- Email: [email protected]
Promoting Prostate-Specific Antigen Informed Decision-Making
Research Articles
Promoting Prostate-Specific Antigen Informed Decision-Making Evaluating Two Community-Level Interventions David L. Driscoll, PhD, MPH, Douglas J. Rupert, MPH, Carol E. Golin, MD, Lauren A. McCormack, PhD, MSPH, Stacey L. Sheridan, MD, MPH, Brandon M. Welch, MPH, Jon A. Poehlman, PhD Background: Most medical associations recommend that patients make informed decisions about whether to be screened for prostate cancer with the prostate-specific antigen (PSA) test. Studies assessing how to promote PSA informed decision-making (IDM) have been conducted almost exclusively in healthcare settings; there is a need for similar research in community settings. Methods:
This paper describes the results of a 5-year study (2002–2007) in which two communitylevel interventions were developed, implemented, and evaluated in matched upper- and lower-SES comparison communities in Greensboro and Wilmington, North Carolina. Both interventions promoted PSA informed decision-making. One intervention (PSA-Only) consisted of educational information about prostate cancer and the PSA test, and the other (Men’s Health) included additional information about recognizing and preventing heart attack, stroke, and colon cancer. Structured survey, semistructured interview, and structured observational data were combined to compare participating community residents’ pre/post changes in knowledge, intentions, and behaviors related to PSA IDM.
Results:
The community-level interventions successfully engaged community participants in discussions, educated individuals, encouraged deliberation of information, and facilitated PSA test discussions with physicians. Men who participated in the PSA-Only educational sessions were more likely than those who attended the Men’s Health educational sessions to discuss the PSA test with their physician (p⫽0.037).
Conclusions: When prospective SES-related confounding factors are matched across comparison communities, PSA IDM interventions can be shown to promote IDM. Framing the PSA test decision relative to less-ambiguous screening decisions does not appear to increase the likelihood of PSA IDM. (Am J Prev Med 2008;35(2):87–94) © 2008 American Journal of Preventive Medicine
Introduction
T
he clinical evidence is insufficient to determine whether screening and early treatment for prostate cancer reduce prostate cancer mortality, and the treatments currently available for prostate cancer can have serious permanent side effects, including impotence and urinary incontinence.1,2 Most medical organizations recommend that patients be given an opportunity to make informed decisions about whether to be screened for prostate cancer.3–5 Interventions to promote cancer screening informed decision-making (IDM) should (1) educate patients about the cancer and From RTI International (Driscoll, Rupert, McCormack, Welch, Poehlman), Research Triangle Park, North Carolina; and the Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill (Golin, Sheridan), Chapel Hill, North Carolina Address correspondence and reprint requests to: David Driscoll, PhD, MPH, RTI International, 3040 Cornwallis Road, P.O. Box 12194, Research Triangle Park NC 27709. E-mail: [email protected].
the benefits, risks, limitations, alternatives, and uncertainties of the screening test; (2) encourage patients to arrive at screening decisions consistent with their values and preferences; and (3) facilitate participation in the decision-making process at a level that is comfortable for the patient.6,7 Many interventions promoting IDM have been conducted in healthcare settings, and there is a need for greater research into such interventions in community settings.6 Community-level IDM interventions, or those that seek to promote IDM in community settings, can reach and engage individuals who are unlikely to participate in such interventions in healthcare settings because of income or class,8,9 knowledge or education,10 –12 or the complex integration of such individuallevel social disadvantages associated with health inequities.13,14 Further, interventions can be targeted in response to population-specific contextual factors that place some communities at greater risk than others. Neighborhood income, SES, and social integration
Am J Prev Med 2008;35(2) © 2008 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/08/$–see front matter doi:10.1016/j.amepre.2008.04.016
87
have been linked to the physical health status, testing behaviors, and cancer survival rates of residents regardless of individual SES.15–18 Community-level interventions also allow residents an opportunity to deliberate on important health prevention decisions in a familiar social environment outside the unusual context and time constraints of the clinical encounter.16,17,19 Community-level interventions may be particularly effective in promoting prostate-specific antigen (PSA) IDM. African-American men and older men have higher levels of prostate cancer mortality1 and are the least likely to have healthcare access,20 least likely to be aware of the PSA test,21–24 and least prepared to discuss the PSA test with a physician during a clinical visit.23,25,26 Community interventions may be particularly necessary to prepare African-American men for clinical discussions because they may distrust their healthcare provider’s suggestions not to receive the PSA test.17,21,23 Finally, many community health fairs are offering PSA tests, and community interventions can inform men about factors that play a role in the PSA decision and can promote IDM prior to these screening opportunities. Community-level interventions can be challenging to evaluate. First, participants in community-level interventions are usually not randomly selected, and quasiexperimental evaluation designs involving matched comparison communities are particularly vulnerable to selection bias.17,27 Such evaluation designs can help reduce such bias by matching comparison communities on key risk factors related to the evaluation outcomes.28,29 Second, evaluations of community-level interventions often encounter only modest outcome effects compared to individual-level interventions.30 Evaluations of community-level interventions should assess changes in intermediate measures of effect, including attitudes and knowledge, and such distal measures as the collaborative nature of the clinical decision-making process.30 This paper evaluates the process and outcomes of two community-level interventions to promote PSA IDM developed and conducted in Greensboro and Wilmington, North Carolina over a 5-year period (2002 and 2007). The overarching objective is to evaluate how effectively the interventions informed, encouraged, and facilitated PSA IDM. A secondary objective is to describe the relative utility of the different intervention formats and the effects of SES on PSA IDM.
upper- and lower-SES communities, were selected based on these attributes in a case-specific multidimensional construct as recommended by Braveman et al.8 The process began by collecting these secondary data at the census-tract level in each city from the U.S. Census Bureau. Although the racial or ethnic population distribution within each census tract was not a community selection criterion, the ratio of African-American to European American households was generally comparable between matched census tracts, with a majority of households in the lower-SES communities being African American. Next, maps of matching census tracts in each city were carefully plotted and systematically assessed by pairs of project team members. Team members collected structured observations and conducted informal, intercept interviews with community residents to identify and map community boundaries and organizations in a rapid assessment process.31 Finally, formal structured interviews were conducted with leaders of selected organizations in these communities to collect data on the mission, membership, and willingness of each organization to host communitybased interventions related to PSA IDM. Two sets of intervention formats were employed to inform, encourage, and facilitate PSA IDM. One intervention (PSAOnly) provided information only about prostate cancer and the PSA test, while the second (Men’s Health) included information about screening for and preventing heart attack, stroke, and colon cancer. Prior research has suggested that starting with less-ambiguous screening messages, such as those regarding heart attack, stroke, and colon cancer, might overcome cognitive dissonance with the message that some men might reasonably choose not to have a PSA test for prostate cancer.32 The intervention community sampling design consisted of four cells (Figure 1). One upper-SES community and one lower-SES community with similar levels of community infrastructure were selected in each city. Each intervention community averaged 4600 residents. Residents of the two Wilmington intervention communities received the PSA-Only educational session, and residents of the two Greensboro intervention communities received the Men’s Health educational session. Assigning all organizations in each metropolitan area to the same intervention reduced the potential for confounding due to area word-of-mouth or local media coverage of the educational sessions. Twenty educational sessions took place with 18 community organizations between September 2004 and February 2005.
Methods Socioeconomic characteristics may be as strongly associated with individual awareness of complex health issues and willingness to adopt health protective behaviors as race or ethnicity.9,10 The intervention communities selected for this study were matched on three measures of SES: median household income, percentage of resident-owned homes, and median household educational attainment. Two sets of matched intervention communities, hereafter referred to as
88
Figure 1. Community sampling design. PSA, prostate specific antigen; SES, socioeconomic status
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
No fewer than 10 and no more than 32 eligible community residents attended each session. Intervention participants were recruited through and by community organization members. All residents who wished to do so could attend the sessions, but only men aged between 40 and 80 years and not previously diagnosed with prostate cancer were invited to participate in the study. These interventions were conducted primarily in the evenings when people were not at work and had more flexible schedules. The educational sessions were held at conveniently located and familiar facilities like churches, fraternal organizations, and YMCAs, and each organization received $250 for each session hosted. While multiple educational sessions took place in each community, participants were screened to prevent duplicate responses, and project team members recall no eligible session attendees declining to participate in the study, making it unlikely that any participants received repeated exposures to the intervention. At least three project team members conducted each educational session. First, participants heard a 5- to 10-minute scripted oral presentation delivered by a local physician and study team member. The presentation included a 4= by 6= poster about prostate cancer or men’s health issues, and an 8.5 ⫻ 11-inch copy of the poster was provided to each participant. Following the presentation, participants viewed a 20-minute video in which age- and race-appropriate men discussed and made decisions about the PSA test. Finally, participants asked questions of the presenting physician and participated in moderated group discussions of the featured health issues and decisions. Eligible participants completed a structured survey prior to the community educational sessions (n⫽361), immediately following the community educational sessions (n⫽339), and via standard mail approximately 6 and 12 months postsession (n⫽274 and n⫽254, respectively). All respondents who were eligible and completed the baseline survey were given the opportunity to respond at 6 and 12 months. The 6- and 12-month response rates for the PSA-Only participants were 85% and 72%, respectively; the response rates for the Men’s Health participants were 71% and 72%, respectively. Response rates were calculated using the American Association for Public Opinion Research methodology, and nonrespondents included those who did not return surveys, who were unable to be reached, and who were deemed ineligible because of age or cancer history after baseline. Men who dropped out were more likely to be African American, less educated, younger, and not married/marital status unknown (p⬍0.05). No differences were found in self-reported health status of those who dropped out and those who did not. Semistructured, in-depth interviews were conducted with a total of 47 participants—23 at 6 months and 24 at 12 months postsession. In-depth interview participants were randomly selected from a pool of respondents whose responses to the survey questions on PSA knowledge and preferred level of involvement in the PSA decision were in the top and bottom quartiles on those survey items. The semistructured interviews took place in community organizations and lasted approximately 60 minutes. Finally, team members employed an observational guide that included a description of the issues discussed by participants at the conclusion of the educational session. Observational data for community identification were
August 2008
analyzed in 2004, and evaluation data analysis (reported here) was completed in 2007. Univariate and multivariate statistical analyses were conducted to test for significant differences in structured survey responses across time points. For single-item analyses, McNemar’s test was used when comparing time-related proportions, and Pearson’s chi-square tests were conducted when comparing responses at single time points. Generalized estimating equations33 that account for correlated responses were used to test time effects on key outcomes while controlling for demographic variables. Qualitative semistructured interview responses were analyzed using an ordered metamatrix34 in which responses were displayed graphically and emergent themes across and within respondents were identified and assessed in an iterative analytic process. Structured observations were reviewed in a debriefing session with all participating project staff immediately following each intervention. Three categories of mixed quantitative and qualitative data were analyzed to assess and compare the community-based educational sessions. The first category consisted of pre- and post-intervention responses to survey questions about prostate cancer, the PSA test, and the clinical options available. These data were augmented with in-depth interview questions assessing participants’ recall of the interventions. The second category of data consisted of pre- and post-intervention responses to survey questions on participants’ preferred role in the PSA decision-making process and post-intervention structured observations of group discussions about the decision to have a PSA test. These data were augmented with in-depth interview questions assessing participants’ perspectives and preferences related to screening decision-making. Finally, the third category of data consisted of post-intervention responses to survey questions on the extent to which men discussed the PSA test and shared their feelings about the PSA with their physicians. These data were augmented with in-depth interview questions assessing how well the interventions facilitated men’s participation in the PSA test decision-making process.
Results The community-matching procedure provided two distinct socioeconomic categories of community residents. Participants from the lower-SES and upperSES communities were statistically distinct on every demographic variable except health insurance, primarily because many lower-SES residents had publicly funded health insurance. No significant differences were found between the educational levels or household incomes of study participants within the matched communities across the two metropolitan areas (Table 1). These data were consistent with the tract-level census data, which indicate that the selfselected sample of community residents who participated in the study was representative of census-tract residents on these variables. A considerable proportion of lower-SES community residents had demographic characteristics, such as low income, low education, and publicly funded health insurance, that have been shown to be barriers to accessing Am J Prev Med 2008;35(2)
89
Table 1. Participant demographics by community and SES level (%) Variable Educationa High school or less Some college College graduate or higher Household income ($)a ⬍20,000 20,000–39,999 40,000–59,999 ⱖ60,000 Racea,b Black White Other Insurance status Medicare Medicaid Employer-based Veterans Affairs Military Prescription coverage None Personal doctora Yes a
Greensboro lower SES (nⴝ90)
Wilmington lower SES (nⴝ29)
Greensboro upper SES (nⴝ146)
Wilmington upper SES (nⴝ96)
21.6 41.9 36.5
22.7 36.4 40.9
17.9 14.2 67.9
12.2 26.7 61.1
7.8 17.8 24.4 28.9
3.5 17.2 13.8 34.5
2.7 13.7 20.6 49.3
2.1 17.7 26.1 38.5
65.6 16.7 4.1
37.9 34.5 0.0
13.7 78.1 1.5
11.5 83.3 2.0
44.9 27.0 47.8 16.7 11.1 3.3 5.6
45.5 13.0 65.5 10.3 10.3 3.5 3.5
54.1 9.7 69.2 6.9 4.1 3.4 4.1
67.7 8.7 60.4 10.4 11.5 7.3 3.1
73.6
90.9
85.1
90.2
Difference between lower-SES and upper-SES communities’ p⬍0.05 27% of respondents in the Wilmington lower-SES community refused to respond to the questions on race/ethnicity.
b
health care or participating in IDM.8,10,11,22–27 Approximately 22% of lower-SES community participants had a high school education or less, and almost 25% had annual household incomes of less than $40,000 (Table 1). Almost 20% of the lower-SES respondents in both metropolitan areas had no personal doctor, and more than 12% of upper-SES study participants also did not have a personal doctor.
Educating Community Participants The educational sessions resulted in a marked increase in accurate responses to survey questions about prostate cancer and the PSA test immediately after the educational session, which then declined slightly over the following year. However, knowledge levels were higher at the 12-month follow-up than at baseline (Table 2). Several survey items reveal improvements in participant knowledge related to prostate cancer and the PSA. First, study participants became more knowledgeable about treatment for early-stage prostate cancer. At baseline, only 13% of participants recognized that doctors do not know which treatment is best for early-stage prostate cancer; at the 12-month follow-up, 30% of participants answered accurately. Second, participants retained knowledge regarding the side effects of prostate cancer treatment: 10% of participants at baseline and 22% at the 12-month follow-up knew that about 5 of 10 men would expe90
rience permanent side effects. Third, 59% of participants at baseline and 73% at the 12-month follow-up knew that most prostate cancers are slow growing. Finally, participants learned that prostate cancer can occur even if PSA levels are normal. The proportion of participants who knew this fact rose from 56% at baseline to 72% at 6 months; the proportion then declined to 63% at 12 months. There were no significant differences in knowledge gained between recipients of the two educational session formats. Residents of the upper-SES communities in both metropolitan areas had higher scores on these knowledge items prior to the community sessions, but this disparity disappeared following the community sessions (Figure 2). In-depth interviews with residents scoring highest and lowest on the knowledge items augmented the survey data by revealing which key messages study participants recalled in the absence of structured response options. All interview participants, including those with the lowest knowledge item scores, recalled that most prostate cancers are slow growing, that half of men who receive treatment will have permanent side effects, and that men should make an informed decision about whether to get the PSA test.
Encouraging Deliberation The educational sessions encouraged participant deliberation on the PSA decision. At the conclusion of the
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
Table 2. Participant knowledge of prostate cancer and the PSA test (%) Knowledge variable Most prostate cancers are slow growing Truea False Don’t know Prostate cancer can occur even if PSA levels are normal Truea False Don’t know Which treatment works best for prostate cancer Surgery Radiation Surgery and radiation work equally well Doctors don’t knowa I don’t know Out of 10 men, how many will have permanent side effects from prostate cancer treatment None About 2 About 5a All 10 Don’t know
Baseline (nⴝ361)
Post-intervention (nⴝ339)
6-month follow-up (nⴝ274)
12-month follow-up (nⴝ254)
59.0 8.4 32.7
74.2 11.3 14.6
74.4 8.2 17.4
72.6 5.9 21.5
56.0 4.7 39.3
78.4 4.9 16.8
71.9 6.1 22.0
62.5 7.0 30.5
13.3 16.5 11.4 13.3 45.5
5.8 3.7 13.5 62.3 14.7
8.5 11.6 11.1 39.3 29.5
10.0 10.8 13.8 30.3 35.1
2.6 16.3 9.6 3.5 68.1
— 19.8 54.1 3.0 23.1
0.8 18.9 27.6 5.9 46.8
0.5 17.3 21.9 6.5 53.8
Note: The — symbol indicates that data were not collected for the variable at the time point. a Indicates accurate response option PSA, prostate-specific antigen
presentation, participants asked questions of the physician and participated in a group discussion on issues related to prostate cancer, the PSA test, and the treatment options for prostate cancer. The discussed questions and issues included the risk factors for fastgrowing prostate cancer, the current state of the science in preventing the development or spread of the disease, likely scenarios during the patient–physician discussion of the PSA test, and prospective outcomes of the various decision options currently available. Many participants applied the information provided in the presentation to describe their decision-making process. At the end of each session, many participants described the concluding discussion as the session’s most informative component.
The education sessions encouraged deliberation on the desired level of participation in the PSA decision. Participants generally preferred a more active role in subsequent doctor–patient deliberations on the PSA decision: 33% of participants at baseline wanted to make the final PSA decision compared with 40% of participants at the 12-month follow-up. Likewise, 46% of participants at baseline preferred to make a joint decision with their physician compared with 50% at the 12-month follow-up. There were no significant differences in desired level of participation in the PSA testing decision-making between recipients of the different intervention formats or the upper- and lowerSES communities. In-depth interviews highlighted key differences between residents in the top quartile of preferred level of involvement in the PSA decision. Most men who responded to the survey question that they preferred making the final decision themselves expressed an interest in hearing their physician’s thoughts about the PSA test. However, some of those in the upper quartile of responses expressed the opinion that there was no PSA decision to discuss with their physicians. These men strongly stated an intention to continue receiving the test.
Facilitating Participation in Decision-Making
Figure 2. Knowledge-level changes by socioeconomic status
August 2008
Both educational session formats led the majority of men to discuss the PSA test and share their feelings about the test with their physicians. By the 12-month Am J Prev Med 2008;35(2)
91
Table 3. Participant intentions and behaviors regarding the PSA test by program content (%) PSA-Only content Variable PSA test intention Plan to have test within next year Do not plan to have test within next year Don’t know Preferred PSA involvement Doctor decision Doctor decision with patient input Joint doctor–patient decision Patient decision with doctor input Patient decision Discussed PSA test with physiciana Yes No Shared feelings about PSA test with physicianb Yes, a lot Yes, a little No
Men’s Health content
Baseline
6 months
12 months
Baseline
6 months
12 months
75.6 10.1 14.3
— — —
59.3 17.4 23.3
66.4 16.2 17.5
— — —
53.8 15.0 31.3
2.6 19.7 47.9 28.2 1.7
3.8 1.9 40.6 50.0 3.8
— 5.8 54.0 36.8 3.5
10.4 10.4 45.1 30.2 4.1
4.8 8.4 42.2 38.0 6.6
5.6 6.8 47.5 34.0 6.2
— —
44.8 54.3
64.0 33.7
— —
40.6 58.8
55.5 44.5
— —
27.1 33.3 39.6
19.6 41.1 39.3
— — —
21.9 37.5 40.6
18.9 37.8 43.3
Note: The — symbol indicates that data were not collected for the variable at the time point. a Indicates significant difference (p⬍0.05) between PSA-Only and Men’s Health formats b Asked only of respondents who discussed PSA test with physician PSA, prostate-specific antigen
follow-up, 59% of men had discussed the PSA decision with their physicians, and 58% had shared their feelings to some extent about the test during that discussion. Men were equally likely to discuss the PSA test with a physician regardless of whether they had visited a healthcare professional in the year before the intervention. Men who received the PSA-Only format were more likely to discuss the PSA test with their physicians (p⫽0.037) (Table 3). There were no differences between upper- and lower-SES community residents on these variables 12 months after the interventions. Subsequent in-depth interviews with men who had discussed the PSA with their physician revealed several factors constraining PSA IDM. The most frequently mentioned constraint was the limited time available during the clinical encounter to sufficiently discuss the complex PSA decision. Even when given sufficient time to discuss the PSA decision, many patients felt IDM was limited by physicians’ reluctance to discuss the lack of medical certainty about the efficacy of the PSA or current treatment options for reducing mortality from prostate cancer. The limited sample of in-depth interview data did not allow for statistical analysis of differences in constraints experienced between intervention community participants. Although the education sessions did not direct men to a particular decision, it did facilitate consideration of the decision to have the PSA test. Participants were generally less likely to plan on receiving a PSA test in the year after the interventions. At baseline, 70% of participants intended to receive a PSA test compared with only 56% of participants at the 12-month follow-up. 92
Moreover, the percentage of men who were unsure whether they wanted to have a PSA test rose from 16% at baseline to 29% at the 12-month follow-up. The interventions, however, had little effect on actual testing behaviors (Table 3). At baseline, 61% of men had received a PSA test within the past year, as had 63% of men at the 12-month follow-up.
Discussion In this evaluation of two community-level interventions promoting IDM for prostate cancer screening, both interventions increased knowledge necessary for IDM to take place, encouraged the application of this knowledge in considering values and preferences related to the PSA, and were generally successful in facilitating patient–physician PSA discussions. Men who participated in the PSA-Only educational sessions were more likely to discuss the PSA decision with their physicians. This finding suggests that framing the PSA test decision in the context of less-ambiguous screening decisions is unnecessary to promote patient participation in the decision-making process. The community-selection process employed in this study identified and engaged many residents of lowerSES communities who would have been unlikely to receive such interventions in healthcare settings. The inclusion of both lower- and upper-SES communities also revealed an association between SES and the knowledge necessary to participate in IDM related to the PSA screening decision. Several previous studies have revealed an association between elements of SES at the individual or community level and health status,
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
testing behaviors, and cancer survival rates.8 –18 This study demonstrated that community-level educational sessions can reduce disparities in the knowledge necessary for PSA IDM among lower- and upper-SES community residents. Community-level interventions have been successful in promoting deliberation on the prostate cancer screening decision,18,22 and the interventions evaluated in this study were likewise effective. The community interventions encouraged men to apply the knowledge gained in the interventions in considering their values and preferences related to the PSA test. The interactive nature of the session allowed community residents to discuss the PSA decision with peers and neighbors. If one views patients’ values as socially derived and shared beliefs about acceptable and desirable choices and behaviors,35 the collaborative nature of these sessions gave community residents an opportunity to deliberate on how shared values and preferences related to the PSA decision. The qualitative and observational data provide important insights into the outcomes of the educational sessions. All study participants, including those who scored lowest on the structured knowledge questions, were able to recall key elements of the educational sessions 12 months after the interventions without the benefit of prompts or the closeended response fields commonly employed in structured surveys. In-depth interview data also revealed that the most common constraint on PSA IDM was insufficient time in the patient–physician encounter to discuss the PSA test decision. This finding emphasizes the importance of community-level interventions in which patients are provided sufficient time, and support from family and friends, to deliberate on a clinical decision. We thank the CDC and the Association for Prevention Teaching and Research for their financial support under Cooperative Agreement No. U50/CCU300860, Project TS0845. The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of CDC or the Association for Prevention Teaching and Research. We also thank the following local physicians and project team members for their leadership and commitment to community-based health promotion and disease prevention: Dr. Cedric Bright, Old North State Medical Society; Dr. Sam Cykert, Moses H. Cone Memorial Hospital; Dr. Jorge Gonzalez, New Hanover Regional Medical Center; and Dr. Joseph Pino, New Hanover Regional Medical Center. No financial disclosures were reported by the authors of this paper.
References 1. National Cancer Institute. Screening for prostate cancer. 2006. Available at: http://www.cancer.gov/cancerinfo/pdq/screening/prostate/ healthprofessional.
August 2008
2. U.S. Preventive Services Task Force (USPSTF). Screening for prostate cancer: recommendation and rationale. Ann Intern Med 2002;137: 915– 6. 3. Harris R, Lohr KN. Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137:917–29. 4. Chan EC, Vernon SW, O’Donnell FT, Ahn C, Greisinger A, Aga DW. Informed consent for cancer screening with prostate-specific antigen: how well are men getting the message? Am J Public Health 2003;93: 779 – 85. 5. Sheridan SL, Harris RP, Woolf SH, and the Shared Decision-Making Workgroup of the U.S. Preventive Services Task Force. Shared decisionmaking about screening and chemoprevention: a suggested approach from the U.S. Preventive Services Task Force. Am J Prev Med 2004;26:56 – 66. 6. Briss P, Rimer B, Reilley B, et al. Promoting informed decisions about cancer screening in communities and healthcare systems. Am J Prev Med 2004;26:67– 80. 7. Rimer BK, Briss PA, Zeller PK, Chan EC, Woolf SH. Informed decisionmaking: what is its role in cancer screening? Cancer 2004;101 (Suppl 5):1214 –28. 8. Braveman PA, Cubbin C, Egerter S, et al. Socioeconomic status in health research: one size does not fit all. JAMA 2005;294:2879 – 88. 9. Isaacs SL, Schroeder SA. Class: the ignored determinant of the nation’s health. N Engl J Med. 2004;351:1137– 42. 10. Viswanath K, Breen N, Meissner H, et al. Cancer knowledge and disparities in the information age. J Health Commun 2006;11:1–17. 11. Ross CE, Van Willigen M. Education and the subjective quality of life. J Health Soc Behav 1997;38:275–97. 12. Ross CE, Wu C. The links between education and health. Am Sociol Rev 1995;60:719 – 45. 13. Braveman P. Health disparities and health equity: concepts and measurement. Annu Rev Public Health 2006;27:18.1–18.28. 14. Braveman P, Egerter S, Cubbin C, Marchi K. An approach to studying social disparities in health and health care. Am J Public Health 2004;94:2139 – 48. 15. Feldman PJ, Steptoe A. How neighborhoods and physical functioning are related: the roles of neighborhood socioeconomic status, perceived neighborhood strain, and individual health risk factors. Ann Behav Med 2004;27:91–9. 16. Boyd C, Zhang-Salomons JY, Groome PA, Mackillop WJ. Associations between community income and cancer survival in Ontario, Canada, and the United States. J Clin Oncol 1999;17:2244 –55. 17. Bridge PD, Berry-Bobovski LC, Gallagher RE. Promoting informed decision-making: evaluating a community-based prostate health awareness program. J Cancer Educ 2004;19:174 –9. 18. Reyes-Ortiz CA, Goodwin JS, Freeman JL, Kuo YF. Socioeconomic status and survival in older patients with melanoma. J Am Geriatr Soc 2006;54:1758 – 64. 19. Sikkema KJ, Kelly JA, Winett RA, et al. Outcomes of a randomized community-level HIV prevention intervention for women living in 18 low-income housing developments. Am J Public Health 2000;90:57– 63. 20. Myers RE. African-American men, prostate cancer early detection examination use, and informed decision-making. Semin Oncol 1999;26:375– 81. 21. Barber KR, Shaw R, Folts M, et al. Differences between African American and Caucasian men’s participation in a community-based prostate cancer screening program. J Community Health 1998;23:441–51. 22. Richardson JT, Webster JD, Fields NJ. Uncovering myths and transforming realities among low-SES African-American men: implications for reducing prostate cancer disparities. J Natl Med Assoc 2004;96:1295–302. 23. Gwede CK, Forthofer MS, McDermott RJ. Awareness of prostate cancer screening tests among African American men in underserved neighborhoods: need for community outreach. J Oncol Manag 2002;11:36 – 41. 24. Demark-Wahnefried W, Strigo T, Catoe K, et al. Knowledge, beliefs, and prior screening behavior among Blacks and Whites reporting for prostate cancer screening. Urology 1995;46:346 –51. 25. McCoy CB, Anwyl RS, Metsch LR, Inciardi JA, Smith SA, Correa R. Prostate cancer in Florida: knowledge, attitudes, practices, and beliefs. Cancer Pract 1995;3:88 –93. 26. O’Dell KJ, Volk RJ, Cass AR, Spann SJ. Screening for prostate cancer with the prostate-specific antigen test: are patients making informed decisions? J Fam Pract 1999;48:682– 8. 27. Allen JD, Kennedy M, Wilson-Glover A, Gilligan TD. African-American men’s perceptions about prostate cancer: implications for designing educational interventions. Soc Sci Med 2007;64:2189 –200.
Am J Prev Med 2008;35(2)
93
28. Merzel C, D’Afflitti J. Reconsidering community-based health promotion: promise, performance, and potential. Am J Public Health 2003;93:557–74. 29. Farquahar S, Wing S. Methodological and ethical considerations in community-driven environmental justice research. In: Minkler M, Wallerstein N, eds. Community-based participatory research for health. San Francisco: Jossey-Bass, 2003:221– 41. 30. Fishbein M. Great expectations, or do we ask too much from communitylevel interventions? Am J Public Health 1996;86:1075– 6. 31. Beebe J. Rapid assessment process: an introduction. Walnut Creek CA: Altamira Press, 2001.
32. Driscoll D, Harris-Kojetin L. Final patient messages for prostate-specific antigen (PSA) screening for prostate cancer, Medicare Screening Messages Project. Final report to AHRQ/CMS. Research Triangle Park NC: RTI International, 2002. 33. Liang KY, Zeger SL. Longitudinal data analysis using general linear models. Biometrika 1986;73:13–22. 34. Miles M, Huberman MA. Qualitative data analysis. 2nd ed. Thousand Oaks CA: Sage Publications, 1994. 35. Kluckhohn CK. Values and value orientations in the theory of action. In: Parsons T, Shils EA, eds. Toward a General Theory of Action. Cambridge MA: Harvard University Press, 1951.
Did you know? You can access back issues of American Journal of Preventive Medicine online. Visit www.ajpm-online.net today to see what’s new online!
94
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
Promoting Prostate-Specific Antigen Informed Decision-Making Evaluating Two Community-Level Interventions David L. Driscoll, PhD, MPH, Douglas J. Rupert, MPH, Carol E. Golin, MD, Lauren A. McCormack, PhD, MSPH, Stacey L. Sheridan, MD, MPH, Brandon M. Welch, MPH, Jon A. Poehlman, PhD Background: Most medical associations recommend that patients make informed decisions about whether to be screened for prostate cancer with the prostate-specific antigen (PSA) test. Studies assessing how to promote PSA informed decision-making (IDM) have been conducted almost exclusively in healthcare settings; there is a need for similar research in community settings. Methods:
This paper describes the results of a 5-year study (2002–2007) in which two communitylevel interventions were developed, implemented, and evaluated in matched upper- and lower-SES comparison communities in Greensboro and Wilmington, North Carolina. Both interventions promoted PSA informed decision-making. One intervention (PSA-Only) consisted of educational information about prostate cancer and the PSA test, and the other (Men’s Health) included additional information about recognizing and preventing heart attack, stroke, and colon cancer. Structured survey, semistructured interview, and structured observational data were combined to compare participating community residents’ pre/post changes in knowledge, intentions, and behaviors related to PSA IDM.
Results:
The community-level interventions successfully engaged community participants in discussions, educated individuals, encouraged deliberation of information, and facilitated PSA test discussions with physicians. Men who participated in the PSA-Only educational sessions were more likely than those who attended the Men’s Health educational sessions to discuss the PSA test with their physician (p⫽0.037).
Conclusions: When prospective SES-related confounding factors are matched across comparison communities, PSA IDM interventions can be shown to promote IDM. Framing the PSA test decision relative to less-ambiguous screening decisions does not appear to increase the likelihood of PSA IDM. (Am J Prev Med 2008;35(2):87–94) © 2008 American Journal of Preventive Medicine
Introduction
T
he clinical evidence is insufficient to determine whether screening and early treatment for prostate cancer reduce prostate cancer mortality, and the treatments currently available for prostate cancer can have serious permanent side effects, including impotence and urinary incontinence.1,2 Most medical organizations recommend that patients be given an opportunity to make informed decisions about whether to be screened for prostate cancer.3–5 Interventions to promote cancer screening informed decision-making (IDM) should (1) educate patients about the cancer and From RTI International (Driscoll, Rupert, McCormack, Welch, Poehlman), Research Triangle Park, North Carolina; and the Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill (Golin, Sheridan), Chapel Hill, North Carolina Address correspondence and reprint requests to: David Driscoll, PhD, MPH, RTI International, 3040 Cornwallis Road, P.O. Box 12194, Research Triangle Park NC 27709. E-mail: [email protected].
the benefits, risks, limitations, alternatives, and uncertainties of the screening test; (2) encourage patients to arrive at screening decisions consistent with their values and preferences; and (3) facilitate participation in the decision-making process at a level that is comfortable for the patient.6,7 Many interventions promoting IDM have been conducted in healthcare settings, and there is a need for greater research into such interventions in community settings.6 Community-level IDM interventions, or those that seek to promote IDM in community settings, can reach and engage individuals who are unlikely to participate in such interventions in healthcare settings because of income or class,8,9 knowledge or education,10 –12 or the complex integration of such individuallevel social disadvantages associated with health inequities.13,14 Further, interventions can be targeted in response to population-specific contextual factors that place some communities at greater risk than others. Neighborhood income, SES, and social integration
Am J Prev Med 2008;35(2) © 2008 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/08/$–see front matter doi:10.1016/j.amepre.2008.04.016
87
have been linked to the physical health status, testing behaviors, and cancer survival rates of residents regardless of individual SES.15–18 Community-level interventions also allow residents an opportunity to deliberate on important health prevention decisions in a familiar social environment outside the unusual context and time constraints of the clinical encounter.16,17,19 Community-level interventions may be particularly effective in promoting prostate-specific antigen (PSA) IDM. African-American men and older men have higher levels of prostate cancer mortality1 and are the least likely to have healthcare access,20 least likely to be aware of the PSA test,21–24 and least prepared to discuss the PSA test with a physician during a clinical visit.23,25,26 Community interventions may be particularly necessary to prepare African-American men for clinical discussions because they may distrust their healthcare provider’s suggestions not to receive the PSA test.17,21,23 Finally, many community health fairs are offering PSA tests, and community interventions can inform men about factors that play a role in the PSA decision and can promote IDM prior to these screening opportunities. Community-level interventions can be challenging to evaluate. First, participants in community-level interventions are usually not randomly selected, and quasiexperimental evaluation designs involving matched comparison communities are particularly vulnerable to selection bias.17,27 Such evaluation designs can help reduce such bias by matching comparison communities on key risk factors related to the evaluation outcomes.28,29 Second, evaluations of community-level interventions often encounter only modest outcome effects compared to individual-level interventions.30 Evaluations of community-level interventions should assess changes in intermediate measures of effect, including attitudes and knowledge, and such distal measures as the collaborative nature of the clinical decision-making process.30 This paper evaluates the process and outcomes of two community-level interventions to promote PSA IDM developed and conducted in Greensboro and Wilmington, North Carolina over a 5-year period (2002 and 2007). The overarching objective is to evaluate how effectively the interventions informed, encouraged, and facilitated PSA IDM. A secondary objective is to describe the relative utility of the different intervention formats and the effects of SES on PSA IDM.
upper- and lower-SES communities, were selected based on these attributes in a case-specific multidimensional construct as recommended by Braveman et al.8 The process began by collecting these secondary data at the census-tract level in each city from the U.S. Census Bureau. Although the racial or ethnic population distribution within each census tract was not a community selection criterion, the ratio of African-American to European American households was generally comparable between matched census tracts, with a majority of households in the lower-SES communities being African American. Next, maps of matching census tracts in each city were carefully plotted and systematically assessed by pairs of project team members. Team members collected structured observations and conducted informal, intercept interviews with community residents to identify and map community boundaries and organizations in a rapid assessment process.31 Finally, formal structured interviews were conducted with leaders of selected organizations in these communities to collect data on the mission, membership, and willingness of each organization to host communitybased interventions related to PSA IDM. Two sets of intervention formats were employed to inform, encourage, and facilitate PSA IDM. One intervention (PSAOnly) provided information only about prostate cancer and the PSA test, while the second (Men’s Health) included information about screening for and preventing heart attack, stroke, and colon cancer. Prior research has suggested that starting with less-ambiguous screening messages, such as those regarding heart attack, stroke, and colon cancer, might overcome cognitive dissonance with the message that some men might reasonably choose not to have a PSA test for prostate cancer.32 The intervention community sampling design consisted of four cells (Figure 1). One upper-SES community and one lower-SES community with similar levels of community infrastructure were selected in each city. Each intervention community averaged 4600 residents. Residents of the two Wilmington intervention communities received the PSA-Only educational session, and residents of the two Greensboro intervention communities received the Men’s Health educational session. Assigning all organizations in each metropolitan area to the same intervention reduced the potential for confounding due to area word-of-mouth or local media coverage of the educational sessions. Twenty educational sessions took place with 18 community organizations between September 2004 and February 2005.
Methods Socioeconomic characteristics may be as strongly associated with individual awareness of complex health issues and willingness to adopt health protective behaviors as race or ethnicity.9,10 The intervention communities selected for this study were matched on three measures of SES: median household income, percentage of resident-owned homes, and median household educational attainment. Two sets of matched intervention communities, hereafter referred to as
88
Figure 1. Community sampling design. PSA, prostate specific antigen; SES, socioeconomic status
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
No fewer than 10 and no more than 32 eligible community residents attended each session. Intervention participants were recruited through and by community organization members. All residents who wished to do so could attend the sessions, but only men aged between 40 and 80 years and not previously diagnosed with prostate cancer were invited to participate in the study. These interventions were conducted primarily in the evenings when people were not at work and had more flexible schedules. The educational sessions were held at conveniently located and familiar facilities like churches, fraternal organizations, and YMCAs, and each organization received $250 for each session hosted. While multiple educational sessions took place in each community, participants were screened to prevent duplicate responses, and project team members recall no eligible session attendees declining to participate in the study, making it unlikely that any participants received repeated exposures to the intervention. At least three project team members conducted each educational session. First, participants heard a 5- to 10-minute scripted oral presentation delivered by a local physician and study team member. The presentation included a 4= by 6= poster about prostate cancer or men’s health issues, and an 8.5 ⫻ 11-inch copy of the poster was provided to each participant. Following the presentation, participants viewed a 20-minute video in which age- and race-appropriate men discussed and made decisions about the PSA test. Finally, participants asked questions of the presenting physician and participated in moderated group discussions of the featured health issues and decisions. Eligible participants completed a structured survey prior to the community educational sessions (n⫽361), immediately following the community educational sessions (n⫽339), and via standard mail approximately 6 and 12 months postsession (n⫽274 and n⫽254, respectively). All respondents who were eligible and completed the baseline survey were given the opportunity to respond at 6 and 12 months. The 6- and 12-month response rates for the PSA-Only participants were 85% and 72%, respectively; the response rates for the Men’s Health participants were 71% and 72%, respectively. Response rates were calculated using the American Association for Public Opinion Research methodology, and nonrespondents included those who did not return surveys, who were unable to be reached, and who were deemed ineligible because of age or cancer history after baseline. Men who dropped out were more likely to be African American, less educated, younger, and not married/marital status unknown (p⬍0.05). No differences were found in self-reported health status of those who dropped out and those who did not. Semistructured, in-depth interviews were conducted with a total of 47 participants—23 at 6 months and 24 at 12 months postsession. In-depth interview participants were randomly selected from a pool of respondents whose responses to the survey questions on PSA knowledge and preferred level of involvement in the PSA decision were in the top and bottom quartiles on those survey items. The semistructured interviews took place in community organizations and lasted approximately 60 minutes. Finally, team members employed an observational guide that included a description of the issues discussed by participants at the conclusion of the educational session. Observational data for community identification were
August 2008
analyzed in 2004, and evaluation data analysis (reported here) was completed in 2007. Univariate and multivariate statistical analyses were conducted to test for significant differences in structured survey responses across time points. For single-item analyses, McNemar’s test was used when comparing time-related proportions, and Pearson’s chi-square tests were conducted when comparing responses at single time points. Generalized estimating equations33 that account for correlated responses were used to test time effects on key outcomes while controlling for demographic variables. Qualitative semistructured interview responses were analyzed using an ordered metamatrix34 in which responses were displayed graphically and emergent themes across and within respondents were identified and assessed in an iterative analytic process. Structured observations were reviewed in a debriefing session with all participating project staff immediately following each intervention. Three categories of mixed quantitative and qualitative data were analyzed to assess and compare the community-based educational sessions. The first category consisted of pre- and post-intervention responses to survey questions about prostate cancer, the PSA test, and the clinical options available. These data were augmented with in-depth interview questions assessing participants’ recall of the interventions. The second category of data consisted of pre- and post-intervention responses to survey questions on participants’ preferred role in the PSA decision-making process and post-intervention structured observations of group discussions about the decision to have a PSA test. These data were augmented with in-depth interview questions assessing participants’ perspectives and preferences related to screening decision-making. Finally, the third category of data consisted of post-intervention responses to survey questions on the extent to which men discussed the PSA test and shared their feelings about the PSA with their physicians. These data were augmented with in-depth interview questions assessing how well the interventions facilitated men’s participation in the PSA test decision-making process.
Results The community-matching procedure provided two distinct socioeconomic categories of community residents. Participants from the lower-SES and upperSES communities were statistically distinct on every demographic variable except health insurance, primarily because many lower-SES residents had publicly funded health insurance. No significant differences were found between the educational levels or household incomes of study participants within the matched communities across the two metropolitan areas (Table 1). These data were consistent with the tract-level census data, which indicate that the selfselected sample of community residents who participated in the study was representative of census-tract residents on these variables. A considerable proportion of lower-SES community residents had demographic characteristics, such as low income, low education, and publicly funded health insurance, that have been shown to be barriers to accessing Am J Prev Med 2008;35(2)
89
Table 1. Participant demographics by community and SES level (%) Variable Educationa High school or less Some college College graduate or higher Household income ($)a ⬍20,000 20,000–39,999 40,000–59,999 ⱖ60,000 Racea,b Black White Other Insurance status Medicare Medicaid Employer-based Veterans Affairs Military Prescription coverage None Personal doctora Yes a
Greensboro lower SES (nⴝ90)
Wilmington lower SES (nⴝ29)
Greensboro upper SES (nⴝ146)
Wilmington upper SES (nⴝ96)
21.6 41.9 36.5
22.7 36.4 40.9
17.9 14.2 67.9
12.2 26.7 61.1
7.8 17.8 24.4 28.9
3.5 17.2 13.8 34.5
2.7 13.7 20.6 49.3
2.1 17.7 26.1 38.5
65.6 16.7 4.1
37.9 34.5 0.0
13.7 78.1 1.5
11.5 83.3 2.0
44.9 27.0 47.8 16.7 11.1 3.3 5.6
45.5 13.0 65.5 10.3 10.3 3.5 3.5
54.1 9.7 69.2 6.9 4.1 3.4 4.1
67.7 8.7 60.4 10.4 11.5 7.3 3.1
73.6
90.9
85.1
90.2
Difference between lower-SES and upper-SES communities’ p⬍0.05 27% of respondents in the Wilmington lower-SES community refused to respond to the questions on race/ethnicity.
b
health care or participating in IDM.8,10,11,22–27 Approximately 22% of lower-SES community participants had a high school education or less, and almost 25% had annual household incomes of less than $40,000 (Table 1). Almost 20% of the lower-SES respondents in both metropolitan areas had no personal doctor, and more than 12% of upper-SES study participants also did not have a personal doctor.
Educating Community Participants The educational sessions resulted in a marked increase in accurate responses to survey questions about prostate cancer and the PSA test immediately after the educational session, which then declined slightly over the following year. However, knowledge levels were higher at the 12-month follow-up than at baseline (Table 2). Several survey items reveal improvements in participant knowledge related to prostate cancer and the PSA. First, study participants became more knowledgeable about treatment for early-stage prostate cancer. At baseline, only 13% of participants recognized that doctors do not know which treatment is best for early-stage prostate cancer; at the 12-month follow-up, 30% of participants answered accurately. Second, participants retained knowledge regarding the side effects of prostate cancer treatment: 10% of participants at baseline and 22% at the 12-month follow-up knew that about 5 of 10 men would expe90
rience permanent side effects. Third, 59% of participants at baseline and 73% at the 12-month follow-up knew that most prostate cancers are slow growing. Finally, participants learned that prostate cancer can occur even if PSA levels are normal. The proportion of participants who knew this fact rose from 56% at baseline to 72% at 6 months; the proportion then declined to 63% at 12 months. There were no significant differences in knowledge gained between recipients of the two educational session formats. Residents of the upper-SES communities in both metropolitan areas had higher scores on these knowledge items prior to the community sessions, but this disparity disappeared following the community sessions (Figure 2). In-depth interviews with residents scoring highest and lowest on the knowledge items augmented the survey data by revealing which key messages study participants recalled in the absence of structured response options. All interview participants, including those with the lowest knowledge item scores, recalled that most prostate cancers are slow growing, that half of men who receive treatment will have permanent side effects, and that men should make an informed decision about whether to get the PSA test.
Encouraging Deliberation The educational sessions encouraged participant deliberation on the PSA decision. At the conclusion of the
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
Table 2. Participant knowledge of prostate cancer and the PSA test (%) Knowledge variable Most prostate cancers are slow growing Truea False Don’t know Prostate cancer can occur even if PSA levels are normal Truea False Don’t know Which treatment works best for prostate cancer Surgery Radiation Surgery and radiation work equally well Doctors don’t knowa I don’t know Out of 10 men, how many will have permanent side effects from prostate cancer treatment None About 2 About 5a All 10 Don’t know
Baseline (nⴝ361)
Post-intervention (nⴝ339)
6-month follow-up (nⴝ274)
12-month follow-up (nⴝ254)
59.0 8.4 32.7
74.2 11.3 14.6
74.4 8.2 17.4
72.6 5.9 21.5
56.0 4.7 39.3
78.4 4.9 16.8
71.9 6.1 22.0
62.5 7.0 30.5
13.3 16.5 11.4 13.3 45.5
5.8 3.7 13.5 62.3 14.7
8.5 11.6 11.1 39.3 29.5
10.0 10.8 13.8 30.3 35.1
2.6 16.3 9.6 3.5 68.1
— 19.8 54.1 3.0 23.1
0.8 18.9 27.6 5.9 46.8
0.5 17.3 21.9 6.5 53.8
Note: The — symbol indicates that data were not collected for the variable at the time point. a Indicates accurate response option PSA, prostate-specific antigen
presentation, participants asked questions of the physician and participated in a group discussion on issues related to prostate cancer, the PSA test, and the treatment options for prostate cancer. The discussed questions and issues included the risk factors for fastgrowing prostate cancer, the current state of the science in preventing the development or spread of the disease, likely scenarios during the patient–physician discussion of the PSA test, and prospective outcomes of the various decision options currently available. Many participants applied the information provided in the presentation to describe their decision-making process. At the end of each session, many participants described the concluding discussion as the session’s most informative component.
The education sessions encouraged deliberation on the desired level of participation in the PSA decision. Participants generally preferred a more active role in subsequent doctor–patient deliberations on the PSA decision: 33% of participants at baseline wanted to make the final PSA decision compared with 40% of participants at the 12-month follow-up. Likewise, 46% of participants at baseline preferred to make a joint decision with their physician compared with 50% at the 12-month follow-up. There were no significant differences in desired level of participation in the PSA testing decision-making between recipients of the different intervention formats or the upper- and lowerSES communities. In-depth interviews highlighted key differences between residents in the top quartile of preferred level of involvement in the PSA decision. Most men who responded to the survey question that they preferred making the final decision themselves expressed an interest in hearing their physician’s thoughts about the PSA test. However, some of those in the upper quartile of responses expressed the opinion that there was no PSA decision to discuss with their physicians. These men strongly stated an intention to continue receiving the test.
Facilitating Participation in Decision-Making
Figure 2. Knowledge-level changes by socioeconomic status
August 2008
Both educational session formats led the majority of men to discuss the PSA test and share their feelings about the test with their physicians. By the 12-month Am J Prev Med 2008;35(2)
91
Table 3. Participant intentions and behaviors regarding the PSA test by program content (%) PSA-Only content Variable PSA test intention Plan to have test within next year Do not plan to have test within next year Don’t know Preferred PSA involvement Doctor decision Doctor decision with patient input Joint doctor–patient decision Patient decision with doctor input Patient decision Discussed PSA test with physiciana Yes No Shared feelings about PSA test with physicianb Yes, a lot Yes, a little No
Men’s Health content
Baseline
6 months
12 months
Baseline
6 months
12 months
75.6 10.1 14.3
— — —
59.3 17.4 23.3
66.4 16.2 17.5
— — —
53.8 15.0 31.3
2.6 19.7 47.9 28.2 1.7
3.8 1.9 40.6 50.0 3.8
— 5.8 54.0 36.8 3.5
10.4 10.4 45.1 30.2 4.1
4.8 8.4 42.2 38.0 6.6
5.6 6.8 47.5 34.0 6.2
— —
44.8 54.3
64.0 33.7
— —
40.6 58.8
55.5 44.5
— —
27.1 33.3 39.6
19.6 41.1 39.3
— — —
21.9 37.5 40.6
18.9 37.8 43.3
Note: The — symbol indicates that data were not collected for the variable at the time point. a Indicates significant difference (p⬍0.05) between PSA-Only and Men’s Health formats b Asked only of respondents who discussed PSA test with physician PSA, prostate-specific antigen
follow-up, 59% of men had discussed the PSA decision with their physicians, and 58% had shared their feelings to some extent about the test during that discussion. Men were equally likely to discuss the PSA test with a physician regardless of whether they had visited a healthcare professional in the year before the intervention. Men who received the PSA-Only format were more likely to discuss the PSA test with their physicians (p⫽0.037) (Table 3). There were no differences between upper- and lower-SES community residents on these variables 12 months after the interventions. Subsequent in-depth interviews with men who had discussed the PSA with their physician revealed several factors constraining PSA IDM. The most frequently mentioned constraint was the limited time available during the clinical encounter to sufficiently discuss the complex PSA decision. Even when given sufficient time to discuss the PSA decision, many patients felt IDM was limited by physicians’ reluctance to discuss the lack of medical certainty about the efficacy of the PSA or current treatment options for reducing mortality from prostate cancer. The limited sample of in-depth interview data did not allow for statistical analysis of differences in constraints experienced between intervention community participants. Although the education sessions did not direct men to a particular decision, it did facilitate consideration of the decision to have the PSA test. Participants were generally less likely to plan on receiving a PSA test in the year after the interventions. At baseline, 70% of participants intended to receive a PSA test compared with only 56% of participants at the 12-month follow-up. 92
Moreover, the percentage of men who were unsure whether they wanted to have a PSA test rose from 16% at baseline to 29% at the 12-month follow-up. The interventions, however, had little effect on actual testing behaviors (Table 3). At baseline, 61% of men had received a PSA test within the past year, as had 63% of men at the 12-month follow-up.
Discussion In this evaluation of two community-level interventions promoting IDM for prostate cancer screening, both interventions increased knowledge necessary for IDM to take place, encouraged the application of this knowledge in considering values and preferences related to the PSA, and were generally successful in facilitating patient–physician PSA discussions. Men who participated in the PSA-Only educational sessions were more likely to discuss the PSA decision with their physicians. This finding suggests that framing the PSA test decision in the context of less-ambiguous screening decisions is unnecessary to promote patient participation in the decision-making process. The community-selection process employed in this study identified and engaged many residents of lowerSES communities who would have been unlikely to receive such interventions in healthcare settings. The inclusion of both lower- and upper-SES communities also revealed an association between SES and the knowledge necessary to participate in IDM related to the PSA screening decision. Several previous studies have revealed an association between elements of SES at the individual or community level and health status,
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net
testing behaviors, and cancer survival rates.8 –18 This study demonstrated that community-level educational sessions can reduce disparities in the knowledge necessary for PSA IDM among lower- and upper-SES community residents. Community-level interventions have been successful in promoting deliberation on the prostate cancer screening decision,18,22 and the interventions evaluated in this study were likewise effective. The community interventions encouraged men to apply the knowledge gained in the interventions in considering their values and preferences related to the PSA test. The interactive nature of the session allowed community residents to discuss the PSA decision with peers and neighbors. If one views patients’ values as socially derived and shared beliefs about acceptable and desirable choices and behaviors,35 the collaborative nature of these sessions gave community residents an opportunity to deliberate on how shared values and preferences related to the PSA decision. The qualitative and observational data provide important insights into the outcomes of the educational sessions. All study participants, including those who scored lowest on the structured knowledge questions, were able to recall key elements of the educational sessions 12 months after the interventions without the benefit of prompts or the closeended response fields commonly employed in structured surveys. In-depth interview data also revealed that the most common constraint on PSA IDM was insufficient time in the patient–physician encounter to discuss the PSA test decision. This finding emphasizes the importance of community-level interventions in which patients are provided sufficient time, and support from family and friends, to deliberate on a clinical decision. We thank the CDC and the Association for Prevention Teaching and Research for their financial support under Cooperative Agreement No. U50/CCU300860, Project TS0845. The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of CDC or the Association for Prevention Teaching and Research. We also thank the following local physicians and project team members for their leadership and commitment to community-based health promotion and disease prevention: Dr. Cedric Bright, Old North State Medical Society; Dr. Sam Cykert, Moses H. Cone Memorial Hospital; Dr. Jorge Gonzalez, New Hanover Regional Medical Center; and Dr. Joseph Pino, New Hanover Regional Medical Center. No financial disclosures were reported by the authors of this paper.
References 1. National Cancer Institute. Screening for prostate cancer. 2006. Available at: http://www.cancer.gov/cancerinfo/pdq/screening/prostate/ healthprofessional.
August 2008
2. U.S. Preventive Services Task Force (USPSTF). Screening for prostate cancer: recommendation and rationale. Ann Intern Med 2002;137: 915– 6. 3. Harris R, Lohr KN. Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137:917–29. 4. Chan EC, Vernon SW, O’Donnell FT, Ahn C, Greisinger A, Aga DW. Informed consent for cancer screening with prostate-specific antigen: how well are men getting the message? Am J Public Health 2003;93: 779 – 85. 5. Sheridan SL, Harris RP, Woolf SH, and the Shared Decision-Making Workgroup of the U.S. Preventive Services Task Force. Shared decisionmaking about screening and chemoprevention: a suggested approach from the U.S. Preventive Services Task Force. Am J Prev Med 2004;26:56 – 66. 6. Briss P, Rimer B, Reilley B, et al. Promoting informed decisions about cancer screening in communities and healthcare systems. Am J Prev Med 2004;26:67– 80. 7. Rimer BK, Briss PA, Zeller PK, Chan EC, Woolf SH. Informed decisionmaking: what is its role in cancer screening? Cancer 2004;101 (Suppl 5):1214 –28. 8. Braveman PA, Cubbin C, Egerter S, et al. Socioeconomic status in health research: one size does not fit all. JAMA 2005;294:2879 – 88. 9. Isaacs SL, Schroeder SA. Class: the ignored determinant of the nation’s health. N Engl J Med. 2004;351:1137– 42. 10. Viswanath K, Breen N, Meissner H, et al. Cancer knowledge and disparities in the information age. J Health Commun 2006;11:1–17. 11. Ross CE, Van Willigen M. Education and the subjective quality of life. J Health Soc Behav 1997;38:275–97. 12. Ross CE, Wu C. The links between education and health. Am Sociol Rev 1995;60:719 – 45. 13. Braveman P. Health disparities and health equity: concepts and measurement. Annu Rev Public Health 2006;27:18.1–18.28. 14. Braveman P, Egerter S, Cubbin C, Marchi K. An approach to studying social disparities in health and health care. Am J Public Health 2004;94:2139 – 48. 15. Feldman PJ, Steptoe A. How neighborhoods and physical functioning are related: the roles of neighborhood socioeconomic status, perceived neighborhood strain, and individual health risk factors. Ann Behav Med 2004;27:91–9. 16. Boyd C, Zhang-Salomons JY, Groome PA, Mackillop WJ. Associations between community income and cancer survival in Ontario, Canada, and the United States. J Clin Oncol 1999;17:2244 –55. 17. Bridge PD, Berry-Bobovski LC, Gallagher RE. Promoting informed decision-making: evaluating a community-based prostate health awareness program. J Cancer Educ 2004;19:174 –9. 18. Reyes-Ortiz CA, Goodwin JS, Freeman JL, Kuo YF. Socioeconomic status and survival in older patients with melanoma. J Am Geriatr Soc 2006;54:1758 – 64. 19. Sikkema KJ, Kelly JA, Winett RA, et al. Outcomes of a randomized community-level HIV prevention intervention for women living in 18 low-income housing developments. Am J Public Health 2000;90:57– 63. 20. Myers RE. African-American men, prostate cancer early detection examination use, and informed decision-making. Semin Oncol 1999;26:375– 81. 21. Barber KR, Shaw R, Folts M, et al. Differences between African American and Caucasian men’s participation in a community-based prostate cancer screening program. J Community Health 1998;23:441–51. 22. Richardson JT, Webster JD, Fields NJ. Uncovering myths and transforming realities among low-SES African-American men: implications for reducing prostate cancer disparities. J Natl Med Assoc 2004;96:1295–302. 23. Gwede CK, Forthofer MS, McDermott RJ. Awareness of prostate cancer screening tests among African American men in underserved neighborhoods: need for community outreach. J Oncol Manag 2002;11:36 – 41. 24. Demark-Wahnefried W, Strigo T, Catoe K, et al. Knowledge, beliefs, and prior screening behavior among Blacks and Whites reporting for prostate cancer screening. Urology 1995;46:346 –51. 25. McCoy CB, Anwyl RS, Metsch LR, Inciardi JA, Smith SA, Correa R. Prostate cancer in Florida: knowledge, attitudes, practices, and beliefs. Cancer Pract 1995;3:88 –93. 26. O’Dell KJ, Volk RJ, Cass AR, Spann SJ. Screening for prostate cancer with the prostate-specific antigen test: are patients making informed decisions? J Fam Pract 1999;48:682– 8. 27. Allen JD, Kennedy M, Wilson-Glover A, Gilligan TD. African-American men’s perceptions about prostate cancer: implications for designing educational interventions. Soc Sci Med 2007;64:2189 –200.
Am J Prev Med 2008;35(2)
93
28. Merzel C, D’Afflitti J. Reconsidering community-based health promotion: promise, performance, and potential. Am J Public Health 2003;93:557–74. 29. Farquahar S, Wing S. Methodological and ethical considerations in community-driven environmental justice research. In: Minkler M, Wallerstein N, eds. Community-based participatory research for health. San Francisco: Jossey-Bass, 2003:221– 41. 30. Fishbein M. Great expectations, or do we ask too much from communitylevel interventions? Am J Public Health 1996;86:1075– 6. 31. Beebe J. Rapid assessment process: an introduction. Walnut Creek CA: Altamira Press, 2001.
32. Driscoll D, Harris-Kojetin L. Final patient messages for prostate-specific antigen (PSA) screening for prostate cancer, Medicare Screening Messages Project. Final report to AHRQ/CMS. Research Triangle Park NC: RTI International, 2002. 33. Liang KY, Zeger SL. Longitudinal data analysis using general linear models. Biometrika 1986;73:13–22. 34. Miles M, Huberman MA. Qualitative data analysis. 2nd ed. Thousand Oaks CA: Sage Publications, 1994. 35. Kluckhohn CK. Values and value orientations in the theory of action. In: Parsons T, Shils EA, eds. Toward a General Theory of Action. Cambridge MA: Harvard University Press, 1951.
Did you know? You can access back issues of American Journal of Preventive Medicine online. Visit www.ajpm-online.net today to see what’s new online!
94
American Journal of Preventive Medicine, Volume 35, Number 2
www.ajpm-online.net