Appendix. Interview Guide and Participant Questionnaire
Customize your JAMA Network experience by selecting one or more topics from the list below.
Schoenborn NL, Lee K, Pollack CE, et al. Older Adults’ Views and Communication Preferences About Cancer Screening Cessation. JAMA Intern Med. 2017;177(8):1121–1128. doi:10.1001/jamainternmed.2017.1778
How do older adults think about stopping cancer screening when life expectancy is limited, and how do they prefer to discuss it with clinicians?
In this qualitative interview study with 40 community-dwelling older adults, participants were amenable to stopping cancer screening in the context of a trusting relationship with their clinician. Participants did not often consider life expectancy important in screening or prefer to hear about life expectancy when discussing screening.
Better delineating patient-centered approaches to discuss screening cessation when life expectancy is limited is important for optimizing cancer screening in older adults.
Older adults with limited life expectancy are frequently screened for cancer even though it exposes them to risks of screening with minimal benefit. Patient preferences may be an important contributor to continued screening.
To examine older adults’ views on the decision to stop cancer screening when life expectancy is limited and to identify older adults’ preferences for how clinicians should communicate recommendations to cease cancer screening.
Design, Setting, and Participants
In this semistructured interview study, we interviewed 40 community-dwelling older adults (≥ 65 years) recruited at 4 clinical programs affiliated with an urban academic medical center.
Main Outcomes and Measure
We transcribed the audio recorded discussions and analyzed the transcripts using standard techniques of qualitative content analysis to identify major themes and subthemes.
The participants’ average age was 75.7 years. Twenty-three participants (57.5%) were female; 25 (62.5%) were white. Estimated life expectancy was less than 10 years for 19 participants (47.5%). We identified 3 key themes. First, participants were amenable to stopping cancer screening, especially in the context of a trusting relationship with their clinician. Second, although many participants supported using age and health status to individualize the screening decision, they did not often understand the role of life expectancy. All except 2 participants objected to a Choosing Wisely statement about not recommending cancer screening in those with limited life expectancy, often believing that clinicians cannot accurately predict life expectancy. Third, participants preferred that clinicians explain a recommendation to stop screening by incorporating individual health status but were divided on whether life expectancy should be mentioned. Specific wording of life expectancy was important; many felt the language of “you may not live long enough to benefit from this test” was unnecessarily harsh compared with the more positive messaging of “this test would not help you live longer.”
Conclusions and Relevance
Although research and clinical practice guidelines recommend using life expectancy to inform cancer screening, older adults may not consider life expectancy important in screening and may not prefer to hear about life expectancy when discussing screening. The described communication preferences can help inform future screening discussions. Better delineating patient-centered approaches to discuss screening cessation is an important step toward optimizing cancer screening in older adults.
Cancer screening in older adults with limited life expectancy may inappropriately subject them to short-term harms when chance of benefit is minimal.1-12 Research and clinical practice guidelines recommend incorporating life expectancy to inform cancer screening.1-12 Specifically, several Choosing Wisely recommendations mention that clinicians should not routinely screen for cancer among patients with limited life expectancy.13-16 However, many older adults with limited life expectancy still frequently receive cancer screening,17-22 with screening rate as high as 55% in one study.17
Only a few small studies have explored older adults’ views on stopping cancer screening. The results suggest that some older adults may be strongly enthusiastic about cancer screening23-27 and may not always consider life expectancy important in the screening decision.25,26 In one survey study25 of 116 mostly healthy older adults, 62% did not think their physician’s life expectancy estimate was important in cancer screening decisions. In an interview study26 of 28 older adults, most of whom were 75 years or younger, 52% did not think their health would influence their screening decision even if they had limited life expectancy. Previous studies have not explored the views of older adults with a wide range of age, health status, and life expectancies. Studies also have not explored patient preferences for how clinicians should discuss screening cessation.
Clinicians’ recommendations are among the most influential factors in cancer screening decisions,18,28 and clinicians are frequently uncomfortable stopping cancer screening.23,29,30 One study30 showed that primary care clinicians often considered that their patients had limited life expectancy but were hesitant to recommend against cancer screening. Clinicians were concerned that patients may react negatively to screening cessation and also struggled with discussing life expectancy.30
In this context, patient perspectives and preferences about stopping cancer screening among those with limited life expectancy are critical to address clinician concerns, inform screening discussions, and help align clinical guidelines with patient preference. Because relatively little is known about this area, we use qualitative methods in this study to explore the range of perspectives from older adults to generate hypotheses. We aim to examine community-dwelling older adults’ perspectives on the decision to stop cancer screening when life expectancy is limited and to identify their preferences for how clinicians should communicate recommendations to cease cancer screening.
This was a qualitative study in which semistructured interviews lasting 30 to 60 minutes were conducted with community-dwelling older adults (≥65 years). Participants were recruited from 4 clinical programs affiliated with an urban academic medical center, including a geriatric ambulatory clinic, a general internal medicine ambulatory clinic, a house-call program for homebound older adults, and a Program of All-Inclusive Care for the Elderly (PACE). This project was approved by a Johns Hopkins School of Medicine institutional review board.
Older adults were eligible to participate if they were 65 years or older, English-speaking, and able to provide informed consent. Because cognitive impairment affects the risk-benefit ratio of cancer screening,31 we sought to include those with cognitive impairment as long as the person could still participate meaningfully in the interview and provide written informed consent. For potential participants who had a diagnosis of cognitive impairment or dementia and were interested in the study, we consulted with the person’s family members and/or clinician to ensure that the person could provide written informed consent and could participate in a 30- to 60-minute interview.
We used maximum variation sampling (targeting a wide range of diverse participants) to recruit community-dwelling older adults with a wide range of age, health and functional status, and life expectancies.32,33 We purposefully recruited from 4 different clinical programs to include older adults in ambulatory clinics and those who were home-bound (house-call) and/or had limitations in daily activities (PACE). We also asked clinicians from all 4 clinical programs to help identify potential participants. Recruitment stopped when theme saturation in the data was reached,32 as described in subsequent section. Each participant was provided a $50 gift card, and parking voucher when needed.
The interview guide (in the Supplement) was developed and iteratively revised during pretesting with 10 older adults who were not included in the study. At the beginning of the interview, we provided a brief overview of the benefits and potential harms of cancer screening, using breast and/or prostate and colorectal cancer screenings as examples. We chose these examples because they are routinely recommended in older adults, and overscreening has been shown to occur.3-6,17-22 We explicitly mentioned that it may take up to 10 years before a cancer grows to the point of causing health problems, so that someone who will not live 10 years may not benefit and may be harmed from screening.2,6 We then asked questions about 2 domains: decision-making and communication.
In the decision-making domain, we explored the following: (1) considerations around the decision to stop screening, (2) reactions if their clinician were to recommend that they stop screening, (3) perceived importance of health and functional status (important predictors of life expectancy) in screening decisions,34,35 and (4) views on hypothetical examples about screening when age and health status were discordant.
In the communication domain, we tested 4 communication strategies for clinicians to discuss screening cessation to a hypothetical patient (Box 1). We then explored attitudes on discussing life expectancy in the cancer screening context. We asked participants for their reaction to one of the Choosing Wisely statements that address cancer screening and limited life expectancy.13-16 We presented the statement from the Society of General Internal Medicine owing to its brevity: “Don’t recommend cancer screening if patient is not likely to live 10 years.”16 The interviews were semistructured and allowed for new topics to emerge.
Mentioning risks and benefits only: “This test would harm you more than benefit you.”
Mentioning risks and benefits plus age: “Around your age, this test is usually not recommended because it would cause more harm than benefit.”
Mentioning risks and benefits plus health status: “When people have these medical conditions like you and need help from day to day activities like you, this test can cause more harm than benefit.”
Mentioning risks and benefits plus life expectancy: “This test would not help you live longer; it would harm you more than benefit you.”
We then asked about a second way to phrase life expectancy: Instead of saying: “this test would not help you live longer,” what about saying: “you may not live long enough to benefit from this test”?
A structured questionnaire prior to the interview collected information on demographic characteristics, health, and functional status as part of a mortality risk index,34 health literacy,36 numeracy,37 and trust in the clinician (“All in all, you have complete trust in your doctor”; 1 = strongly disagree, 5 = strongly agree).38 We also asked about past screenings for breast (women only), prostate (men only), colorectal cancers, and intention to continue or discontinue screenings at the time of the study. We extracted information on health conditions and duration of relationship with clinician from the medical records. Using the collected data, we estimated 4-year and 10-year mortality risks using a validated index.34
One investigator (N.L.S.) conducted the interviews in person between December 2015 and March 2016. The investigator had no direct clinical relationship with any participant. For participants in the 2 ambulatory clinics, interviews occurred in a private room before or after a scheduled clinic visit or (for 2 participants) at the participants’ homes. For participants in the house-call program, interviews occurred at the participants’ homes at times chosen by the participants. For participants in the PACE program, interviews occurred in a private conference room at the PACE program site at times chosen by the participants. All interviews were audio recorded. The audio recordings were then transcribed verbatim and analyzed using Atlas.ti textual data analysis software. The transcripts were continuously reviewed and assessed for the emergence of new ideas or themes; data collection continued until no new ideas were emerging and theme saturation was reached.32 Standard techniques of directed qualitative content analysis were used to code the transcripts.32,39 A preliminary coding scheme based on the interview guide was iteratively refined and applied to analyze the data using the constant comparative approach.32,40 Open coding procedures allowed new theme identification in addition to the established scheme. Revisions to the coding scheme were applied to all previously coded transcripts. Two investigators (N.L.S. and K.L.) independently coded all transcripts. Differences were reconciled by consensus until 100% agreement was reached. Content analysis generated major themes and subthemes.
Forty older adults participated in the study (Table 1). The participants’ average age was 75.7 years. Of the 40 participants, 23 (57.5%) were female; 25 (62.5%) were white. Estimated life expectancy was less than 10 years for 19 participants (47.5%), including 8 participants with less than 4-year life expectancies.34 All participants had regular clinicians and reported high level of trust (average score, 4.7 out of 5.0). Duration of relationship with clinician averaged 3.7 years. Among the 29 participants who had up-to-date cancer screenings, 5 participants (all female) reported that they had decided to stop screening and 4 participants (2 males, 2 females) reported that they were unsure about continued screening.
Qualitative content analysis revealed 3 major themes with subthemes (Box 2); these are presented herein and illustrated using representative quotes.
Theme 1: Older adults were amenable to considering cessation of cancer screening.
Participants mentioned several reasons to stop screening.
Cessation of screening was acceptable within trusting relationship.
Theme 2: Older adults supported using health status to individualize screening decision but often did not understand the role of life expectancy.
Health status was considered important.
Some used health status in different ways to inform screening decisions.
Many did not understand the role of life expectancy in cancer screening.
Theme 3: Preferences varied regarding how to discuss cessation of screening.
Incorporating health status was the preferred communication strategy.
Views were divided on whether life expectancy should be discussed.
Specific wording of life expectancy was important.
Some preferred little to no explanation.
Older adults were amenable to considering cessation of cancer screening. In addition to the participants who already decided to stop screening, others mentioned hypothetical scenarios in which they would consider screening cessation (Table 2). Older age was the most common reason; one 84-year-old woman said: “I just feel like at my age I don’t need a colonoscopy, what’s gonna be is gonna be.”
The mentioned age threshold to stop screening varied in age from 65 to 100 years.
When the participants were asked how they would react if their regular clinician suggested that they stop screening, many said that they would view such a suggestion as acceptable or positive. One participant said: “I’d feel good that I didn’t need [another screening].”
Some commented that they would think more highly of their clinician: One participant said, “If the doctor says to me we don’t have to do this no more I’d say: ‘thank you very much doc’… I probably [would] think more of him.” Another participant said, “I would think he would be personalizing my personal health when he suggested that.”
Many of the participants attributed their responses to their trust and confidence in their clinician: “I have all the confidence in her and if she told me to stop it I would stop.”
Some participants were skeptical of a suggestion to stop screening or said that they would still insist on screening. Even among these, participants mentioned that the clinician’s suggestion would not necessarily make them think less of the clinician or trust the clinician less. One person described that she eventually accepted her clinician’s recommendation to cease screening because of trust: “I told [my doctor] that I would want another [mammogram]. He said: ‘But at your age, 75, [we] don’t usually give another test.’ I said: ‘What difference does it make [what] my age is… I’m still human you know, why not another test?’ But he said: ‘you are fine, you ain’t got nothing to worry about.’ I said: ‘Well, I must be fine.’... He’s a good doctor and I trust him.”
Older adults supported using health status to individualize screening decision but often did not understand the role of life expectancy. When asked about the importance of health and functional status in cancer screening decisions, most participants believed that these were important factors to consider along with age: “There are people much younger whose health is very poor, and people who are 80 and 90 whose health is very good, so age is not the only determining factor [in cancer screening] in my opinion.”
On the one hand, when provided the example of a healthy older person who would have been recommended to stop screening based on age, many participants supported continuation of screening. On the other hand, when provided the example of a sick younger person who would have been recommended to receive screening based on age, many agreed that foregoing screening made sense: “Don’t do it. [Cancer] isn’t gonna be the thing that kills the people if they’ve got all those [health] problems.”
For some participants, however, using health status to inform screening decision led to opposite conclusions. Despite explaining that screening by definition looks for cancer in the absence of symptoms, some participants viewed screening as a way to evaluate poor health: “If a person is sick all the time any test they do has got to help, it can’t hurt … if they were really sick they’d probably need more tests.”
Although many participants supported using health status, along with age, to inform the screening decision, they did not perceive life expectancy as being directly related to health status and age and were perplexed when shown the Choosing Wisely statement “Don’t recommend cancer screening if patient is not likely to live 10 years.”16 All except 2 participants objected or questioned the statement. Reasons for objection included skepticism about life expectancy predictions, skepticism about screening’s lag-time to benefit, and perceived negativity of the statement (Table 3). One person described her doubt about the life expectancy prediction even if someone had multiple health problems: “How do you actually know the patient is not gonna live 10 years? I mean you look at it statistically I guess, you look and say…because she has this [health problem] and this [health problem], she probably won’t live, but you never know. There’s always that one person who’s able to get over the hump.”
Preferences varied regarding how to discuss cessation of screening. Although all 4 tested communication strategies were acceptable to most participants (Box 1), the one that mentioned health and functional status tended to be the preferred approach. One participant commented that it sounded better because it included details about him as an individual: “It sounds like my doctor has done a study on me and my issues, and she recommends that I not have this [test] done… I would have to be guided by her knowledge and her decision.”
Some participants mentioned that they would want to discuss what health issues, instead of cancer screening, would be the focus of their care. Others mentioned the importance of giving patient the ultimate choice to make the decision.
Participants were divided about whether clinicians should mention life expectancy in the context of discussing cancer screening if the clinician’s recommendation was informed by the patient’s life expectancy. The most common reason for wanting to discuss life expectancy was full disclosure of information: “I would want to know. [There are] a lot of things in life that is upsetting but you deal with it. It’s better, I think it’s better, to know than to not know.” For another participant, the phrase that “the test is not going to help you live longer” was the most compelling reason to stop screening: “That would really make me think and probably change my mind [about continuing screening].”
Others did not want to hear about life expectancy in the context of cancer screening: “Even though the doctor may have the feeling that the patient is not going to live very long I don’t think he should express it… No, just say the test is not going to be helpful.”
A common reason was that the participants did not believe that clinicians can predict life expectancy: “A doctor cannot tell you how long you gonna live, he can’t tell you if you gonna die tomorrow, next week, or 10 years from now…the only one [who] knows about that is God.”
A second reason that participants did not think life expectancy should be discussed in cancer screening was that the information would depress or worry the patient: “I would rather for them to focus on the test because to tell me about life expectancy then my mind is gonna be: ‘oh, I’m not gonna live long.’ You could shut a person down quicker.”
We provided 2 example phrases that mentioned life expectancy to explain screening cessation: “This test would not help you live longer” and “You may not live long enough to benefit from this test.” Most participants preferred the former phrase and called the latter phrase “harsh.” One person contrasted the 2 phrases: “You say you don’t think I’m gonna live that long. That raises a red flag to me…It’s more negative than the previous way you discussed it….By saying ‘it’s not gonna help me live longer’—it’s a positive statement.”
Although many participants appreciated the discussion on the risks and benefits of screening, some participants preferred little to no explanation: “A short ‘yes’ or ‘no’ would be ok if [the doctor] feels truly in his heart that is the way it should be.”
Other participants preferred omission of cancer screening discussion altogether if the doctor intended to recommend no screening: “If [the doctor] wasn’t gonna give [the screening test] to them he shouldn’t have mentioned it… Leave it alone.”
Building on the small body of literature that explored older adults’ attitudes about cancer screening and life expectancy,25,26 we characterized perspectives from older adults with a wide range of life expectancies regarding cancer screening cessation when life expectancy is limited. To our knowledge, this is the first project to describe older adults’ preferences for different strategies to discuss screening cessation.
Contrary to previous studies that suggested that patients may be reluctant to stop screening,23-27 we found that many participants in our study were amenable to considering screening cessation. Recommendations from a trusted clinician made screening cessation acceptable or positive to our study participants, whereas most participants in prior studies stated that they had never discussed stopping cancer screening with their clinicians.24,25 The results also contrast with a study that suggested a recommendation to stop screening may undermine trust in the clinician.24 The difference in results may stem from the fact that our study participants had high levels of trust and long relationships with their clinicians. Our results need to be tested in larger populations but suggest that it is possible for clinicians to find acceptable ways to recommend screening cessation without negatively affecting their on-going relationship with the patient.
We found that many participants agreed that age and health status are important in screening decisions. While these factors are key predictors of life expectancy,34,35 our participants did not perceive life expectancy as being directly related to them or that life expectancy itself was important in cancer screening. The perceived irrelevance of life expectancy was also found in 2 previous studies.25,26 It is not surprising then that many participants objected to the Choosing Wisely statement16 about not recommending cancer screening in those with limited life expectancy. We realize that this statement is intended for clinicians and may have been worded differently if directly addressing patients. This result demonstrates that complex concepts and terms used in research and clinical guidelines, such as life expectancy, can be misinterpreted by the public and call for careful, sensitive discussions to avoid misperceptions. We found similar views among participants with varying predicted life expectancies. An important next step is to examine the potential heterogeneity in views by life expectancy in a larger study.
Using health status to inform screening led to opposite decisions in some participants. These participants perceived someone with poor health status to need more screening. This misunderstanding of the purpose of screening, even after providing a brief explanation, confirms the importance to both educate patients and check for understanding in clinical practice.
How to best discuss screening cessation has not been previously studied. A systematic review of the impact of clinician-patient communication on cancer screening showed that whether or not a clinician recommended screening only explained some of the screening behavior; the quality and content of the communication were also influential.28 Currently, clinicians use a variety of approaches to discuss cessation of screening without clear best practices.41 Our results add to the literature by showing that many participants preferred explanations that included individualized health status but disliked explanations that mentioned life expectancy with a negative framing (“you may not live long enough to benefit from this test”).
Our findings highlight a potential dilemma around incorporating life expectancy in cancer screening. Although many participants were receptive to stopping cancer screening, most did not feel that life expectancy was the key decision-making factor, were skeptical that clinicians could accurately assess life expectancy, and did not want the clinician to discuss life expectancy in the screening context. An important implication of this finding is that clinicians may be caught between, on the one hand, patients who do not believe or want to hear about life expectancy, and, on the other hand, research and guidelines that emphasize the importance of using life expectancy to inform cancer screening.3-6,13-16 The scientific literature increasingly recognizes the importance of incorporating patient preference into clinical guidelines, including in the language of guideline presentation.42,43 Our results need to be tested in larger populations but suggest that misperceptions around the term life expectancy may be a barrier to guideline acceptance; a different term to represent the predicted health trajectory based on age and health status may be more preferred by patients. Even if the role of life expectancy in cancer screening is appropriately understood, the participants’ questions about the accuracy of life expectancy estimates may remain, since inherent uncertainty exists in applying population-based prediction models to any individual. A threshold of 10-year life expectancy may not sufficiently account for this uncertainty. Patient input is important to help better define how to apply life expectancy predictions to guide cancer screening decisions. These implications are also relevant to a number of other clinical decisions that incorporate life expectancy.44,45
This study has several limitations. It was conducted with participants from clinical programs affiliated with 1 academic center and all participants had regular clinicians in whom they had high levels of trust. The results may not represent experiences of older adults elsewhere or among those without regular clinicians. This study was not designed to be representative of all older adults, but rather to gain in-depth perspectives about a topic about which little was previously known. There was only 1 interviewer, and both transcript coders were physicians which may have had an impact on data collection and interpretation. The study design relied on self-report, and the results are prone to recall and social desirability biases. A number of the interview questions used hypothetical scenarios, and participants’ responses may not be fully congruent with behavior in real life.
We found that older adults were generally receptive to recommendations to stop cancer screening, especially in the context of a trusting physician relationship and when the discussions framed the recommendations around age, health status, and helping people live longer; in contrast, discussing life expectancy in the screening context was more controversial. Testing these communication preferences in larger populations to better delineate patient-centered approaches to discuss screening cessation is an important step toward optimizing cancer screening in older adults.
Corresponding Author: Nancy L. Schoenborn, MD, The Johns Hopkins University School of Medicine, 5200 Eastern Ave, Mason F. Lord Building Center Tower, Room 711, Baltimore, MD 21224 (email@example.com).
Accepted for Publication: February 22, 2017.
Published Online: June 12, 2017. doi:10.1001/jamainternmed.2017.1778
Author Contributions: Dr Schoenborn had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Schoenborn, Armacost, Dy, Bridges, Boyd.
Acquisition, analysis, or interpretation of data: Schoenborn, Lee, Pollack, Armacost, Bridges, Xue, Wolff, Boyd.
Drafting of the manuscript: Schoenborn, Armacost.
Critical revision of the manuscript for important intellectual content: Schoenborn, Lee, Pollack, Dy, Bridges, Xue, Wolff, Boyd.
Statistical analysis: Schoenborn, Xue.
Obtained funding: Schoenborn.
Administrative, technical, or material support: Lee, Armacost, Bridges.
Study supervision: Schoenborn, Pollack, Dy, Bridges, Boyd.
Conflict of Interest Disclosures: None reported.
Funding/Support: Research reported in this publication was supported by the National Institute On Aging of the National Institutes of Health under Award No. R03AG050912. This project was also made possible in part through the support of the Maryland Cigarette Restitution Fund Research Grant to the Johns Hopkins Medical Institutions. In addition, Dr Schoenborn was supported by a T. Franklin Williams Scholarship Award; Funding provided by Atlantic Philanthropies Inc, the John A. Hartford Foundation, the Alliance for Academic Internal Medicine–Association of Specialty Professors, and the American Geriatrics Society.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Information: Ms Armacost is also a patient and caregiver partner.
Previous Presentations: An earlier version of this study was presented as a poster at the American Society of Clinical Oncology National Meeting; June 3-7, 2016; Chicago, Illinois; and as a poster at the Grants for Early Medical/Surgical Specialists' Transition to Aging Research (GEMSSTAR) U13 Conference; September 21-23, 2016; Bethesda, Maryland.