Pre-implementation
Of the 3 FBOs selected, 2 were located on the outskirts of Bridgetown – the capital of Barbados, while the third was located in a rural community.
An engaged leadership
All 3 churches have taught on health-related issues at some time in the past; 2 sites had sessions in the last year – both had talks on cancer. In addition, one site organised a series of NCD-related talks on diabetes, obesity, exercise and nutrition. The same site has an ongoing programme for prevention of NCDs in the form of a bi-weekly fitness group. All 3 sites indicated intentions of improving or implementing new health programmes but no strategic policies had been documented or enacted. Leaders at all sites were receptive to the idea of a diabetes prevention programme and reported perceptions that there was a high burden of NCDs in their community. Leaders also expressed feeling a responsibility to the congregation and communities to not only offer spiritual guidance but support towards physical health as well. They described this intervention as ‘timely’ and ‘necessary’.
Adequate resources
The leaders reported that the physical plants were adequate with private space for interviews and secure storage facilities. All but one of the facilities had a designated area for cooking activities. A site visit performed by the lead investigator confirmed these reports.
During the interviews, the organizational leaders across the 3 churches identified several health professionals amongst the congregation: 20 nurses, 7 physicians, 2 physiotherapists, 1 paramedic and 6 community health aides with care of the elderly experience; however, the sites had neither structured health programmes nor designated health leaders.
In response to the request for volunteers, 40 persons volunteered for training: 20 with no previous health experience, 9 nurses, 4 doctors, 3 first aiders, 1 pharmacist, 1 social worker, 1 paramedic and 1 community health aide.
Easy Access to intervention characteristics
Sixteen CHAs responded to the access to intervention-related information survey: 9 nurses, 2 of whom were retired and 2 in training; 3 persons with no health training, 2 with care of the elderly exposure, 1 paramedic and 1 first aider. Responses were condensed from the 5-point Likert scale (strongly agree, agree, neutral, disagree, strongly disagree) to a 3-point scale (agree, neutral, disagree) for easy reporting (Table 3). The majority of answers indicated that the content and delivery was adequate. One responder mentioned the need for additional dietary information including dietary allowances and menus. Another person thought that the time allotted to the course was inadequate especially for persons who had no previous health exposure; interestingly, all persons who had no previous exposure agreed that the time allotted to the course was adequate. One person disagreed that their knowledge and skills base was expanded while on this course; in each case, the person had previous health training.
Table 3
Responses of CHAs on the content, delivery and acceptability of the intervention
Question n (%)
|
agree
|
neutral
|
disagree
|
1. The content of the course was of an acceptable standard
|
15 (94)
|
0
|
1 (6)
|
2. The information was delivered in such a way that it was difficult to understand
|
1 (6)
|
0
|
15 (94)
|
3. The time allotted to the course was adequate
|
14 (88)
|
1 (6)
|
1 (6)
|
4. My knowledge base was expanded while on this course
|
14 (88)
|
1 (6)
|
1 (6)
|
5. My skills base was expanded while on this course
|
15 (94)
|
0
|
1 (6)
|
6. The course is good enough to be offered for a fee
|
13 (81)
|
3 (19)
|
0 (0)
|
7. I am likely to recommend this course to a friend or colleague should it become publicly available
|
16 (100)
|
0
|
0 (0)
|
8. Seeing participants according to the intervention schedule will disrupt the church schedule
|
0 (0)
|
3 (19)
|
13 (81)
|
9. Seeing participants according to the intervention schedule fits into my schedule
|
13 (81)
|
3 (19)
|
0 (0)
|
10. Utilising the church as the intervention site is convenient for participants
|
15 (94)
|
1 (6)
|
0 (0)
|
11. Utilising the church as the intervention site hinder participants from freely sharing information
|
0 (0)
|
2 (12)
|
14 (88)
|
Individuals’ attitudes to the intervention
There were mixed reactions from the FBO leaders regarding the congregations’ utilisation of health services of CHAs chosen from among them; two sites answered in the affirmative and the other was unsure. CHAs however, unanimously agreed that the FBO was a convenient and acceptable location for clients and that the schedule was not disruptive to the FBO or to themselves. This was confirmed by that attendance register which showed that all clinical measurement stations were covered by CHAs over the intervention period, barring one visit at one site during a church convention when the facility was closed and CHA and participants were away from the site.
In addition, the participant register showed high attendance throughout the study. Three participants prematurely discontinued the intervention; 1 at week 3 and 1 at week 10, both due to travel and another 1 at week 6 who was no longer interested in continuing; all 3 had pre-diabetes. The lowest attendance of 82% occurred during the convention week; apart form that, the lowest attendance was 89% at weeks 8 and 11. There were 21 absences in all (out of 429 visits); 1 absence occurred in 5 people, 2 in 2 people, 3 in 1 person, 4 in 1 person and 5 in 1 person.
A changing organisation
When applying the TTM to the leaders’ and by extension, the organizations’ intent to change; leaders have used the opportunities afforded via this study to progress from the contemplation phase to the preparation phase i.e. from a point where there was no concerted effort to rectify the health challenges in their organisation to a place where they have facilitated the assembly of a team of CHAs to support the intervention.
Implementation
Community Health Advocates
Forty persons volunteered for the CHA training program (35 females: 5 males); of these, 32 (80%) attended at least 1 session. Four of the 32 discontinued the program during the first 2 weeks; the reasons were: work commitments, family commitments, change in church membership and the program was not what was expected. The remaining 28 (70%) completed the training and 27 (68%) of these took the clinical assessment; 10 at 2 sites and 7 at the third site. Of those 27, 14 had previous health worker experience including paramedics, nurses and clinicians.
Implementation Effectiveness
Amongst those who took the competence examination, the success rate for each station was: BP (100%), weight (93%) [2 persons were not assessed and hence failed by default; all who were assessed passed], hip circumference (89%), waist circumference (78%) and glucose (74%). All CHA passed at least 1 station. The odds of success were better in persons with previous clinical training only at the glucometer station (OR = 8.3, 1-224; p = 0.05) (Table 4).
Table 4
Percentage success rate at each clinical measurement station for Community Health Advocates with previous health care experience vs those with no experience
|
Experience (%)
|
No experience (%)
|
OR (95% CI)
|
waist
|
85
|
71
|
2.1 (0.3, 19.8)
|
hip
|
92
|
93
|
0.9 (0.02, 38.9)
|
weight
|
100
|
85*
|
5.3 (0.4, 198.0)
|
glucose
|
92
|
57
|
8.3 (1.0, 224)
|
BP
|
100
|
100
|
1
|
* 2 (15%) persons did not take the test and hence failed by default |
Intervention Effectiveness
One hundred and fifty-six persons were interviewed over a 6-month period; 15 were immediately eligible; of these 13 participated; the other 2 migrated. An additional 77 persons were potentially eligible based on age and BMI; 64 attended their appointments for screening for glucose status and of these 18 were enrolled.
In all, 31 participants were enrolled across the 3 FBOs; further recruitment was interrupted by the lockdown due to COVID19 pandemic. There were 28 (90%) females and 3 (10%) males, 11 (35%) with T2DM (10f: 1m), of whom 3 were diagnosed during screening for this study, all female. The 8 persons who were previously diagnosed were all on glucose-lowering medication; 7 on metformin; 3 of these in combination with a sulphonylurea and 2 with a dipeptidyl peptidase-4 (DPP-4) inhibitor; 1 person was on a sodium-glucose cotransporter-2 (SGLT-2) inhibitor alone. Twenty participants (65%) were enrolled with pre-diabetes, 13 were diagnosed during screening for this study (12f: 1m); 3 of the 7 who were previously diagnosed were on glucose-lowering medication; all 3 on metformin, 1 in combination with a DPP-4 inhibitor. Eleven participants had a history of HTN; all were on anti-HTN medication. Nine participants had a history of high cholesterol, 4 were on HMG-CoA reductase inhibitors.
Average entry measurements were as such mean (95%CI): weight (kg) 93.7 (86.2 to 101.1) 90.6 in females vs. 116.0 in males; FBG (mmol/l) was 6.4 (6.1 to 6.7); 6.6 in T2DM vs. 6.3 in pre-diabetes and HbA1C was 6.6% (6.3 to 6.9); 6.6% in both T2DM and pre-diabetes (Table 5).
Table 5
Change in weight, FBG and HbA1C post LCD intervention
|
Pre-intervention
Med (IQR)
|
Post-intervention
Med (IQR)
|
Change
Med (IQR); p value
|
HbA1C (% ) [all]
|
6.3 (6.0-6.9)
|
6.2 (5.8–6.5)
|
0.2 (0-0.6); p = 0.0004
|
HbA1C (%) [T2DM]
|
6.7 (6.2–7.1)
|
6.2 (5.9–6.7)
|
0.3 (0.2–0.6)
|
HbA1C (%) [predb]
|
6.2 (6.0-6.5)
|
6.2 (5.8–6.5)
|
0.1 (-0.1-0.5)
|
FBG (mmol/L) [all]
|
6.3 (5.8–6.9)
|
6.0 (5.4–6.4)
|
0.4 (-0.1-0.9); p = 0.01
|
FBG (mmol/L) [T2DM]
|
6.6 (6.0-7.1)
|
6.3 (6-6.8)
|
0.4 (-0.4-0.8)
|
FBG (mmol/L) [predb]
|
6.2 (5.8–6.6)
|
6.0 (5.4–6.3)
|
0.4 (-0.1-0.9)
|
Weight (kg) [all]
|
92.5 (79.8-101.6)
|
88.2 (72.5–95.3)
|
7.6 (4.5–8.8); p < 0.00001
|
Weight (kg) [male]
|
117.1 (104.0-128.7)
|
106.0 (97.0-118.2)
|
9.8 (6.3–10.5)
|
Weight (kg) [female]
|
91.6 (78.0-100.6)
|
87.0 (71.7–94.8)
|
7.2 (4.6–8.6)
|
Mean (95%CI) change post-intervention was as such: weight loss 6.8kgs (5.4 to 8.2), p < 0.00001; 7.9kgs (-3.4 to 19.2) in males vs 6.6kgs (5.2 to 8.1) in females. FBG (mmol/l) reduction was 0.4 (0.1 to 0.7), p = 0.004; 0.5mmol/l (0.1 to 0.8) in persons with prediabetes and 0.3 (-0.3 to 0.8) in persons with T2DM. HbA1C decreased by 0.4% (0.1 to 0,7), p = 0.003; with a similar reduction in both persons with prediabetes and T2DM. SBP and DBP decreased by 10mmHg; p = 0.003 and 8mmHg p = 0.005 respectively.
Of the 11 persons with T2DM who completed the LCD intervention, 10 persons were available for FBG and HbA1C testing at week 12. Based on an HbA1C threshold of < 6.5%, 6 (60%) were in remission. Based on the FBG of < 7mmol/l, 9 (90%) were in remission.
Of the 20 persons with pre-diabetes, 17 completed the study and were available for HbA1C testing at week 12; only 15 were fasted and hence had FBG done. Based on an HbA1C threshold of < 5.7%, 3 (18%) reverted to normal glucose status. Based on FBG, 6 (40%) reverted to normal glucose status (< 5.6 mmol/L). 1 person achieved normal glucose status by both FBG and HbA1C.
Five of the 11 participants that were anti-hypertensive medications either decreased or discontinued the medication.