The SA-CHAMP is unique in that it demonstrated the feasibility of using trained lay community volunteers to run a sustainable, culturally adapted opportunistic CVD risk factor screening project in places of worship within the SA community in Calgary, Alberta, Canada. Furthermore, despite this being a pilot trial, modest but significant improvement in cholesterol measures were noted. Much of the success and acceptability of this program is a reflection of the high level of community engagement and volunteer participation throughout all stages of the program including the design, implementation and evaluation; key factors in the well accepted CBPR approach to reducing health disparities in minority populations . The degree of community “buy-in” or ownership of the program and the strong commitment of the lay community members that has led to the dissemination of the program beyond its original setting to SA communities across Canada also makes this program unique in Canada.
Places of worship play a strong cultural and social role in many communities across the world, and are a place where community members come together. They have been shown to provide a particularly relevant and culturally comfortable setting for health-related interventions [19, 43, 44, 56, 57]. The bulk of this evidence is derived from studies done in the United States within African American and Hispanic communities. However, there are two recent studies other than ours suggesting the feasibility of using places of worship in SA communities to identify individuals with CVD risk factors. Davachi et al.  demonstrated that when diabetes screening sessions for those without known diabetes were held in the context of religious gatherings in temples and mosques in Calgary, Alberta, it was highly acceptable to community members and is reflected in the high participation rate (922 screened in 14 four hour screening sessions). Similarly, Rao et al.  screened for CVD risk in two Hindu temples in London, England. In this study medical/paramedical community members were able to screen 434 self-selected participants in 9 six-hour sessions. This group is currently carrying out a qualitative assessment of the acceptability of the program among staff and participating community members.
Use of trained lay health workers as part of the health-care team in community-based program settings is recognized as a way to improve access and the health of underserved populations . Community members trained as peer health advisors, particularly those who are members of the target community, may play a significant role in addressing cardiovascular risk factors because of their connectedness to individuals and communities, as well as their understanding of cultural and contextual issues [28, 58, 59]. A noteworthy example is the recent study looking at the effectieveness of barbers who were trained and paid to become health educators, monitor BP, and promote physician referral for patrons of local African-American-owned barbershops in Dallas County, Texas, USA. While improvement in the primary outcome of hypertension among African-American male patrons was modest (8.8%, 95% confidence interval: 0.8%-16.9%), 98% of participants and all 29 participating barbers reported that they would like the program continued indefinitely.
Until recently, there has been a paucity of literature on the effectiveness of trained lay community members as volunteers, specifically in the context of community-based CVD risk factor screening and improved CVD outcomes. In a cluster randomized controlled trial of 39 communities in Ontario, Canada , volunteer lay seniors were trained to perform BP and cardiovascular risk assessment on other seniors invited by their physicians to attend pharmacy-based screening clinics. Within one year of the screening program, a 9% relative reduction in CVD-related hospital admissions was noted, translating to 3.02 fewer hypertension-related hospitalizations per 1000 people aged 65 and older in the intervention communities. Our study and that of Davachi et al.  differ from the Kaczorowski study in that they were both adapted for the SA community and as such, the lay trained volunteers were SA and the screenings took place in religious facilities rather than community pharmacies. The use of trained lay community volunteers in this setting shows promise and our study helps to support the feasibility of implementing a similar larger RCT in SA communities to evaluate whether such an intervention can reduce CVD in this group.
The value (in terms of identifying at-risk individuals that would benefit from intervention) of opportunistic CVD risk screening in SA places of worship has not been well described. Using a methodology similar to ours, Davachi et al.  opportunistically screened adult non-diabetic SA for diabetes in places of worship, and observed very high rates of obesity (67%), family history of diabetes (43%) and results suggestive of potential glucose intolerance (36%). Despite excluding of those with self reported CVD, hypertension, diabetes or use of lipid therapy,  Rao et al. found that 52% of participants had hypertension, 75% had central obesity, 10% had TC/HDL ratios > 6, and 15% were assessed as high risk using the QRISK2 model for 10-year CVD risk. The higher prevalence of both poorly controlled risk factors and 10-year CVD risk among participants in our study are consistent with their higher average age and their relatively high prevalence of self-reported CVD (10%), diabetes (23%), hypertension (52%) and dyslipidemia (49%). While these three studies are not directly comparable given the variable inclusion criteria, screening techniques, ages and ethnic groups represented, a high prevalence of modifiable, poorly controlled CVD risk in the SAs that self-presented for screening was a consistent finding. The majority of participants were at moderate to high risk for CVD and therefore represented an ideal target group for improving awareness and risk factor management. Further, participants in our study were satisfied with the program and barriers to healthcare provider access did not appear to be a problem. Overall, these three studies provide support for the feasibility, acceptability and value of screening SA adults for CVD risk factors in places of worship.
Our study has limitations. Participants self-selected for study and it is not known if participants in this study are representative of the SA community at large. Self-selected participants may have been more motivated, particularly those who returned for follow-up, which could partially explain the observed reduction in TC/HDL ratios and high rates of follow-up with family physicians. Our study design (single group, before-and-after) was also vulnerable to regression to the mean. While significant TC/HDL ratio reductions were still observed when the analysis was limited to those without outlying values at baseline, random variation, rather than follow-up care, may explain some of the observed improvements in lipid control. Furthermore, while there are small variations in lipid measures between the Cholestech® desk top reflometer and full laboratory-based lipid profiles, especially in the case of elevated triglycerides, Cholestech® measures are considered acceptable under these circumstances . However, we cannot rule out such variability as contributing a measure of uncertainty to our cholesterol measures. Additionally, relying on a single measure of blood pressure at baseline may have led to an overestimation of the prevalence of hypertension among those screened.
With respect to data collection, we had no way of validating self-reported information. The validity of self-report of CVD risk factors and follow-up behavior in the context of reporting such sensitive information to a community volunteer possibly known to the participant is an interesting area for future study. Further, the study protocol depended to a great extent on trained lay volunteers and may not be generalizable to all SA communities where volunteerism is not as prevalent.
Given that this study was primarily qualitative, and that no a priori power/sample size calculation was performed, the significance of the pre-post changes (or lack thereof) in clinical parameters should not be over-interpreted. These analyses were conducted on an exploratory basis only. The variability in the studies measures will be used to determine a sample size for a future, larger study.
Finally, while modified versions of the program have been initiated in Calgary and 5 other cities in Canada, it is not yet known if the programs will prove feasible on a large and sustainable scale and demonstrate effectiveness within the target population.