SEARCH

SEARCH BY CITATION

Keywords:

  • cancer screening;
  • aging;
  • research agenda;
  • barriers

Abstract

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Cancer is the second most common cause of death among those aged ≥65 years, and is a major cause of morbidity. There is some evidence that screening, by detecting precancerous lesions in asymptomatic patients, is effective in reducing cancer-related morbidity and mortality among older men and women. The objectives of the current review article were to identify some of the barriers to and opportunities for research in cancer screening among older individuals. Using expert opinion, the authors developed a taxonomy of barriers to research among those aged ≥65 years at 3 levels: the macro (policy and population), the organizational and provider, and the interindividual and intraindividual. There are numerous barriers to high-quality screening research among older individuals, across all 3 levels of the taxonomy. Overall, there are limited evaluations of repeat adherence; follow-up for positive findings; screening, diagnostic, and treatment delays; and access to and acceptance of screening among those aged ≥65 years. There are particular barriers to research in colorectal cancer screening. There has been limited development and testing of evidence-based and theory-based intervention approaches to enrich screening adherence over time by those aged ≥65 years in which screening has demonstrated effectiveness. Professional groups differ in their recommendations for screening older asymptomatic patients, and implementation varies across healthcare systems in the US and the UK. The authors propose an agenda for cancer screening research in older populations, based on US and UK experiences. Cancer 2008;113(12 suppl):3493–504. © 2008 American Cancer Society.

The population in the Western world is aging and living longer, with those aged ≥65 years among the fastest-growing segments.1 Cancer is the second leading cause of death in the US among those aged ≥65 years.2 Screening has been shown to reduce breast and colorectal cancer-related mortality among all individuals aged ≥65 years. There is also evidence that cervical cancer deaths may be reduced through early detection among the older population. Evidence of prostate cancer screening efficacy continues to emerge but to our knowledge is thus far equivocal.3 As a result, a recent review of cancer screening in the US concluded that the decision to screen older individuals should be based on individual health status, the specific benefits and harms of the test, and patient preferences, rather than solely the age of the patient.4

Given the growth in the number of older people, expansion in the field of geriatric oncology, and a consequent interest in cancer screening decision-making among the elderly, the objectives of the current article were to identify some of the conceptual, methodologic, and practical challenges to and opportunities for research on cancer screening among older individuals. We propose a taxonomy of barriers to research; these barriers have limited the accumulation of evidence for shared decision-making5 regarding cancer screening.

First, we describe screening rates among older individuals, evidence for effectiveness, and the benefits and costs of screening. Next, we detail barriers at 3 levels, the macro level (policy and population), the patient contact level (organization and provider), and the group and individual patient level. Policy-level and population-level barriers include varied professional guidelines for screening by the individual's age, differing screening programs and reimbursement strategies, and limited national cohorts of older individuals for clinical research. Organizational and provider-level impediments include limited enrollment in clinical trials, insufficient and inconsistent communication regarding screening between providers and patients, and limited provider skills in shared decision-making. At the interindividual and intraindividual level, comorbidities, functional disabilities, reduced awareness of screening, cognitive declines, differing values for more life, varied perceptions of cancer, and limited social support or the need for family involvement may delimit the individual's ability to make informed decisions concerning screening. We will conclude by suggesting future directions for cancer screening research, based on US and UK experiences.

Screening Rates Among Older Cohorts

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Rates of cancer screening tend to be highest among the age groups in which there is the greatest evidence that early detection is effective. As discussed in more detail below, this pattern largely reflects policy and population-level barriers to early detection, in which screening is only formally offered as part of an organized program (eg, via invitation) or through reimbursement to age groups in whom the evidence of efficacy exists. For example, in mammography, uptake is highest among those ages 50 to 64 years, and lower among women aged ≥65 years and those aged <50 years.6 Rates of mammography adherence appear to be similar among women aged ≥65 years, however, compared with women ages 40 to 50 years, in whom the recommendations are much more contested.6 In colorectal cancer, screening uptake is highest among people ages 55 to 74 years, with lower adherence reported among younger and older groups, although the upper age at which adherence rates decline has varied slightly across ethnic/racial groups (age 70 years for Hispanics vs age 74 years for whites and blacks).7 In cervical cancer screening, women aged ≥65 years are least likely to be screened, and are most likely to present with late-stage disease.8 In addition, older, minority, and non-English-speaking women are disproportionately represented among the unscreened and underscreened.9

In contrast with other recommended cancer screening, prostate cancer testing via the digital rectal examination and the prostate-specific antigen (PSA) test uptake does not decrease with increasing age. Uptake is highest among men aged ≥65 years, with 76.6% having had a PSA over the past 2 years, according to the Behavioral Risk Factor Surveillance System.

Screening Effectiveness Among Older Cohorts

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Because a healthy man can expect to live 75 years, and a woman to age 80 years,10 screening issues are increasingly important among older cohorts (that is, those aged ≥65 years). The aging population is heterogeneous physiologically, psychologically, socially, and culturally.11 Many among the general public believe that cancer screening is universally beneficial and represents appropriate personal responsibility, even for those aged 80 years.12 However, there is continued controversy regarding whether screening in the oldest age groups results in clear health benefits.

Despite a relative lack of data, as will be discussed later, researchers often generalize the evidence from randomized clinical trials (RCTs) to older age groups, provided they have reasonable life expectancy and are free from significant comorbidities (such as myocardial infarction and diabetes13). In the case of breast cancer, the probability of gaining benefit from screening could theoretically increase with age because cancer incidence is higher among older women than younger,14 with the peak incidence among those ages 75 to 79 years.15 Among older women, breast cancer also demonstrates a more favorable diagnostic profile, with disease characteristics that may require less aggressive therapies; for example, older women are more likely to have estrogen receptor-positive tumors that can be treated with hormonal therapy rather than chemotherapy.16 Compared with younger women, breast cancer screening among older women is also more sensitive.17 Hence the ratio of benefit to harm appears to improve with increasing age.

Although the evidence for screening older individuals for colorectal cancer is to our knowledge particularly sparse,18 most professional guidelines do not recommend an age at which to stop, leaving the decision to clinical judgment.19 The prevalence of advanced neoplasia continues to increase with age, suggesting the importance of continued screening in the elderly.20 A recent clinical literature review21 concluded that because the majority of deaths from colorectal cancer were among those aged ≥65 years, population-based screening of the aged could yield considerable benefits. A current study using Surveillance, Epidemiology, and End Results (SEER) data to model the risks and benefits of screening in patients ages 70 to 90 years22 suggested, however, that the potential benefit from screening varied with age, life expectancy, and screening modality; these criteria may be particularly relevant to morbidity and mortality among those aged ≥75 years.21

Recently, Lin et al23 examined the effect of screening colonoscopy on life expectancy of 3 age groups (ages 50-54 years, ages 75-79 years, and aged ≥80 years) and found that, although older people had more colorectal polyps detected, the life expectancy gain was smaller for them. The results of a retrospective trial among those aged ≥80 years24 suggest that, because of low diagnostic yield, colonoscopy should be limited to elderly patients with symptoms or specific findings. Therefore, for any individual patient the potential for harm from screening must be weighed against the likelihood of benefit, as well as the acceptability of an individual test (fecal occult blood test [FOBT], sigmoidoscopy, and colonoscopy), especially for those with a shorter life expectancy. Additional data to assist with this clinical decision-making for older individuals are sorely needed.

To the best of our knowledge, there is no consensus across the professional guidelines regarding the age at which to stop cervical cancer screening.20, 21 Mandelblatt et al25 applied a Markov model to hypothetical cohorts of women ages 20 to 65 and 75 years, or to death, with screening via the Papanicolaou (Pap) test plus oncogenic human papillomavirus (HPV) DNA test, Pap test alone, and HPV testing alone performed every 2 to 3 years. Their findings revealed that lifetime screening (particularly the cost-effective combination of the HPV and Pap test every 2 years) saved lives, although nearly all of the benefits are achieved by ages 65 to 75 years. After this age, the benefits must be carefully weighed against the harms, including those of false-positive results. Among those who are unscreened, however, testing is critical to reducing cervical cancer mortality. A greater life expectancy yields greater gains from screening,26 and women must have access to routine follow-up for positive findings.

Prostate cancer screening is more contested than cervical cancer screening, particularly among older individuals. The conclusions from 2 recent reviews suggested that screening men aged ≥70 years should be reserved for those with very long life expectancies,27 or for otherwise healthy men.28 In addition, Catalona et al28 recommended that PSAs for older men be compared with the median biomarker value for the age group in men without known prostate cancer to determine whether to perform a biopsy. One set of authors contend that treating men aged ≥75 years is unlikely to increase survival, and could raise the risk for serious complications and therefore population-based screening appears to be unwise.28 Another set suggests tailoring PSA testing to men age ≥70 years who have few comorbid conditions and who may still benefit from treatment.29 Analyses of older men from a major ongoing trial29 may provide additional evidence in support of a single strategy, although the predilection for prostate cancer screening may still persist for a small proportion of older patients.30

Benefits and Costs of Screening

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

The balance of benefits to harms become more favorable for screening in older age groups with reductions in cancer mortality, false-positive rates, anxiety because of unnecessary workups, and overtreatment. There is an issue of offering screening to groups who would not accept treatment if it were offered.31 At the clinical level, as mentioned previously, the patient's life expectancy, values, and preferences should also inform the decision to screen.32 Furthermore, RCTs report the average effectiveness of an intervention, without addressing the influence of individual patient characteristics, such as comorbid conditions, that are highly variable and increase with age, changing the balance of individual risks and benefits.33

Access to follow-up and acceptance of treatment for positive findings from screening may also vary by age. Rates of follow-up after abnormal test results from the Pap smear appear to be lower among older women than younger.33 Just as with younger groups, a negative screening experience, particularly a false-positive finding,34 may remain a barrier to subsequent screening, even after the findings of a follow-up examination are normal.

A recent study using 8 years of linked SEER-Medicare claims data reported that breast cancer treatment delay is less among the older old (aged >80 years) than the younger women (ages 65 years-79 years), potentially improving survival.35, 36 Nonetheless, several studies have documented that older patients with cancer receive less treatment than younger patients,37–39 and this alters the benefit calculus for older women. Accordingly, there is a need for additional studies of screening and follow-up for positive findings, diagnosis and treatment delay, access, and acceptance among older individuals; when benefits are favorable compared with risks, intervention studies are also warranted.

Repeat Screening Adherence and Surveillance

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Although there is controversy concerning the protective effect of 1-time screening in an elderly population, recently Lash et al40 have been the first to demonstrate an important reduction in the odds of death from breast cancer associated with repeated mammography in women aged >74 years who already had a diagnosis of stage I and II breast cancer. This survival advantage was because of the identification of new primary tumors as well as recurrences of the original disease. Adherence issues are a problem with cancer survivors as well as elderly women drawn from the general population, as the Lash et al study demonstrates. Even in the older age group assessed within a managed care setting, up to 60% of these survivors were not fully adherent with repeat mammography. Demonstrating the promise for measuring adherence over time, however, Rakowski et al41 found that predictors of repeat mammography were comparable to those for recent mammography in a cross-sectional national probability sample.

There remain some important methodologic challenges in assessing screening adherence over time among older people because few cohort studies include measures of cancer screening, older adult enrollment in RCTs is sparse, and survival bias is a strong confound. These and other barriers to research on cancer screening among older individuals are discussed under the following multilevel taxonomy.

Policy- and Population-level Barriers to Research on Screening Among Older Individuals

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

The organization of screening

Organized (or single-payer) healthcare systems usually target screening to the age group most likely to benefit in terms of life–years saved, in an attempt to reduce (premature) mortality from cancer at a population level with reasonable cost (eg, UK National Screening Committee). By contrast, in systems in which private healthcare is the norm, screening may be tailored more to individual risk status, although evidence-based guidelines may still exist. (The US Preventive Services Task Force recommendations closely mirror the screening provision offered in the UK.) Countries endorse screening for those cancers that have reduced cancer-specific mortality (eg, breast and colorectal42, 43) through the provision of national screening programs that are free at the point of delivery or through reimbursement schemes (such as via Medicare or Medicaid in the US). Consistent with the view that screening is likely to be beneficial for people aged >74 years, no upper age limit has been set for breast or colorectal cancer screening in either the US or the UK. (Again, the US Preventive Services Task Force recommendations closely mirror the prostate screening provision offered in the UK when it is endorsed [such as by the American Cancer Society]).

In the UK, women ages 50 years to 69 years are sent a timed, dated appointment for a mammogram every 3 years. Once they reach age 70 years, women are no longer sent invitations to attend but are encouraged to make their own appointments for screening.44 Similarly, in the UK national colorectal cancer screening program (using FOBT), men and women ages 60 years to 69 years are routinely sent an FOBT kit; but once they reach age 70 years, the tests are only sent on request. Timed, dated appointments have been demonstrated to improve uptake of screening,45 so the cessation of the call-recall system around the age of 70 years would be expected to have a detrimental effect on uptake.

Although the chance of being diagnosed with cancer increases with age, 1 notable exception is cervical cancer, in which the overall risk is lower among older than among younger women.46 The US Preventive Services Task Force recommends discontinuing cervical cancer screening at age 65 years. In the UK, women aged ≥65 years are only entitled to be screened if they have never had a Pap test, have not had 1 since the age of 50 years, or have had recent abnormal tests. Otherwise, once they have had 3 consecutive negative smears, they are taken off the national call-recall system and the test is no longer available. Cervical screening in older adults is therefore targeted at higher-risk groups, rather than the whole population who are at average risk.

Although its practice is quite widespread in the US,47 there is no formal screening program for prostate cancer in the UK and rates of PSA testing are much lower, at approximately 6%.48 The Prostate Cancer Risk Management Program (PCRMP) was introduced in the UK in 2001, which enables men to have a PSA test on request provided they have been given full information regarding the possible harms and benefits; this is the first screening service to be offered on the basis of an individualized consultation.

Interestingly, there is only 1 national ‘1-stop’ cancer screening site in the US, the National Breast and Cervical Cancer Early Detection Program (NBCCEDP), which offers free or low-cost breast, cervical, and, in some states, colorectal cancer screening. Only 6.2% of women screened by the Pap smear in the NBCCEDP are aged ≥65 years, however.49

Organized programs differ by their outreach, requirements, and their reimbursement for screening by age, screening adherence, methods of enhancing uptake, performance (in cancer detection rates), and the psychologic impact of screening. The impact of these variations on screening rates among the elderly are confounded by differences among those who select to attend organized programs and those who decline. It is critical to examine how and in what way these varied organizational structures and their components influence initial uptake, and establish adherence over time, among older people as they age.

Organizational and Provider-level Barriers to Research on Screening among Older Adults

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Provider recommendations for screening

Physician recommendation is the strongest predictor of screening uptake in the US, but there is evidence that this recommendation is more forthcoming in younger age groups for both Pap smears50 and mammography.51 As women age, experience more health difficulties, and lose critical body cues to reproductive health such as regular menses, external prompts such as physician screening reminders, prevention-oriented office tools, and techniques for tracking and prompting screening52 become more important.

Physician awareness of guidelines that screening should only be recommended provided older adults have good life expectancy (eg, 10 more years for the PSA) is unclear. PSA testing among men aged ≥70 years demonstrated no correlation with life expectancy, as recommended by most of the professional guidelines. These findings suggest that patients and/or physicians may not be considering prognosis when deciding on a PSA test.53

Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Some of the influences on nonattendance at screening, such as poor health54 and functional disability,55 would be expected to increase with age. Other factors such as socioeconomic status and previous cancer screening may exert similar influences on both older and younger people.56 However, to our knowledge, research into whether there are factors that affect older people disproportionately has been scant.

In the UK at least, research into beliefs concerning screening and the understanding of cancer screening information leaflets has often been conducted on the age group actively invited to attend. For example, research concerning understanding of breast cancer screening information has focused on women ages 49 to 64 years (who used to be) invited to national programs.57 In addition, research regarding risk perceptions of colorectal cancer has been performed primarily in individuals ages 55 to 64 years, although older age has been associated with lower perceived risk.58 New forms of screening and intervention approaches tend to be concentrated on the group for whom the evidence base is strongest, increasing the need for high-quality research among a broader age range of older people.

Awareness of the availability of screening

To our knowledge, little is known regarding the correlation between levels of awareness and the availability of screening on request for those aged ≥65 years. A recent UK survey (conducted in 2007) found that awareness of NHS cancer screening programs for breast and cervical cancer were lower among women aged ≥65 years than among younger women (unpublished data). In the US, a random digit dialing telephone survey of women aged ≥65 years59 found that awareness of Medicare reimbursement for mammography was >75%, but minority women were twice as likely to be unaware of it as other women.60 The authors pointed to the need to provide information to older women about Medicare coverage, particularly in cases in which they were not claiming reimbursement for screening.

Cognitive decline

Many healthcare systems have shifted from external to self-determined screening, with the individual expected to play a greater role in these decisions. This is reflected in moves toward informed and shared decision-making in the US. In the UK, the transition from a call-recall system to screening on request at age 70 years places decisional responsibility on individuals at a time when their cognitive abilities to make good decisions may be declining.

A review of age differences in rational versus emotional processing61 found evidence demonstrating that older adults process less information less quickly than younger adults, have deficits in explicit memory and learning, are less aware of the factors that influence their judgments and decisions, and are less able to control the impact of automatic processing on their judgments. Older adults also tend to focus more on the emotional context of information. The authors concluded that reliance on affective as opposed to deliberative (slower and more rational) processes may increase with age. Given the limited study of this area, there is therefore an urgent need to examine the role of age on judgments and decision-making around screening, particularly when the information is likely to be complex or the behavior relatively novel. Similarly, more investigation of decision aids that providers can use with older individuals to enhance shared decision-making would be useful.

Valuing life

Another issue to consider is how people value extended life years at older ages, when many of those years may be dogged by illness and disability. Older adults may be reluctant to give up a relatively healthy year of life to potentially painful and disabling cancer treatments in exchange for a slightly longer overall lifespan in later life. In other words, the value placed on current health may be considerably higher than the value placed on longer-term health in people aged ≥65 years compared with those in younger groups.

Perceptions of cancer

People's perceptions about cancer play an important role in whether they adhere to the recommended guidelines about screening. Despite increased breast cancer incidence and mortality with older age, knowledge of age as a risk factor for cancer is poor.62 A population representative survey of 996 women ages 16 to 96 years in the UK63 found that the majority of women (70%) believed the age range of greatest risk is ages 35 years to 59 years, followed by ages 60 years to 74 years (10%). Patients ages 74 years to 90 years were least frequently endorsed as those at highest risk. Consistent with this, 17% of the overall sample reported themselves less likely to develop breast cancer than other women, but this increased to approximately 30% among women aged ≥65 years. In open-ended explanations of reduced risk, 25% of women with a perceived risk lower than average believed that they were ‘too old’ to get cancer, expressing the belief that the ‘danger has passed,’ ‘they stop mammograms at 65.’ Hence, an unintended consequence of stopping the British call-recall system at age 70 years (rather than formerly at age 65 years) is that it sends the message that breast cancer risk declines with older age.

Perceptions of being ‘too old’ for screening have also emerged in qualitative studies in people who declined colorectal cancer screening despite being aged approximately 60 years,64 and these also appeared to be based partly on perceptions of life stage and life expectancy. Such beliefs need to be explored in relation to the decision to decline screening; they may reflect quite rational responses to the offer of screening and an appropriate caution in individuals who have comorbidities, or alternatively a lack of awareness of how effective screening might be in extending life expectancy.

Perceptions of the severity of cancer also appear to reduce with age. For example, Eisner et al59 found that approximately one-third of their sample (aged ≥65 years) said they were not as concerned about getting breast cancer as they had been when they were younger, and this proportion was higher among women aged ≥70 years than among those ages 65 years to 69 years. The perceived threat and hence the fear associated with cancer therefore appears to decrease with advancing age, although the authors noted that this reduction may reflect stronger beliefs in the efficacy of screening.

To our knowledge, however, little work has been done to assess older people's perceptions of the benefits and harms of screening. A key benefit is the reduction in cancer mortality that screening affords. Research into screening efficacy beliefs in older adults is scant. Older patients are as likely to know their diagnosis but less likely to understand the stage of their cancer as younger individuals.65 This may point to a lack of understanding regarding the importance of early detection, although perceptions of the benefits of mammography in reducing mortality have demonstrated little variation with age.66 The latter study, which examined beliefs concerning mammography among women ages 50 to 85 years who were nonadherent, found that perceived barriers to mammography such as time, pain, and embarrassment declined with age but only among white women. An increase in perceived barriers was observed with age among African American women, controlling for education and income. However, in the absence of population representative data, it is difficult to assess whether these findings are robust. In their population-based survey of women's beliefs about breast cancer, Grunfeld et al63 found older women (aged ≥65 years) were more likely to believe that breast cancer treatments were more effective in younger women. In addition, older women were more likely to believe that breast cancer treatments always led to disfigurement of some kind, and they were less likely to believe that a good quality of life could be enjoyed by women who had been through such treatment.

Social support

Living alone67 increases with age, and amplifies the risk for screening nonadherence. In the absence of a cohabitee to promote screening adherence, the impact of the beliefs and behaviors of an older person's close friends and relatives on screening has to our knowledge rarely been explored. Studies of cancer survivors suggest that some older patients want family to make decisions for them68, 69; inclusion of significant others in discussions of cancer management is also associated with older patients reporting higher levels of shared decision-making.70 These data suggest that involving significant others in discussions regarding cancer screening may improve older adults' decision-making as well, and better protect their interests, particularly if that communication is tailored to their needs and concerns.71 To our knowledge, few studies to date have been conducted among individuals aged ≥65 years, however.

Comprehensive measures

Studies of the influences on screening adherence among older people would benefit from additional research using comprehensive geriatric assessments (CGA) and multidimensional geriatric assessments (MGA) that evaluate comorbidities, functional limitations (dependence on at least 1 Activity of Daily Living), and/or geriatric syndromes (such as urinary incontinence, falls, failure to thrive, osteoporosis).72, 73 Some preliminary data suggest that the findings from these comprehensive assessments predict morbidity and mortality. Additional psychometric studies of CGAs or MGAs, particularly to distinguish their factorial structure (eg, dementia is included in the Charlson Comorbidity Index,74 and as a geriatric syndrome in many CGAs or MGAs) are warranted. The subsequent application of the revised instruments to studies of cancer screening adherence could inform the design of interventions.75

Quality assurance for screening

There are important quality assurance concerns for screening among older individuals. Because of limited mobility and comorbidities, it may be more difficult to obtain adequate biologic samples from the cervix, or the colon, or adequate images of the breast among older people compared with younger individuals.

For people with mobility problems, self-sampling techniques are a solution, but we know little about the ease of use among older age groups. Different methods of stool collection may enhance uptake; there is evidence of stool sampling preferences (‘card with stick,’ ‘tube with probe,’ and ‘wipe’) among those ages 50 years to 69 years, but older ages have not been surveyed.76 Similarly, attitudes toward self-sampling a Pap test among women not attending organized cervical screening for more than 6 years were explored among women ages 35 years to 55 years,77 but not among older women. Additional studies of the performance characteristics and feasibility of varied types of screening tests among older individuals are warranted.

Implementing a Research Agenda

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Enrollment in RCTs

One of the problems with providing evidence-based screening is that efficacy data are scarce for older age groups. Exploring accrual to National Cancer Institute (NCI)-sponsored clinical trials over a 12-month period of time, Sateren et al78 found that greater than half of patients ages 5 years to 9 years were enrolled in trials compared with <1% of patients between the ages of 75 and 79 years. RCTs for mammography have tended to focus on women ages 50 to 74 years,79, 80 with some RCTs conducted in younger groups.40–49 In the absence of RCTs in women aged ≥75 years, evidence from observational studies supports the efficacy of mammography for women ages 75 years to 84 years, but to our knowledge there are few data regarding women aged ≥85 years,81 or for those with common comorbidities.

In general, physician, patient, organizational, and trial-related barriers have impeded enrollment in clinical trials.82, 83 Physician barriers84–87 include time constraints, treatment preferences and age-related biases, concern regarding the clinical relationship, lack of experience with trials, lack of interest in research in general or the specific study protocol, and economic disincentives. Patients are frequently unaware of the availability of clinical trials, unwilling to be randomized, and mistrustful of or confused about medical experimentation or the informed consent language, and they frequently present other social, cultural, and sociodemographic barriers to participation in trials.88, 89 More general characteristics of cancer-related decision-making among the elderly, such as concern with quality rather than length of life, as discussed earlier in this article, also affect enrollment in clinical trials.

Organizational barriers,90 such as a lack of research infrastructure or the presence of an organizational culture that is unfriendly to research, and perceived increased medical care costs for participation inhibit trial enrollments. Trial-related barriers91–93 include strict eligibility criteria, lack of equipoise/uncertainty, or trial-recommended therapy that is not appropriate for all eligible patients (because of comorbidities or performance status, for example).

Leveraging existing cohorts for research

Over the past 10 years, strategic initiatives have been developed to explore and develop a research agenda for collaborative and integrative efforts between the fields of aging and cancer (Table 1).94 At the current time, the National Institute on Aging (NIA) supports 55 longitudinal cohort studies of older individuals. Many of the NIA-sponsored longitudinal studies have excellent retention over time, are population-based, collect standardized measures of quality of life, and all-cause mortality. However, measures of cancer screening and surveillance behaviors, or cancer-specific measures of morbidity and mortality, are relatively rare in any of these prospective cohorts of the aging.

Table 1. Selected Cohorts for Research on Cancer Screening and Aging
  1. SEER indicates the National Cancer Institute's Surveillance, Epidemiology, and End Results Program; NCI, National Cancer Institute.

In the UK, the Whitehall studies recruited adults in 1985 and will follow them through 2009 to measure disease-specific mortality and quality of life. The English Longitudinal Study of Aging studies quality of life among those aged ≥50 years, with 10% of the cohort aged ≥80 years, but has no measures of cancer screening.

There are several well-designed, cross-national studies of aging; for example, the Health and Retirement Study; Survey of Health, Aging, and Retirement in Europe; and the Korean Longitudinal Study of Aging that promise increased understanding of the impact of different policies, systems of care, and social and economic factors on screening among the aging. Their contributions to the understanding of cancer screening could be further enhanced by the use of comparable datasets, a common core of health measures, and reliable measures of health outcome (from death certificates, for example), with systematic correction for survival bias and triangulation of data sources (eg, from self-report and biomarker or health administrative data).95

Unlike randomized clinical trials or population-based, cross-sectional studies (such as the National Health Interview Survey in the US), retrospective and prospective cohorts could illuminate the rates of initial and repeat adherence over time among individuals who are exposed to similar contextual influences (such as changes in professional guidelines, media coverage, and physician training).

Retrospective cohort studies using linked SEER-Medicare claims data offer opportunities for population-based research in cancer screening and surveillance among those aged ≥65 years, particularly in the exploration of patient comorbidities, physician characteristics, and health service system influences. State and country-based cancer registries, including those linked to projects under the aegis of the NCI's Breast Cancer Surveillance Consortium or the International Cancer Screening Network, offer additional promise for both prospective and retrospective community-based research on breast cancer screening, particularly as participating mammography sites collect standardized performance criteria.

The national membership organizations for the aging, including the American Association of Retired Persons (AARP) and the National Council on Aging (NCOA) in the US, the Association of Retired Persons over 50 (UK), and the International Federation on Aging (Canadian-based) are potential sources for research participants, although their associates are self-selected and they are not research cohorts. Because some of the larger affinity organizations, such as the AARP, also offer discounts on insurance and legal services, their members may vary in health and functional status, enabling the recruitment of a diverse participant pool. Furthermore, because many of these organizations include education or information-sharing in their mission statements, they may be particularly receptive to studies that could demonstrate concrete benefits to their members.

Future Directions

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Although there may be consensus that breast and colorectal cancer screening can save lives if offered to older adults, there are numerous critics of any moves toward offering screening to adults aged ≥65 years because of the low number of life–years saved. Some argue that although screening offers some benefit, resources would be better expended elsewhere. It is also possible that age biases in the healthcare system interfere with the delivery of effective, optimal, prevention and early detection services.96

Funding of screening research in the elderly is not proportionate to that for younger populations. Although greater than half of cancers occur in the elderly, only 10% of health services research is targeted toward them.97 This is of special concern with regard to new technologies. For example, in a recent trial examining magnetic resonance imaging, new breast cancers were detected that were not seen on clinical examination or mammography, but no women aged ≥65 years were included.98

Cancer screening research with older adult populations is unjustifiably limited and far too few empiric studies have been conducted to date. Future research in cancer screening should be devoted to developing more definitive screening guidelines,99 decreasing barriers to screening, leveraging existing cohorts and outreach to volunteer organizations to enhance the participation of older adults in screening trials,100 and increasing the development and testing of evidence-based and theory-based intervention approaches. There is an urgent need for a concerted effort to increase the sum of knowledge regarding screening in older adults at a time when life expectancy is increasing in most parts of the world.

Summary of Recommendations for Future Research on Cancer Screening Among Older Adults

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES

Overall

  • Increase high-quality screening research among a broader age range of older people, including evaluations of repeat adherence; follow-up for positive findings; screening, diagnostic, and treatment delays; and access to and acceptance of screening. The need is particularly acute for colorectal cancer screening.

  • Increase the development and testing of evidence-based and theory-based intervention approaches to enrich screening adherence over time by those aged ≥65 years, in whom screening has demonstrated effectiveness.

  • Evaluate evidence-based interventions designed to increase enrollment by older adults in RCTs, particularly racial/ethnic minorities.

  • Add to existing longitudinal cohorts sponsored by NIA core measures of screening predictors, screening behavior, and cancer outcomes. Triangulate self-report and administrative claims or biomarker data; analyze with systematic correction for survival bias.

  • Leverage existing cohorts of older individuals in affinity organizations such as the AARP for research on cancer screening and its outcomes, including morbidity, mortality, and quality of life.

Policy and population levels

  • Examine how varied organizational structures for screening and their components influence initial uptake, and establish adherence over time, among older people as they age.

Organizational and provider-levels

  • Explore novel approaches to increasing provider awareness and implementation of professional screening guidelines.

  • Develop and test decision aids that providers can use with older individuals to enhance shared decision-making for screening.

Interindividual and intraindividual levels

  • Research the factors, in addition to comorbidities and functional disabilities, that disproportionately influence screening adherence among older individuals relative to younger people.

  • Explore approaches to enhance awareness of the availability of screening, particularly among the never-screened.

  • Examine the role of age on judgment and decision-making around screening, particularly when the information is likely to be complex or the behavior relatively novel.

  • Consider how people value extended life years at older ages, and the impact of varied valuations on screening behavior.

  • Enrich the scant literature on perceptions of cancer, including older people's perceptions of the benefits and harms of screening, screening efficacy beliefs, risk perceptions, and beliefs in the severity of cancer as influences on screening adherence over time.

  • Explore the role of the social network in screening decisions, particularly their attitudes and beliefs toward screening, as well as the implications of their absence.

  • Conduct additional psychometric studies of CGAs or MGAs. Apply revised instruments to the study of screening among those aged ≥65 years in community and institutional settings.

  • Enrich studies of the performance characteristics and feasibility of varied types of screening tests with older individuals.

REFERENCES

  1. Top of page
  2. Abstract
  3. Screening Rates Among Older Cohorts
  4. Screening Effectiveness Among Older Cohorts
  5. Benefits and Costs of Screening
  6. Repeat Screening Adherence and Surveillance
  7. Policy- and Population-level Barriers to Research on Screening Among Older Individuals
  8. Organizational and Provider-level Barriers to Research on Screening among Older Adults
  9. Interindividual and Intraindividual Barriers to Research on Screening Among Older Adults
  10. Implementing a Research Agenda
  11. Future Directions
  12. Summary of Recommendations for Future Research on Cancer Screening Among Older Adults
  13. Acknowledgements
  14. REFERENCES