: Torke, Schwartz, and Sachs.

Funding/Support: This work was supported by award P30AG024967 from the National Institute on Aging to the Roybal Center at Indiana University. Dr Torke was supported by a career development award from the National Institute on Aging (K23AG031323).

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging or the National Institutes of Health.