Assessing Effectiveness of a Chronic Disease Self-management Program in Faith-based Organizations in Barbados: a Cluster Randomized Trial

Setting: This cluster randomized trial will be conducted among members of selected church groups in Barbados - a small island developing state in the Caribbean with a population of approximately 287 025 (52% female). According to the Barbados' 2010 census, approximately three-quarters of the population identify as Christian. This study is intended for Christian churches in Barbados, from any denomination. Barbados is divided into 11 parishes. Given the small sizes of many churches in Barbados we have chosen to cluster by parish and divide the most densely populated parish (St. Michael) into two divisions. This gives us a total of 12 geographic divisions. Each division consists of 7-9 of our target churches comprising a total of about 350 members per division. The study will consider each division as a single unit which will be randomized to either intervention or control. The treatment of a division as one assembly allows the study to have sufficient participants at each recruitment event to be able to meet the required sample size of persons with uncontrolled hypertension.

Recruitment: Potential participants will be invited to recruitment days where they are screened for hypertension. All persons with a previous diagnosis of hypertension or currently on anti-hypertensives AND the occurrence of two or more blood pressure readings above 140 mmHg (systolic) or 90mmHg (diastolic) on the day of recruitment will be deemed eligible. Persons not known to have hypertension but who have two or more blood pressure readings above 140 mmHg (systolic) or 90mmHg (diastolic) on the day of recruitment are also eligible.

Clinical screening will also be conducted and persons will be included in the study if their systolic blood pressure (SBP) and diastolic blood pressure (DBP) readings are greater than or equal to 140 mmHg or 90 mmHg respectively on two or more readings taken at rest more than 5 minutes apart on the day of screening. Persons who are well controlled on blood pressure lowering drugs will be excluded. To obtain eligible participants, fliers and information on the study will be circulated on social media and via telephone. Recruitment days will be conducted with strict physical distancing protocols in place which are in keeping with the stage of restriction Barbados is experiencing at the time of recruitment. Recruitment will not occur during times of day time curfew and will conform to government directives.

Randomization of the churches will occur after completion of all baseline assessments, and the study statistician will randomly assign each church to either intervention or control groups based on a 1:1 ratio, using a pre-specified random number generator (Stata statistical software version 16). Using a 3-church block, we will assign 2 churches to intervention and 2 churches to control in each randomization block. Given the nature of the intervention, the trial will not be blinded. However, data collectors will be blinded to church randomization status until the end of the six-week recruitment period.

Sample size: Previous studies have found group mean SBP reductions ranging from 4.5 mmHg to 6.6 mm Hg17. Using the lowest group mean SBP reduction of 4.5 mmHg (SD 10)21 among the persons living with uncontrolled hypertension and given a study of 12 clusters we estimated that 168 persons in the intervention and 168 in the control group, would provide 80% power to detect a difference of 4.5 mmHg. This is assuming an intra-class correlation (ICC) of 0.0517. This required sample size of 336 provides us cluster sizes of 28. Given that CDSMP groups are typically between 12-16 each, we will recruit 32 per cluster (church) for a total of 192 persons in intervention and control arms each and a total required sample size of 384. Calculations were performed using Stata statistical software version 15.