aFactors are included only if they are reported in more than 1 study.
eTable 1. Search Strategy
eTable 2. Study Quality Appraisal Using Joanna Briggs Institute Critical Appraisal Checklists (Quantitative Studies)
eTable 3. Study Quality Appraisal Using Joanna Briggs Institute Critical Appraisal Checklists (Qualitative Studies)
eAppendix. Results of Risk of Bias Assessment
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Smith J, Dodd RH, Gainey KM, et al. Patient-Reported Factors Associated With Older Adults’ Cancer Screening Decision-making: A Systematic Review. JAMA Netw Open. 2021;4(11):e2133406. doi:10.1001/jamanetworkopen.2021.33406
What are the factors associated with older adults’ cancer screening decision-making?
This systematic review of 21 studies identified that positive screening attitudes and gaining knowledge and reassurance were associated with the decision to undergo screening, whereas negative screening attitudes and not wanting to know whether one has cancer were associated with the decision to forgo screening. Other notable factors with varying associations included poor health status and health problems, limited life expectancy, and a physician’s recommendation to stop cancer screening.
Communication strategies should support older adults to make informed cancer screening decisions by addressing underlying screening beliefs in context with their perceived and actual risk of developing cancer.
Decisions for older adults (aged ≥65 years) and their clinicians about whether to continue to screen for cancer are not easy. Many older adults who are frail or have limited life expectancy or comorbidities continue to be screened for cancer despite guidelines suggesting they should not; furthermore, many older adults have limited knowledge of the potential harms of continuing to be screened.
To summarize the patient-reported factors associated with older adults’ decisions regarding screening for breast, prostate, colorectal, and cervical cancer.
Studies were identified by searching databases from January 2000 to June 2020 and were independently assessed for inclusion by 2 authors. Data extraction and risk of bias assessment were independently conducted by 2 authors, and then all decisions were cross-checked and discussed where necessary. Data analysis was performed from September to December 2020.
The search yielded 2475 records, of which 21 unique studies were included. Nine studies were quantitative, 8 were qualitative, and 4 used mixed method designs. Of the 21 studies, 17 were conducted in the US, and 10 of 21 assessed breast cancer screening decisions only. Factors associated with decision-making were synthesized into 5 categories: demographic, health and clinical, psychological, physician, and social and system. Commonly identified factors associated with the decision to undergo screening included personal or family history of cancer, positive screening attitudes, routine or habit, to gain knowledge, friends, and a physician’s recommendation. Factors associated with the decision to forgo screening included being older, negative screening attitudes, and desire not to know about cancer. Some factors had varying associations, including insurance coverage, living in a nursing home, prior screening experience, health problems, limited life expectancy, perceived cancer risk, risks of screening, family, and a physician’s recommendation to stop.
Conclusions and Relevance
Although guidelines suggest incorporating life expectancy and health status to inform older adults’ cancer screening decisions, older adults’ ingrained beliefs about screening may run counter to these concepts. Communication strategies are needed that support older adults to make informed cancer screening decisions by addressing underlying screening beliefs in context with their perceived and actual risk of developing cancer.
Decisions for older adults (aged ≥65 years) and their clinicians about cancer screening are not easy. The evidence of benefits is unclear and the chance of harm becomes greater (eg, overdiagnosis, burdens of additional testing, false-positive results, and psychological impacts),1-3 which can adversely impact an older individual’s remaining quality of life.4-7 Although guidelines recommend when screening is no longer needed (eg, life expectancy <10 years for breast screening),8 many older people with limited life expectancy continue screening for colorectal, breast, cervical, and prostate cancer.9,10
There has recently been a shift in cancer screening communication to support informed decision-making rather than persuasively promote uptake.11 Some clinicians and researchers also advocate for individualized screening decisions for older adults (ie, not solely based on age).4,6 However, using these approaches in practice is challenging. Clinicians have varied knowledge about overscreening and find it difficult to discuss stopping screening.12 Furthermore, some older adults may not appreciate a recommendation to stop screening,13 and positive screening attitudes may be associated with continued screening.14
A greater understanding of the factors associated with cancer screening decision-making will provide helpful insight for clinicians navigating screening discussions with older patients and inform the design of interventions to support informed decision-making. The aim of this review is to examine the patient-reported factors associated with older adults’ breast, prostate, colorectal, and cervical cancer screening decisions.
This systematic review was registered as a protocol with PROSPERO (CRD42020185642). Reporting is guided by the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.15
A comprehensive search strategy was developed in consultation with an academic librarian (eTable 1 in the Supplement). We searched Medline, Pre-Medline, EMBASE, PsycINFO, and CINAHL on June 2, 2020. Duplicates were removed before 2 researchers (J.S. and R.H.D.) independently screened titles and abstracts for inclusion and exclusion. Using an eligibility checklist, decisions were made independently, and disagreements were resolved via discussion. Any uncertainty that arose was taken to the research team for further discussion. Once decisions were finalized, a backward citation search and forward citation search (on January 13, 2021) was conducted.
Studies were included if they empirically assessed patient-reported factors associated with older adults’ cancer screening decision-making (eg, patient-reported decision, choice, or intention) and included adults aged 65 years or older, a subgroup of participants aged 65 years or older or with limited life expectancy (<10 years), or participants with a mean or median age of 65 years or older (if there was also evidence of targeted recruitment of older people). Studies were included that assessed breast (mammography), prostate (prostate-specific antigen testing), colorectal (fecal occult blood testing or colonoscopy), or cervical (Papanicolaou test or human papillomavirus test) cancer screening decisions. Quantitative, qualitative, and mixed-methods peer-reviewed publications or theses were included.
Studies were excluded if they analyzed factors associated with retrospective screening or assessed decision-making regarding diagnostic or surveillance tests. To ensure that decision-making was examined in the general older population, studies that included participants at high risk or who received a diagnosis of the cancer type being examined (unless the proportion was inconsequential [<10% of the sample] or any impact on decision-making was accounted for) were excluded. Reviews, editorials, commentaries, research letters, intervention studies, or protocols were also excluded.
Risk of bias was assessed independently by 2 researchers (J.S. and K.M.G.). The Joanna Briggs Institute critical appraisal tools were used (Prevalence Studies Checklist for survey studies and Qualitative Research Checklist for interview and focus group studies).16 Items relating to different aspects of the methods of included studies were individually scored to determine an overall assessment of study quality (low, moderate, or high risk of bias). See eTable 2 and eTable 3 in the Supplement for more details.
A standardized template was developed and piloted to extract data, guided by previous systematic reviews.17-19 Data were independently extracted by 2 authors (J.S. and K.M.G.) and disagreements were discussed. Narrative synthesis was used to summarize the data. Data analysis was performed from September to December 2020.
We screened 2475 titles and abstracts after removal of duplicates and reviewed 183 full texts, of which 22 met inclusion criteria (Figure 1). Two articles reported data from the same study,14,20 leaving 21 unique studies.
Nine of the included studies were quantitative,14,21-28 8 were qualitative,13,29-35 and 4 used mixed methods.36-39 Ten studies examined breast cancer screening,21,22,24,25,29-32,36,39 3 assessed cervical cancer,26,28,34 1 assessed prostate cancer,23 and 7 assessed some combination of breast, prostate, cervical, and colorectal cancer screening.13,14,27,33,35,37,38 Thirteen studies included women only,21,22,24-26,28-32,34,36,39 1 included men only,23 and 7 included a combination of men and women.13,14,27,33,35,37,38 Most studies were conducted in the US (17 of 21 studies), 2 in the United Kingdom,21,36 1 in Germany,37 and 1 in Brazil.34 Five studies analyzed factors associated with screening intention21,22,24,26 or willingness to stop,28 1 assessed decisions within a discrete choice experiment for a hypothetical patient,14 and 1 examined factors associated with the decision process (eg, goals, concerns, and knowledge).27 Two studies asked closed questions about screening,23,25 9 asked open questions about screening29,31,32,34,39 or specifically about stopping,13,30,33,35 and 3 asked both open and closed questions.36-38 See Table 1 for a summary of the characteristics of the included studies.
Most studies included adults aged 65 years or older (17 of 21 studies),13,14,21,22,25,26,28-32,34-39 3 included a subgroup of adults aged 65 years or older,23,24,27 and 1 study reported a mean or median age 65 years or older and targeted recruitment via a senior health center.33 Most had representation from younger (65-74 years) and older (≥75 years) age ranges,13,14,21-23,25,26,28-30,32-39 except for 3 studies: 2 included only adults aged 75 years or older27 or 80 years or older,31 and 1 did not report an age range.24 Many studies also included participants with self-reported fair to poor health,13,23,28,30,31,39 2 or more health problems or comorbidities,25,32,36,38 mild cognitive impairment,33 life expectancy less than 10 years,14 or appreciable or severe physical disability.21 Two studies included mostly healthy participants,29,37 and 6 did not report health characteristics.22,24,26,27,34,35 Four studies included participants with a personal history of cancer (see Table 1 for reasons for inclusion).25,33,36,38
Risk of bias assessment is reported in the eAppendix in the Supplement and is incorporated throughout the results for studies that report conflicting or stand-alone findings. Main findings from the included studies are categorized and synthesized narratively below with additional details presented in Table 2 (quantitative results) and Table 3 (qualitative results). Figure 2 shows a summary of factors associated with older adults’ cancer screening decision-making.
Nine studies13,14,21,22,26,28,31,33,38 found that older age was associated with the decision to forgo screening (eg, “If I got to be really old, I think I would say to heck with it. Like in my nineties.”). In contrast, another study25 found age was associated with older women’s decisions to both forgo and undergo breast screening. Women also held negative views toward upper age limits for breast screening eligibility in 1 UK study with moderate risk of bias.36
Quantitative studies consistently found that those with higher education28 or social class or income21,28 were more likely to undergo screening. Three quantitative studies23,26,28 and 1 qualitative study33 found that not having insurance (US) was associated with forgoing screening. However, 2 studies found no association of income or insurance with breast screening decisions.22,24 Two studies also found no association with race or ethnicity.24,28 Being married was associated with a greater likelihood of undergoing breast screening in 2 studies,21,24 but another study at moderate risk of bias found that marital status was not associated with cervical screening decisions.28 One qualitative study suggested living in a nursing home was not associated with the decision to forgo screening,30 but 3 studies33,37,38 (2 of which had risk of bias concerns37,38) highlighted that some individuals may be open to stopping for this reason.
The association of overall health status with screening decisions varied across studies. Qualitative studies13,38 suggested that older adults may consider their overall health status or health problems in their cancer screening decision-making. However, quantitative evidence23,25,37,38 suggested that this was not the majority view, and 3 other studies28,30,36 found that health status was not associated with cervical or breast screening decisions. More specific problems considered as reason to forgo screening included memory problems, “dying from something else,” or not being functionally able to undergo screening.31,33,38 Qualitative studies13,31 suggested screening was instead seen as a way to evaluate or validate health status or avoid major surgery or longer hospitalization as a result of early detection.36 Individual studies found that no physical disability (compared with severe disability),21 not having depression or anxiety,21 and greater quality of life increased the likelihood of choosing screening.14
The association of life expectancy with screening decisions was also inconsistent across studies. In a discrete choice experiment,14 participants were more likely to choose screening for a hypothetical patient with 5 or 10 years life expectancy than a patient with 1 year life expectancy (but 57.2% still chose screening for the latter), and a qualitative study33 found that some individuals may consider limited life expectancy as a reason to stop screening. However, 6 qualitative studies13,30,33,35,37,38 highlighted older adults’ resistance to using life expectancy to dictate or communicate about screening decisions. Instead, some thought that screening was a means of increasing life expectancy.33,36,37
Personal or family history of cancer consistently was associated with the decision to undergo screening in 5 studies.25,28,31,33,35 However, another study22 found that the majority of women thought they should continue breast screening even if they did not have any risk factors. Prior screening experience was associated with continued screening,14,22,24,26 but this was also reason to forgo breast screening.25,31 Screening to target treatment appropriately was expressed in 1 qualitative study (with moderate to high risk of bias),37 and women were also hesitant to stop breast screening even if they did not plan to undergo treatment.30 Women who had not had a hysterectomy were more likely to undergo cervical screening,26 and the number of visits to clinicians and exercise were not associated with breast cancer screening decisions.24
Positive screening attitudes consistently were associated with the decision to undergo screening across many studies, including the importance of early detection and testing,23,26,33,37 belief in screening,35,38,39 just wanting to do it,29 fulfilling a duty,37 a way of taking care of oneself,31 and peace of mind.27 In 3 studies,14,37,38 many older adults agreed that they planned to get breast, prostate, and colorectal cancer screening for as long as they live. Similarly, negative screening attitudes consistently were associated with the decision to forgo screening across many studies, including simply not wanting screening,26,36 lack of benefit or appropriateness,33,37 disinterest,37 test discomfort or pain,26,29,34-36 embarrassment or privacy concerns,36 fatalism,32,33,35 the potential trouble obtaining insurance coverage if the individual is identified as high risk,26 and stress, time, or cost.35 Routine and habit also were associated with the decision to undergo screening,23,25,30,37 and forgetting was associated with forgoing screening.29,36
Perceived risk and fear of developing cancer were inconsistently operationalized and assessed across studies, meaning that findings were varied regarding their association with decision-making. High perceived risk was associated with continued screening,22,24,26 and low perceived risk was associated with forgoing screening,25,31 except in 1 study.28 Cancer fear, anxiety, and concern were associated with both the decision to undergo22,33 and forgo screening,32 except in 1 study.26 Fear of screening itself reduced the desire to screen,32-34,37 and fear of not being screened was associated with continued screening.37 Gaining knowledge or reassurance was associated with the decision to undergo screening,23,25,27,33,36,37,39 but not wanting to know also was associated with the decision to forgo screening.31,32,34,39
Risks of screening were also inconsistently assessed, and results varied across the included studies regarding their influence. Qualitative findings suggested that screening risks may be associated with the decision to forgo screening33,38 as well as knowledge of overdetection in 1 quantitative study.39 However, 3 studies found low awareness of potential risks of screening,35 low worry about risks of breast screening (low to moderate risk of bias),36 and resistance to the concept of overdetection.39
The association of screening knowledge or information with decision-making was extremely varied across included studies. Awareness of US Medicare coverage22 and the starting age for screening24 was associated with continued breast screening, and not knowing about access beyond age 70 years was a reason for not screening (low to moderate risk of bias).36 Receiving information was generally expressed by older adults as influential,23 and many specifically desired information about the possibility of false-positive findings.23,37 Older adults desired information for agency in the decision,35,36,39 but looking for cancer information was not associated with breast screening decisions,24 and statistical information was not deemed relevant.33 However, some older adults expressed a desire for specific, tailored information to justify a recommendation to stop screening.35
Seven studies23-26,29,37,39 consistently found that a physician’s recommendation to screen was associated with the decision to undergo screening. Seven other studies13,14,25,29,34,35,38 also found that a physician’s recommendation against screening was associated with the decision to forgo screening. In 1 study,27 discussing the disadvantages of screening was associated with stopping colorectal screening. Not receiving a recommendation was also reason to forgo breast screening for some individuals.31 However, qualitative studies highlighted more nuance surrounding the influence of a physician’s recommendation against screening. Some older adults reported they may respond negatively to a recommendation against screening; they were skeptical, especially when justified with life expectancy,13,30 desired a second opinion,33,35 or continued screening anyway; 1 study38 found that 43% of participants still desired screening despite receiving a recommendation to stop.
Believing that the physician makes important decisions and trusting them to make the best decision on their patients’ behalf were not associated with cervical screening decisions.24 One study38 also found that one-half of the participants thought that physicians know for certain whether cancer screening helps people older than 70 years (low to moderate risk of bias).
Of the studies that examined the influence of family, most (6 studies) found that family was associated with the decision to undergo screening (including wives, daughters, mothers, and mothers-in-law, specifically for breast screening),23,25,29-31,35 but 3 studies32,33,36 found that family was associated with the decision to forgo screening because of older adults feeling they would become a burden33,36 or that they needed to care for family instead of themselves.32 One study26 specifically found that the perceived positive impact of screening for the family was not associated with the intent to undergo cervical screening. Five studies consistently reported the association of friends with older adults’ decisions to undergo breast and prostate screening.23,25,29,31,35
Invitations and reminders consistently were associated with the decision to undergo breast screening (in the United Kingdom and US),25,31,36 but skepticism toward government recommendations to discontinue screening was also expressed in 2 studies.30,33 However, it was unclear how government recommendations were associated with decision-making. The association of media with undergoing breast screening was also expressed in 2 studies.25,29 Single studies found that access to transportation was associated with prostate cancer screening decisions,23 transport difficulties specifically discouraged breast cancer screening attendance,36 and ease of access was associated with women’s decisions to undergo breast cancer screening.31 Finally, societal costs and resource waste were reasons not to screen in 2 studies,33,38 but this may be the minority view because only 30% of participants agreed with the statement that “Screening for cancer in people over the age of 70 may waste healthcare time and money.”38
This systematic review synthesizes study findings regarding the factors associated with older adults’ cancer screening decisions. It builds on previous work in younger populations by including study designs that summarize how commonly studied factors (eg, personal or family history of cancer), psychosocial factors, and factors of more relevance to older people (eg, life expectancy and health status) objectively and subjectively are associated with older adults’ cancer screening decisions. The findings reveal a substantial gap in studies conducted outside the US, which is important given that many health systems vary considerably from the US health system.
Our findings highlight the degree to which aging-related factors associated with the likelihood of benefit and harms of cancer screening are associated with older adults’ decision-making. First, despite older age consistently being associated with older adults forgoing screening, recent US evidence highlights high rates of screening beyond recommended ages.10 Qualitative studies suggest that some older adults find the idea of no longer needing screening when much older more intuitive than stopping at a younger or older age (eg, “If I’m 90” or “dying from something else” vs 75 years). This may be because of widespread information that age is the greatest nonreversible factor associated with the risk of cancer. Second, although limited life expectancy is also associated with lower screening participation,40-44 rates of screening in older adults with limited life expectancy are still high.9,44-46 Findings from this review suggest that life expectancy may not be relevant to some older adults’ screening decisions. A recent survey of veterans (aged ≥50 years) supports this finding, as stopping screening for colorectal cancer because of age and life expectancy did not resonate.47 Finally, some older people may not consider other health problems as relevant to their screening decisions and, instead, could view screening as a means of validating their health. Relevant studies not included in this review similarly suggest that older people may see discontinuing screening as giving up48 and that being invited to participate in cancer screening symbolically demonstrates continued investment in their health.49 Although aging-related concepts are challenging to communicate, older people must be counseled about the reduced benefit and increased chance of harm from screening associated with limited life expectancy and worsening health to make better quality screening decisions.
Many psychosocial factors were consistently associated with older adults’ screening decisions across studies. Many of these factors should not be the main reason for screening decisions (eg, routine or habit and reassurance) and are likely substantial hindrances for older adults in considering the factors that should be considered to make better quality decisions (ie, life expectancy and overall health). One study14 specifically found attitudes toward screening are better at estimating older adults’ screening decisions than age and life expectancy, highlighting that an individual’s intrinsic attitude to screening needs to be understood when trying to foster informed decision-making. Given that aging-related factors are commonly incorporated into screening guidelines, future research should examine how physicians can sensitively communicate with patients regarding these factors in light of their screening attitudes. Our review suggests that some patients may appreciate a recommendation to stop cancer screening, especially if it is individually tailored and justified, but others may not. Although older adults may prefer a recommendation statement that incorporates health status,50 our Australian experimental study51 found that a more-confronting statement incorporating life expectancy could be more effective than a more-preferred statement incorporating health status in reducing screening intention without having adverse outcomes. More research with larger community samples is needed to examine how older adults respond to recommendations against screening and information about potential harms. Responses may differ according to different patient characteristics (eg, screening attitudes, knowledge, patient-physician relationship, and perceived risk) and the way it is communicated. Some older adults may also respond negatively to government recommendations against screening, perhaps because of the belief that it would be a cost-saving measure.52 This warrants further exploration across countries where differences in reimbursement for screening may differentially affect decision-making.
Our review was limited as we chose not to perform a meta-analysis because of the wide variability in designs and outcomes and explanatory variables in the included studies. Instead, we used narrative synthesis to present findings given that estimating the relative magnitude of effect sizes for these factors was not the aim. However, this review builds on existing narrative reviews by using a systematic approach and including a range of study designs with results presented narratively, meaning the nuance of the data could still be captured, as well as the consistencies and inconsistences across studies. Two independent reviewers conducted screening, data extraction, and risk of bias assessment, and gray literature (eg, theses) and non-English studies were searched and included. Another limitation was that the larger scope meant that studies examining combinations of breast, prostate, colorectal, or cervical cancer screening decisions were included, making it difficult to identify trends or differences across separate screening decisions. However, it was important to be more inclusive given the similarly high rates of screening evident across all 4 screening types, especially in the US,9 and the limited recent research in this field. Only 1 study exclusively included men, 7 studies that assessed multiple screening types included mostly women, and only 1 study analyzed sex-specific differences in decision-making (and found no differences). Therefore, future studies should examine sex- and screening-specific differences in decision-making. Large, nationally representative, cross-sectional surveys in different countries would also be useful to guide the targeting of communication strategies to support informed decision-making. More specifically, associations between demographic and health characteristics, screening attitudes and knowledge, and the extent to which factors such as health status, life expectancy, and personal risk are considered in older adults’ decision-making should be examined. An international comparison should also be conducted to explore differences between countries with and without nationally funded cancer screening programs (eg, US vs Australia).
Older adults should consider aging-related factors such as life expectancy and overall health status in their cancer screening decisions, as these factors have important implications for their own personal chance of benefiting from screening and chance of adverse outcomes on health and quality of life due to short-term physical and psychological harm. Currently, their existing screening beliefs may run counter to these concepts, likely hindering efforts to support informed decision-making, and decision support or education about benefits and harms may not be enough to change screening intentions or behavior.48,53 Communication strategies ought to address these underlying screening perceptions and beliefs given older adults’ longer term exposure to positive screening messaging and the adverse impact cancer screening can have on their quality of remaining life.
Accepted for Publication: September 12, 2021.
Published: November 8, 2021. doi:10.1001/jamanetworkopen.2021.33406
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Smith J et al. JAMA Network Open.
Corresponding Author: Jenna Smith, BSc (Hons), Wiser Healthcare, School of Public Health, Faculty of Medicine and Health, The University of Sydney, Edward Ford Bldg, Room 128C, Sydney, NSW 2006, Australia (email@example.com).
Author Contributions: Ms Smith 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.
Concept and design: Smith, Dodd, Naganathan, Cvejic, Jansen.
Acquisition, analysis, or interpretation of data: Smith, Dodd, Gainey, Jansen, McCaffery.
Drafting of the manuscript: Smith, Naganathan.
Critical revision of the manuscript for important intellectual content: All authors.
Administrative, technical, or material support: Dodd.
Supervision: Dodd, Naganathan, Cvejic, Jansen, McCaffery.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by program grant 1113532 from Wiser Healthcare, funded by the National Health and Medical Research Council.
Role of the Funder/Sponsor: The funder 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 Contributions: Bernadette Carr, BA, Grad Dip (Public Health Librarian, The University of Sydney), assisted with the search strategy development. Ms Carr received no financial compensation for her contribution.