Fax: (301) 480-6637
Article first published online: 27 JUL 2004
Published 2004 by the American Cancer Society
Supplement: Promoting Cancer Screening: Lessons Learned and Future Directions for Research and Practice
Volume 101, Issue Supplement S5, pages 1251–1259, 1 September 2004
How to Cite
Meissner, H. I., Vernon, S. W., Rimer, B. K., Wilson, K. M., Rakowski, W., Briss, P. A. and Smith, R. A. (2004), The future of research that promotes cancer screening. Cancer, 101: 1251–1259. doi: 10.1002/cncr.20510
This article is a U.S. Government work and, as such, is in the public domain in the United States of America.
The opinions expressed herein do not necessarily reflect the views of the National Cancer Institute, the Centers for Disease Control and Prevention, or the U.S. Government.
- Issue published online: 18 AUG 2004
- Article first published online: 27 JUL 2004
- Manuscript Revised: 26 MAY 2004
- Manuscript Accepted: 26 MAY 2004
- Manuscript Received: 22 APR 2004
- Behavioral intervention;
- cancer screening;
- standardized measures;
- research design;
- external validity
The authors draw on the lessons highlighted in preceding articles in the current supplement to provide recommendations for cancer screening intervention research and to highlight some of the many questions that will require further investigation. Cancer 2004. Published 2004 by the American Cancer Society.
In the current supplement to Cancer, lessons learned from an evolving literature base were highlighted with the objective of informing future screening intervention research. It is clear, however, that the lessons discussed do not constitute a cookbook. Some lessons are likely to translate to new cancer screening tests. For example, the evidence base developed as a result of experience with mammographic screening already has informed many of the approaches used to promote colorectal cancer screening (e.g., the use of multiple strategies aimed at multiple targets). It is difficult to predict the extent to which these lessons will apply to the next generation of screening tests. For example, how will ethical and economic factors that influence the use of today's screening tests apply to the advent of genetic and molecular tests? These issues are especially challenging when the magnitude of benefits is uncertain. We recognize that differences among tests (e.g., differences in terms of invasiveness) will necessitate further evaluation of certain promotional approaches, which may be unique to each test and to each cancer site.
How, then, should these lessons shape future approaches to cancer screening intervention research? Here, we attempt to answer this question by making seven research recommendations based on consistent themes that were identified in the preceding articles. In addition, we highlight some of the many questions that will require further investigation.
Recommendation 1: Conduct Behavioral and Social Science Research Concurrently with the Development of New Screening Tests or with Changes in Guidelines for Established Tests
Behavioral science, as noted in the current supplement,1 can make important contributions to the design of screening approaches that will be acceptable, effective, and cost effective; to the identification of important barriers and facilitators to uptake; and to screening use itself. Behavioral and social sciences have important roles to play in matching screening approaches to the context in which screening occurs. We can predict that the context for screening will be different in the future. Advances in molecular and genetic testing may redefine the issue of who is susceptible to cancer. New screening technologies will identify cancers much earlier, raising new questions regarding what constitutes appropriate follow-up and treatment and who should be screened.
The social and community context in which each screening test is developed has an important effect both on the extent to which screening diffuses without specific interventions and on which interventions are likely to be feasible, appropriate, and effective. Some tests can and will be used for screening before their efficacy has been established clearly (e.g., prostate-specific antigen [PSA] testing for prostate cancer and spiral computerized tomography [CT] for lung cancer). In contrast, some tests engender slow uptake irrespective of their efficacy (e.g., fecal occult blood tests [FOBT]). Characteristics of the test itself (e.g., the apparent simplicity and low cost of the PSA test) contribute to its potential for dissemination. In general, however, rates of screening are unlikely to be optimal without promotion.
Trends in the prevalence of cervical, breast, colorectal, and prostate cancer screening show that the uptake of screening varies from test to test. One lesson we can draw from more than 20 years of experience with cancer screening is that behavioral research on screening tests should be conducted long before efficacy is demonstrated. Such research can identify possible barriers to acceptance and dissemination from the perspective of providers and patients. Likewise, with the emergence of new scientific evidence, screening guidelines may be revised in ways that could affect their acceptance and adherence. For example, recent changes in cervical cancer screening guidelines (http://www.ahrq.gov/clinic/uspstf/uspscerv.htm) resulted in a more complex message regarding the frequency of Pap smear screening for different age groups and will necessitate the modification of established patient and provider behaviors (e.g., women and their providers are accustomed to the routine of annual Pap smears). Preliminary behavioral research suggests that women may be uncomfortable with the recent recommendation to lengthen the interval between Pap tests from 1 year to 3 years, because they have long been told that annual screening is effective. Qualitative research indicates that some women interpret the shift in screening intervals as a cost-saving maneuver by insurance companies rather than as a change motivated by science.2 This reaction is consistent with reactions to the changes in mammography recommendations that occurred in the 1990s.3 Furthermore, current thinking regarding incorporation of testing for a sexually transmitted virus (human papillomavirus [HPV]) into cervical cancer screening programs will not only necessitate a change in messages directed toward women and clinicians but also will introduce a new population group (namely, women's partners) to cervical cancer control efforts.4
Recommendation 2: Consider Multilevel Conceptual and Analytic Approaches That Recognize the Context in Which the Process of Cancer Screening Occurs
Screening and appropriate follow-up involve complex interactions among individuals, their social networks, health care providers, organizations, and the broader social and economic environment. To date, research has not systematically addressed the contribution of geographic and societal phenomena that may influence the process of screening, although it is known that these factors have independent effects on individual health behavior, morbidity, and mortality.5–7 The ecologic frameworks cited in the current supplement8, 9 and elsewhere recognize the ways in which context can have a direct influence on screening and follow-up by shaping the characteristics of the environment that may facilitate opportunities for or present barriers to use (e.g., policies mandating that insurers offer coverage of screening tests and follow-up services). This context, in turn, can affect the characteristics of the social and health service environments (e.g., the supply of health professionals, area norms, and health-promoting practices) to which members of a community are exposed regardless of their individual characteristics. Ecologic models also account for interactions between the community and health plans, employers, and medical groups. Health plan characteristics that may influence provider performance include features such as how providers are paid and how risk is shared. In turn, the practice setting, in which interactions between providers and patients occur, ultimately shapes decisions regarding adherence to guidelines and follow-up.
Multilevel or contextual approaches to health behavior can offer further insight into how different levels of influence contribute to patterns of screening uptake.10–12 Conceptualizing screening as a product of social and physical environments and of individual characteristics acknowledges that the environment can affect individual behavior through multiple pathways; many of these pathways have not been identified or explored to date. A recent Institute of Medicine (IOM) report addresses the need to conduct more research within the context of multilevel models.6
Multilevel analytic methods allow investigators to distinguish individual sources of variation from contextual sources.13 In addition to assessing the contribution of individual-level and macrolevel effects on the distribution of outcomes, multilevel techniques permit researchers to investigate whether apparent location-to-location variations are the same or different across various populations.14 These new methodologic tools should be explored further both to improve our understanding of how social, political, and economic factors shape screening behavior and to determine whether multilevel measures are more powerful than single measures in predicting behavior. Conducting such analyses requires the inclusion of variables across multiple levels in databases that track the adoption of screening modalities, as is discussed elsewhere in the current supplement.1
In the planning of the current supplement, we considered developing a comprehensive conceptual model to represent all possible factors and pathways that influence screening behavior and follow-up. After much discussion and debate, however, we recognized that this effort was both futile and inappropriate. First, intervention research in this area has been driven by multiple models and frameworks, rather than by a single model. Creating a post hoc model could oversimplify reality and reconstruct history in a way that would not reflect reality. Second, it became apparent that the depiction of multiple influences on the screening process would be too complex for a single model. Like others, we concluded that the utility of a given model depends on the specific scientific question being asked and on the level of specificity needed to derive an answer.15 The different models and frameworks used to guide discussion of the topics addressed in the current supplement complement rather than compete with one another. However, we encourage researchers to use analytic frameworks, like the one used in the Guide to Community Preventive Services16 (http://www.thecommunityguide.org/cancer), as well as theoretically guided models as an integral part of research planning.
Recommendation 3: Use Survey Instruments That Reflect the Complexity of Screening Behavior
Rakowski and Breslau1 discuss the need to include behavioral constructs and measures (e.g., indicators of the range of cancer screening, perceived risk, and intention to screen) in population-based surveys. Such efforts are constrained, however, by costs and by the amount of time required to administer lengthy questionnaires. Nevertheless, to capture the complexity and context of screening behavior, it is critical that key behavioral and social science variables be included at least periodically in surveys. Theory and empiric support should inform the selection of measures for inclusion in surveys to maximize the value of these measures in explaining screening behavior.
Recommendation 4: Consider Alternatives to the Randomized Controlled Trial
A key question to be answered is, “Does this intervention strategy affect screening uptake ?” Although randomized controlled trials (RCTs) are widely accepted to be the strongest method for identifying causal relations between interventions and outcomes, it is becoming increasingly apparent that they are not always practical, affordable, or appropriate for assessing the complexity of approaches needed to improve adherence to screening.17 Moreover, the highly controlled and standardized nature of an RCT may preclude programs that will be successful in practice.18 Further dialogue among researchers and between researchers and practitioners is needed to broaden the repertoire of research designs for assessing the effects of screening promotion. Expansion of the range of alternative methods (e.g., observational methods, time series) is necessary. More explicit guidance regarding the relative appropriateness of alternative study designs for answering questions at different stages of intervention development could optimize efficiency, cost-effectiveness, and the breadth of information captured. For example, less expensive designs (e.g., a pre-post single group) for assessing feasibility can yield important information before a more expensive and potentially less generalizable RCT is conducted.19 Pasick et al.20 also comment on the need for alternative methods (e.g., qualitative research) to explain why a given intervention strategy may be effective, rather than simply assessing that strategy's effectiveness. Understanding why programs do or do not work is of critical importance.
Recommendation 5: Use a Consistent Set of Standardized Behavioral Outcome Measures for Cancer Screening, Uniform Classification Schemes for Intervention Strategies, and Uniform Reporting Methods
The methodologic lessons discussed by Vernon et al.21 highlight ways in which the heterogeneity of outcome measures of cancer screening behaviors used to evaluate intervention studies limits our ability to pool data or make valid comparisons between studies. The need for a standard vocabulary is cited in all areas of cancer screening intervention research.9, 20, 22, 23 Whenever possible, researchers should use standardized measures of cancer screening behaviors for surveillance and evaluation. Additional reliability and validity studies in different populations are needed for all currently available measures (e.g., items from the National Health Interview Survey and the Behavioral Risk Factor Surveillance System). In the absence of these data, we can overcome the problems associated with inconsistency across studies by agreeing to adopt a standard set of core behaviors as outcome measures. Agreement regarding consistent measures that classify where individuals are in the screening process (e.g., never underwent screening, ever underwent screening, recently underwent screening, or undergoing maintenance screening) also is necessary.24 Several initiatives are underway to facilitate the development and use of standardized measures, and Internet-based methods will augment our ability to share them. For example, Vernon and colleagues developed a core set of self-report measures of colorectal cancer screening behavior,25 and this set of measures will soon be available online from the National Cancer Institute (NCI) (http://dccps.nci.nih.gov/acsrb/). Development and evaluation of standardized measures of screening outcomes, stage, and other key variables are critical steps in accelerating our knowledge base regarding the promotion of cancer screening.
In addition to the need for consistent measures of cancer screening behaviors, further agreement is needed on how to assess quantitatively and to evaluate the state of intervention science. The fact that most interventions involve multistrategy approaches presents challenges in developing uniform classification schemes. Several different schemes have been used to classify intervention types (individual-directed, access-enhancing, cognitive, behavioral, etc.),26, 27 but a lack of consistency hampers our ability to adequately conduct the research synthesis necessary to move the field forward. We encourage investigators to describe their studies in typologic fashion when publishing their results.
Because resources available for public health interventions are limited, it is increasingly vital that program implementers capitalize on existing knowledge. The Guide to Community Preventive Services provides a comprehensive review of the effectiveness of population-based strategies for increasing the use of breast, cervical, and colorectal cancer screening and for promoting informed decisions regarding cancer screening. Practitioners are not the only audience for the Guide to Community Preventive Services, however. The Guide's recommendations also should be considered by researchers as they build on the body of evidence and attempt to address gaps in the science. Designing studies that reflect the criteria used in the Guide to Community Preventive Services16 or in related assessments of evidence quality19 will strengthen the body of knowledge available for evaluating interventions aimed at cancer screening promotion. A minimum number of quality intervention studies is required to draw conclusions regarding the effectiveness of a given strategy. Evidence reviews depend not only on the quality of the studies but also on the completeness of reporting. We therefore recommend that intervention scientists adhere to the Consolidated Standards for Reporting of Trials (CONSORT) statement (www.consort-statement.org) or to similar criteria to improve the quality of reporting of intervention results.28, 29 There is no longer any reason for evidence reviews to have to exclude more studies on the basis of inadequate or noncomparable data than they include.
Recommendation 6: Place Greater Emphasis on External Validity
Most of the screening interventions described here and elsewhere have been tested in efficacy studies. Historically, the paradigm for cancer control has emphasized that efficacy should precede dissemination. But what evidence should be required? Glasgow et al.18, 22 argue that the way in which intervention research is conducted limits dissemination into practice. Very few of the efficacy studies that have been conducted have been replicated or evaluated for effectiveness in different populations or settings, perhaps contributing to the failure of many efficacious interventions to find their way into practice. The report by Scott et al.30 on the replication of earlier research is an exception.
The effort to maximize internal validity in efficacy studies, coupled with a lack of emphasis on replication, minimizes our ability to generalize our findings to broader venues and populations. Glasgow et al.22 suggest that this problem is not insurmountable if investigators plan for dissemination by paying more attention to issues of external validity (e.g., participant representativeness) when designing efficacy trials and by not postponing this consideration until effectiveness studies are conducted. Furthermore, replication studies, which constitute an accepted phase of research in the basic sciences, should become routine practice in the behavioral and social sciences. It is critical to investigate how interventions can be adapted across populations and settings.
The wealth of research on interventions designed to enhance cancer screening will do little to improve populationwide health if we are unable to move successful strategies into practice. The first ‘lesson’ suggested by Glasgow et al.22 pertains to this issue. Specifically, we cannot assume that successful programs will naturally disseminate into routine practice. The need for researchers and funding agencies to develop partnerships for dissemination is urgent.
The sooner researchers and funding agencies understand the constraints on intervention delivery in the real world, the more likely programs are to be adopted. For example, the NCI's Health Maintenance Organization (HMO) Cancer Research Network (http://healthservices.cancer.gov/hmo) involves the collaboration of cancer researchers and health care providers within the context of health care delivery systems. The centralized policies and programs of HMOs provide an ideal system not only for research but also for dissemination of results.
A similar collaboration exists between researchers and community groups as part of the Prevention Research Centers of the Centers for Disease Control and Prevention (CDC) (www.cdc.gov/prc). The CDC and the NCI established the Cancer Prevention and Control Research Network (CPCRN; http://ukprc.uky.edu/CPCRN/home.htm) to support the dissemination of research that is carried out through partnerships between communities and researchers. Funding for the CPCRN was set up to allow sufficient local infrastructure to be built and to subsequently distinguish among phases of intervention research with the express purpose of making the eventual dissemination of a proven intervention more likely.
To facilitate the sharing of information and dissemination of intervention research results, the NCI, the CDC, the American Cancer Society (ACS), and the Substance Abuse and Mental Health Services Administration have developed an Internet portal, the Cancer Control PLANET (Plan, Link, Act, Network with Evidence-Based Tools; http://cancercontrolplanet.cancer.gov). PLANET provides access to data and resources that can help cancer control planners, health educators, program staff, and researchers to design, implement, and evaluate evidence-based cancer control programs. The Internet-based resources available on PLANET are the result of a major effort to help practitioners understand research findings and access evidence-based programs and products. Nevertheless, researchers and funding agencies must focus far more attention on designing interventions that can realistically be put into practice.
Recommendation 7: Identify Minimal Intervention Components Needed To Improve Screening and Follow-Up That Can Be Generalized to Diverse Populations and Settings
Notable progress has been made in the use of screening tests nationally. Use of all screening modalities increased in the 1990s; in some geographic areas and for some populations, documented disparities all but disappeared.31, 32 By 1998, racial and ethnic differences in mammography use had diminished significantly.31 Although this trend is encouraging, disparities still persist among groups defined by other indicators. Socioeconomic and health system–related factors—including lower educational level and income, lack of health insurance coverage, and lack of a usual source of health care—are consistently associated with lower levels of screening.31 Although effective interventions exist that increase initial and recent screening in diverse populations, such interventions are underused. More thought must be given to designing interventions that can be adopted outside of research settings. Multicomponent interventions produce changes in screening outcomes, but we need to build a foundation of simpler interventions that are more likely to be disseminated. It is critical for investigators to carefully assess what level of intervention is needed rather than choose interventions that are familiar and have been implemented in the past. Rarely has research established the minimum intervention necessary and for whom that intervention should be performed. Partnerships between the NCI, the CDC, the ACS, and health care delivery organizations could finally create the infrastructure needed to ensure the dissemination of effective interventions.
It is unlikely that research will provide evidence of intervention efficacy for every focal point combination (i.e., every combination of target population, screening objective, and intervention setting), as discussed by Pasick et al.20; however, it should not be assumed that a different intervention is necessary for each focal context. The question that remains, however, is this: How do we know when an intervention is needed for a specific focal point, target population, or setting? A recent IOM report concluded that we cannot assume that specially targeted programs are needed.33 We need to improve our understanding of why a certain combination of intervention strategies works in certain populations and settings. We also should ask whether we can expect these same results in association with different screening behaviors at different stages of adoption in the screening process.
The approaches to intervention described in the articles presented in the current supplement and the lessons learned, as discussed above, attest to how much we have learned about the factors that drive the adoption of cancer screening tests in both clinical and community settings. Although we have identified many lessons from experience in the promotion of cancer screening, in the process, we have raised many issues for future research. In this concluding section, we highlight some of the many questions that require further investigation.
How Hard Should We Try To Screen Every Eligible Individual Once High Levels of Screening Uptake Have Been Achieved?
How much effort should be expended to reach the last 5% or 10% of the population? Interventions are less likely to be cost effective when screening rates are high, intensive efforts are needed to reach increasingly smaller segments of the population, and some informed individuals will elect not to undergo even well supported screening tests. Targeted efforts to increase screening in populations with lower screening rates have met with moderate success, but there is still much that we do not know about barriers to screening in such populations. Groups with lower rates of screening are unlikely to be homogeneous. While many nonusers are more likely than recent screeners to report barriers to accessing screening services,31 access is a necessary but insufficient predictor of screening. Even among individuals who have health insurance and a regular source of health care, screening rates could be improved. Individuals who have access to health care give many reasons for not being screened. Among the most important barriers are the absence of provider recommendation and a lack of awareness of the need to be screened regularly, both of which account for some proportion of underutilization.23, 34 Nonetheless, many individuals who have access to the health care system and receive provider recommendations to be screened still do not follow through. The fact is that we do not truly understand the meaning of responses such as, “I just didn't get around to it.” We also have not determined the proportion of the population that chooses not to participate in screening despite being informed of the associated benefits and harms. Rakowski and Breslau1 note the need to examine the way in which screening fits into an individual's daily life. Further efforts aimed at identifying barriers to screening among individuals who remain unscreened even when population rates are high and at effectively and cost-effectively promoting screening in such populations are required. Also necessary is additional societal discussion regarding the extent to which more intensive efforts to reach relatively small segments of the population are appropriate and cost effective as screening rates rise.
Should Informed Decision-Making, Rather than Screening Uptake, Be the Goal For All Screening Tests?35
Much discussion has occurred in recent years regarding the need to make ‘informed’ screening decisions. If informed decision-making (IDM) is the goal, should screening promotion efforts continue to be more persuasive for tests that can prevent disease or tests for which efficacy is known to be high and for which strong evidence reviews have been conducted? How do we assess the level at which to intervene? If, for example, a health plan covers and recommends a given screening test, then how will this affect an individual's decision-making? Do we know how receptive patients will be to communications regarding uncertainty? Are clinicians prepared to enter into a process of shared decision making (SDM)? Can we learn to use SDM more effectively? Can we expect an SDM model be the standard of care if clinicians still can be held accountable for undetected cancers in patients who make informed choices not to undergo screening?36 Rimer et al.23 state that because research on IDM as it relates to cancer screening is in its infancy, practices in this area still are evolving. Even if we agree that IDM/SDM is an appropriate goal, many questions remain regarding the practical implementation of this model.
What Type of Information Is Helpful to Patients for Making Informed Screening Decisions, and What Factors Influence What Is Communicated?
We still have much to learn about optimal methods for communicating the benefits and risks associated with screening and about the ways in which those communications ultimately affect a patient's intention to undergo screening. This complexity only will increase as screening strategies become more complex and tailored—i.e., as screening becomes tailored to levels of individual risk. We know that specialty training influences what information physicians deem to be important for their patients to know.37 Real and perceived time constraints frequently limit the ability of providers to discuss screening tests, let alone administer services.38 The practicality of adding even more tasks to the physician-patient encounter must be evaluated carefully. A significant challenge for researchers is the development of strategies that will facilitate communication between provider and patient without demanding more of the provider's time. This must be accomplished while creating an environment that enhances patient-provider communication.
How Are Interpersonal Factors, Such as Patient Satisfaction with Provider-Patient Encounters, Associated with Screening Practices?
One recent study39 found that rural residence and Hispanic ethnicity were not closely associated with women's screening practices. Rather, satisfaction with the quality of health care received influenced patients' screening behavior. Although the investigators were limited in their ability to assess which aspects of quality affected their findings, these results suggest that interpersonal factors (e.g., language concordance between provider and patient) can influence whether patients return for screening or follow-up.
A largely unaddressed but increasingly important issue, given the growing ethnic diversity of the United States population, is the need for culturally competent health care systems. In a recent systematic review, the National Task Force on Community Preventive Services addressed this topic and was unable to find sufficient evidence to determine the effectiveness of interventions aimed at improving access to culturally competent health care systems.40 The intervention approaches examined included studies on recruiting and retaining staff members who reflected the cultural diversity of the community served, using interpreter services for clients with limited English-language proficiency, providing cultural competency training for health services workers, using linguistically and culturally appropriate health education materials, and using culturally specific health care settings. Further research on the potential for these strategies to improve satisfaction with care is necessary.
Are Intervention Approaches That Have Been Successful in Improving Recent Uptake of Screening Tests Also Effective in Promoting Maintenance of Such Behavior?
All currently available cancer screening tests require repeated performance at regular intervals to achieve optimal population-level benefits related to morbidity and mortality. To date, research on adherence has focused largely on recent (single-endpoint) screening (e.g., mammography use within the past 2 years). Much less is known about the long-term effectiveness of interventions in maintaining screening behavior over time. Fewer individuals report having maintained such behavior than report having received the screening test recently.24 The correlates of repeat screening (mammography or Pap smear) appear to be similar, but not identical, to variables that frequently are correlated with recent screening or with any screening during an individual's lifetime.41, 42 Zapka and Lemon8 note that we must expand the focus of interventions to reflect not only the regularity of screening but also the continuum of care, including appropriate diagnostic, treatment-related, and surveillance-related services.
Which Strategies Hold the Greatest Potential for Improving Follow-Up of Abnormal Screening Results?
The development of cost-effective patient-level, provider-level, and practice-level strategies to improve follow-up of screen-detected abnormalities may reduce cancer morbidity and mortality. Surprisingly few studies, however, have dealt sufficiently with follow-up after abnormal findings. In their report, Bastani et al.9 emphasize that most intervention research efforts aimed at improving follow-up have used patient-directed strategies for women with abnormal findings on Pap smear. A recent study of clinical follow-up among African Americans after abnormal mammographic findings suggests that there is a need to address patient-related and health care system–related factors to improve timely follow-up.43 That study found no association between socioeconomic status and the timely resolution of findings. Rather, diagnostic resolution was associated with both the severity of the abnormality and with whether patients asked questions or received information regarding the next steps following screening. Women's prior experience with breast abnormalities also was predictive of a delay in follow-up. Interventions aimed at improving communication during the screening process, thus, have the potential to facilitate the timely resolution of abnormal findings.
Are Some Promotional Strategies More Cost-Effective than Others?
Data on the cost-effectiveness of screening promotion and follow-up are scant. Consequently, with few exceptions, we are unable to compare the costs of different interventions. It is possible to speculate that cost-effective interventions are more likely to be adopted compared with interventions that are not as cost effective; however, the extent to which cost will be a deciding factor in determining the allocation of resources for promotional efforts is unknown. In their report, Andersen et al.44 note that screening promotion will not be equally cost effective across population groups, because some groups will require more intensive intervention efforts. Thus, as screening intervention research evolves, so too will the measures used to assess cost-effectiveness. The incorporation of health-related quality of life44 into cost-effectiveness analyses and the investigation of ways to improve the economic efficiency of programs aimed at increasing IDM are but a few of the options to consider.
How Can We Effectively Exploit the Internet and Other Interactive Technologies To Address the Many Challenges Associated with Promoting Adherence to Screening?
Rapid advances in technology are creating exciting new opportunities to communicate information regarding cancer screening. In what ways can technology enhance screening-related education and decision-making? How can new communication technologies alleviate some of the face-to-face time constraints associated with the patient-provider encounter? How can we implement health care system–related interventions more effectively without burdening physicians and other staff?38 How many messages can be delivered simultaneously and to whom can they be delivered without overburdening the recipients?
How Does Screening Participation Affect Other Health Practices?
In one study, Miles et al.45 reported positive changes in smoking-, eating-, and exercise-related behaviors after colorectal cancer screening. Further investigation of this finding, as well as the apparent clustering of negative health behaviors in the never-screened group, is necessary.45 Similarly, Clark et al.46 and others have shown that smokers are less likely to undergo screening. Some data also indicate that overweight women are less likely to undergo mammography.47 To maximize changes in behavior, additional research focusing on behavioral clusters and how they can be understood is required.
How Much Does the Organization of a Cancer Screening Program Influence Participation Rates?
A review of screening intervention research conducted within the context of health care systems outside the U.S. shows the potential for learning from comparisons among countries.48 A major difference between the screening programs described by Miles et al.48 and those conducted in the U.S. is the way in which screening is organized; that is, whether the program is centrally organized or largely opportunistic (i.e., based on each individual's own schedule as it evolves with his or her provider). Because few studies have compared different systems of health care delivery, this area is ripe for investigation of the effects of different contexts on screening participation. For example, is the introduction of a new screening test accomplished more easily in a population in which screening is organized through both providers and the target population? Although the prevalence of recent screening may be similar across health care systems, is that also the case with respect to the maintenance of screening behavior? Assuming that organized screening increases screening uptake more effectively than opportunistic screening,49 which components should countries without central organization of screening attempt to emulate? Furthermore, assuming that opportunistic screening possesses advantages, how do countries with organized screening programs incorporate those advantages? Moreover, central organization of a screening program does not necessarily signify that a population is making informed decisions regarding screening or that behavioral and other interventions are not required to reach underusers of screening.
The lessons highlighted in the current supplement complement national efforts to improve cancer-related health outcomes. It is important to approach these initiatives in their totality to maximize progress in achieving desired objectives. The U.S. Department of Health and Human Services Healthy People 2010 (http://www.healthypeople.gov) program's goal of increasing the proportion of adults who receive breast, cervical, and colorectal cancer screening examinations will be realized only with continued efforts to improve screening compliance and follow-up. Toward this end, public health practitioners and decision-makers are encouraged to implement evidence-informed strategies, such as those recommended in the Guide to Community Preventive Services. Likewise, we urge national organizations charged with addressing the cancer burden, such as the ACS, the CDC, and the NCI, to partner and to provide coordinated leadership to support and implement the recommendations identified in this supplement. The lessons learned will be beneficial only if progress can be made through their application.
- 1Perspectives on behavioral and social science research on cancer screening. Cancer. 2004; 101 (5 Suppl ); xxx–xxx., .
- 5The society and population health reader: income inequality and health. New York: The New Press, 1999., , .
- 6http://www.nap.edu/books/0309071755/html [accessed Sep 9, 2003]., . Promoting health: intervention strategies from social and behavioral research [monograph online]. Available from URL:
- 7Unhealthy societies: the afflictions of inequality. London: Routledge, 1996..
- 8Interventions for patients, providers, and health care organizations. Cancer. 2004; 101(5 Suppl ): xxx–xxx., .
- 9Interventions to improve follow-up of abnormal findings in cancer screening. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , .
- 15Multilevel approaches to understanding social determinants. In: BerkmanLF, KawachiI, editors. Social epidemiology. New York: Oxford University Press, 2000: 349–367..
- 20Lessons learned from community-based cancer screening intervention research. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , .
- 21Some methodological lessons learned from cancer screening research. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , .
- 22Disseminating effective cancer screening interventions. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , .
- 23Informed decision making: what is its role in cancer screening? Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , , .
- 27Mammography use in the U.S.: trends and the impact of interventions. Ann Behav Med. 1994; 16: 317–326..
- 33Institute of Medicine. Speaking of health: assessing health communication strategies for diverse populations. Washington, DC: National Research Council, Institute of Medicine, 2002.
- 44Examining the cost-effectiveness of cancer screening promotion. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , .
- 48A perspective from countries using organized screening programs. Cancer. 2004; 101(5 Suppl ): xxx–xxx., , , .