|Home | About | Journals | Submit | Contact Us | Français|
Men with a family history (FH) of prostate cancer (PCA) and African American (AA) men are at higher risk for PCA. Recruitment and retention of these high-risk men into early detection programs has been challenging. We report a comprehensive analysis on recruitment methods, show rates, and participant factors from the Prostate Cancer Risk Assessment Program (PRAP), which is a prospective, longitudinal PCA screening study.
Men 35–69 years are eligible if they have a FH of PCA, are AA, or have a BRCA1/2 mutation. Recruitment methods were analyzed with respect to participant demographics and show to the first PRAP appointment using standard statistical methods
Out of 707 men recruited, 64.9% showed to the initial PRAP appointment. More individuals were recruited via radio than from referral or other methods (χ2 = 298.13, p < .0001). Men recruited via radio were more likely to be AA (p<0.001), less educated (p=0.003), not married or partnered (p=0.007), and have no FH of PCA (p<0.001). Men recruited via referrals had higher incomes (p=0.007). Men recruited via referral were more likely to attend their initial PRAP visit than those recruited by radio or other methods (χ2 = 27.08, p < .0001).
This comprehensive analysis finds that radio leads to higher recruitment of AA men with lower socioeconomic status. However, these are the high-risk men that have lower show rates for PCA screening. Targeted motivational measures need to be studied to improve show rates for PCA risk assessment for these high-risk men.
Prostate cancer (PCA) is the second leading cause of cancer-related deaths in men in the United States.1 Men with a family history of PCA and African American (AA) men are at 2–7 times increased risk for the disease.1, 2 PCA screening remains controversial in men at average risk. Recent randomized PCA screening studies reported either no benefit or marginal benefit to PCA screening.3, 4 However, multiple professional societies have recommended early detection measures for high-risk men (men with a family history and AA men) after careful discussion regarding PCA screening with their physicians.5, 6 Therefore, there is a particular need to study screening strategies in high-risk men in order to personalize early detection approaches and PCA risk assessment.
Recruitment and poor show rates to screening appointments pose major challenges to enrolling high-risk men for PCA screening studies. While several previous studies have reported on patient characteristics related to PSA testing for PCA or factors related to seeking PCA screening in high-risk men, none of these studies have specifically described the effectiveness of various recruitment methods for high-risk men for PCA screening. Spencer et al. evaluated self-reported rates of PSA testing for PCA in men with a family history of PCA and AA men using the 2001 California Health Interview Survey.7 While this study found that older age, higher socioeconomic status, and having a family history of PCA were predictors of PSA testing, this study was not based on a prospective screening population of high-risk men, a more motivated group of men seeking screening. Melia et al. evaluated recruitment of PCA patients through postal recruitment and recruitment through clinics to obtain permission to contact their first-degree relatives for PCA screening in England.8 While marital status was correlated with higher rates of PCA screening in first-degree relatives of PCA patients in this study, no other participant characteristics were analyzed with respect to recruitment method. Recruitment of a diverse population of high-risk men was also not performed. Taylor et al. conducted a randomized trial evaluating video vs. print material vs. a wait list control arm with respect to knowledge, decisional conflict and self-reported PCA screening.9 Self-reported screening rates did increase with any intervention, highlighting the importance of the communication of issues surrounding PCA early detection. This study did not specifically focus on effective recruitment methods to PCA screening studies, where PCA education can take place. Pruthi et al. studied the impact of knowledge of PCA on screening behaviors in siblings of men diagnosed with PCA.10 Improvement in PCA knowledge was found to be predictive of seeking subsequent PCA screening. Therefore, recruitment methods need to be evaluated for their effectiveness in recruitment and show rates for PCA screening studies for high-risk men.
The Prostate Cancer Risk Assessment Program (PRAP) at Fox Chase Cancer Center (FCCC) is a prospective, longitudinal PCA early detection and education program for men at high risk. A 10-year analysis of cancer detection in PRAP found that approximately 9% of high-risk men were diagnosed with PCA at a mean age of 55.9 years for white men and 57.4 years for AA men.11 These findings support ongoing investigation of optimal PCA early detection strategies in high-risk men. However, significant efforts are involved in recruitment to PRAP, with 50–60% of scheduled men failing to show to the first appointment. A previous analysis from PRAP showed that radio advertisements resulted in higher recruitment of AA men. 12 AA recruitment was guided by a framework called Social Marketing where relationships are built with customers (men at risk for PCA in this case) in order to develop a value-creating exchange process. 13 One method of Social Marketing has been a focus of PRAP, Direct Response Radio (DRR). DRR is a marketing method that seeks to elicit a consumer response by having listeners respond to a toll-free number. Such social marketing strategies have historically been successful in communities where AA men reside.14, 15 Therefore, emphasis has been placed over the past ten years of the PRAP study to employ DRR to increase recruitment of AA men to PRAP.
This current analysis is a follow-up, comprehensive assessment of ten years of recruitment efforts to PRAP and is building on results from the previous analysis. We have also expanded on the previous analysis by examining show rates to the initial screening appointment by recruitment method in order to tailor recruitment approaches for these men in the future.
The objectives and design of PRAP have been described previously.16 Briefly, PRAP was established in 1996 with the objective of screening high-risk men for PCA and to provide medical and psychosocial interventions. Eligibility criteria include being male between ages 35 to 69 years and the presence of at least one of the following: (a) ≥ 1 first-degree relative with PCA; (b) ≥ 2 second-degree relatives on the same side of the family with PCA; (c) being AA regardless of a family history of PCA; or (d) having a known mutation in BRCA1 or BRCA2. The PRAP study is approved by the Institutional Review Board at FCCC.
Recruitment is defined as those who agree to participate and receive a scheduled visit date. Show rate is defined as those who participate in PRAP by attending the first visit.
Several recruitment methods have been used since 1996. Radio advertisements encouraging direct response (DRR) using a toll-free telephone number have been the primary recruitment method. Most of the advertisements aired in flighted segments three times a year in the Greater Philadelphia area. The spots run on AM and FM radio stations with majority AA listenership and have been previously described in detail.12 At the time of this analysis, PRAP was not advertised in the newspaper or on television; however PRAP has previously been promoted via news stories in newspaper or on television. PRAP has been publicized at various community events, health fairs, and in physician offices using a brochure designed by the FCCC Marketing and Communications Department. Since 1999, a printed newsletter specifically about PRAP has been distributed quarterly to a similar audience. Referrals (physician, family, friends, and acquaintances) have provided another major source of recruitment to PRAP. PRAP has been mentioned in FCCC generic newsletters (either Internet or hard copy) at least annually. Finally, although no direct recruitment to PRAP is done over the internet, PRAP is described on the FCCC website.
PRAP has also recruited participants from community Partner hospitals. Currently, three community hospitals actively recruit to the PRAP study. All recruitment materials at these sites are provided by PRAP and all intake materials are uniform throughout. This analysis focuses only on recruitment and show rates at FCCC due to variable missing data from Partner hospitals and to keep the analysis focused in one geographic region (the Greater Philadelphia area).
Figure 1 shows the flow of participants to their initial appointment. Demographic variables and recruitment methods are elicited using an extensive script in the initial phone intake. The phone-script data is entered by data managers into an Oracle RDBMS (Version 22.214.171.124.0) using custom-built Oracle Forms (Version 126.96.36.199.2).
For the statistical methods, a series of chi-square analyses were conducted to examine whether there were differences in participants’ characteristics (age, ethnicity, education, marital status, employment status, income, health insurance coverage, family history of PCA) according to their method of recruitment (radio, referral, other). Chi-square analyses were also conducted to test whether the show rate for the initial PRAP appointment varied according to the aforementioned participant characteristics. These analyses were also repeated separately according to participants’ ethnicity (white, AA) and recruitment method (radio, referral, other). For each group in which two or more participant characteristics were associated with the show rate, we conducted a follow-up multiple logistic regression analysis with the significant characteristics from the chi-square analyses included as independent variables and show rate as the dependent variable. We report the results of likelihood-ratio tests for each independent variable in these multiple logistic regression analyses. A cutoff of p < .05 was used to determine statistical significance for all analyses.
Table 1 displays the frequencies for the participant variables in the full PRAP cohort and according to recruitment method. In the full sample, there was considerable variability in age (M = 48.7 years, SD = 8.6), education, and income. Three-quarters of the participants were AA, under two-thirds were married or partnered, and most reported being employed full time. More than three-quarters of study participants reported having health insurance and just over half had a family history of PCA. More individuals were recruited via radio than from referral or other methods (χ2 = 298.13, p < .0001). There were statistically significant differences in several participant characteristics according to recruitment method. Compared to referral or other recruitment methods, individuals recruited via radio were more likely to be AA, less educated, not married or partnered, and were less likely to have a family history of PCA. Participants recruited via referral were more likely to have a higher income compared to other individuals.
Table 2 displays the show rates for the initial PRAP visit by participant characteristics and recruitment method. Across all participants, the show rate for the initial PRAP appointment was 64.9%. Characteristics associated with a lower show rate included being AA, low education, not being married or partnered, lack of employment, low income, and having no family history of PCA. Each characteristic, with the exception of family history of PCA (p = .998), was also significantly associated (ps ≤ .027) with show rate in a multiple logistic regression analysis. In subgroup analyses, the show rate for white participants (83.2% show rate overall) did not differ significantly according to the other participant characteristics. However, AA participants (overall show rate = 59.7%) were less likely to attend their initial PRAP appointment if they had a low level of education, were not married or partnered, lacked current employment, or had a low income. Each of these characteristics remained significantly associated with the show rate in a multiple logistic regression analysis (ps ≤ .036).
In the full sample, individuals recruited via referral were more likely to attend their initial PRAP appointment than those recruited by radio or other methods (χ2 = 27.08, p < .0001). In subgroup analyses, the show rate among individuals recruited via radio was lower if they were AA, had a low level of education, were not married or partnered, or did not have a family history of PCA. In a multiple logistic regression analysis, the lower show rate for African Americans (p = .006) and those not married or partnered (p = .008) remained significant, but neither education (p = .086) nor family history of PCA (p = .695) was associated with the show rate. Among individuals recruited via referral, the only participant characteristic associated with the show rate was employment status, with a lower show rate found among individuals not currently employed. The show rate for individuals recruited via methods other than radio or referral did not vary according to any of the participant characteristics.
The importance of studying PCA risk assessment in high-risk men is coming to light in order to discuss PCA screening and gain knowledge of how to personalize approaches to diagnose aggressive PCA at a curable point. Studying PCA screening strategies is of particular importance given the controversy surrounding PCA screening for the general male population.3, 4 Recruitment of high-risk men to PCA screening studies should lead to a careful discussion between doctor and patient regarding the pros, cons, and unknowns regarding PCA screening in high-risk men. However, there are significant challenges in recruiting high-risk men to take part in screening studies. Enhancing show rates to early detection studies will help to educate high-risk men to make an informed decision regarding PCA screening. In this report, our objective was to identify participant factors that are associated with recruitment to PRAP and show to the first visit by recruitment method. To our knowledge, this is the first such comprehensive analysis regarding high-risk men recruited to a PCA early detection program.
Radio advertisements led to the majority of recruitment to PRAP, with 63.5% of men recruited via radio recruitment efforts. In addition, we found that radio advertisements resulted in the recruitment of high-risk men who have less education, are less frequently married, and have lower rates of a family history of PCA. Therefore, radio remains a viable avenue for conveying health information to underserved populations, many of whom are at high-risk for PCA. However, the dilemma arises that these same underserved, minority males recruited through radio advertisements have lower show rates to their initial PCA screening visits. Targeted measures that enhance men’s attendance at screening visits should be studied. Such measures may include mailed educational material regarding PCA early detection, discussing fears related to PCA screening tests, and tailoring radio advertisements by including the importance discussing PCA screening with medical professionals. We plan in the future to evaluate radio messages that stress the importance of discussing PCA screening with medical professionals because, while PCA screening remains controversial, it is still the second leading cause of cancer-related deaths in US men. 1
Referrals remain an important recruitment method for PCA screening studies for high-risk men. In our study, 22.6% of men recruited to PRAP were by referrals, the majority of whom were white men. Show rates to the first PRAP visit were enhanced if men had a higher income and/or were employed. This is likely related to having health insurance coverage for PCA screening services. Men who were recruited to PRAP via referrals did have higher show rates to the first appointment. This highlights how personal communication with high-risk men from doctors, family, friends, religious advisors, and others may enhance men’s seeking and showing for PCA education and discussion of screening.
Health insurance coverage remains a barrier to participation. AA men represent a large subset of high-risk men without health insurance. In our study, AA men were less likely to have health insurance coverage than White men (χ2 = 7.05, p = .005). Ironically, these AA men are also primarily recruited via radio advertisements by which more AA men are recruited to PRAP. Our findings remain of concern since one report found that health insurance is a determinant of the stage of presentation in men treated with radical prostatectomy for localized PCA.17 Continued efforts are needed to provide AA men access to PCA screening studies where the pros and cons of PCA screening can be discussed.
Our report has several limitations primarily related to missing information. There were varying degrees of missing information regarding each of the demographic variables (Table 1 and Table 2), which is likely due to information being obtained via phone intake and by questionnaire. Another limitation was that data from the Partner hospitals were not able to be included due to the amount of missing information. Barriers may exist at the community level regarding adherence to high-risk screening protocols and acquiring demographic information from participants via a self-reported questionnaire. Data regarding health insurance information was limited as health insurance coverage was not required for entry into PRAP prior to 2003. Health insurance coverage remains a key factor to study regarding access to PCA early detection studies.
Participation of high-risk men in PCA screening studies is necessary in order to gain knowledge of how to personalize the early detection of PCA, educate high-risk men regarding the disease, and discuss the pros and cons regarding PCA screening. This analysis found that radio advertisements led to the highest recruitment, while referrals led to superior show rates. Radio resulted in higher recruitment of AA men and those with lower education status. These men also have decreased show rates to the initial appointment. Education and personal communication from physicians, family, and friends should lead to improved follow-through with show to PCA screening visits in order to discuss the issues regarding screening. Studies regarding targeted recruitment methods to motivate high-risk men to show to PCA risk assessment visits are needed in the future.
We are grateful to all participants of the Prostate Cancer Risk Assessment Program.
PA Department of Health Grant #98-PADOH-ME-98155, NIH CCSG (CA06927)
Institutional Review Board Approval:
The Prostate Cancer Risk Assessment Program is an approved study by the Institutional Review Board at Fox Chase Cancer Center – IRB # 96-091