Implementation of a couple-based HIV prevention program: a cluster randomized trial comparing manual versus Web-based approaches

Table 4 Estimated effect of assignment to web-based versus manual-based connect at both agency and individual/participant levels: longitudinal multilevel analysis

	Agency level (N=80)¹			Individual level (N=253)²
	Negative Binomial Regression Estimates			Negative Binomial Regression Estimates
	IRR	CI	p	IRR	CI	p
*Total number of couples*
Average of 3 waves	0.35*	0.13 - 0.94	0.038	0.45	0.14 - 1.45	0.181
At 6-month only	0.27*	0.09 - 0.84	0.024	0.30	0.08 - 1.10	0.070
At 12-month only	0.36*	0.13 - 0.99	0.047	0.43	0.13 - 1.40	0.163
At 18-month only	0.49	0.16 - 1.55	0.226	0.62	0.18 - 2.20	0.464
	Linear Regression Estimates			Linear Regression Estimates
	b	CI	p	b	CI	p
*Mean number of sessions completed*
Average of 3 waves	-0.31	-0.68 - 0.06	0.101	-0.25	-0.57 - 0.08	0.137
At 6-month only	-0.38	-0.80 - 0.04	0.079	-0.29	-0.66 - 0.08	0.124
At 12-month only	-0.31	-0.68 - 0.06	0.101	-0.25	-0.58 - 0.08	0.133
At 18-month only	-0.25	-0.67 - 0.18	0.257	-0.21	-0.59 - 0.17	0.279
	Logistic Regression Estimates			Logistic Regression Estimates
	OR	CI	p	OR	CI	p
*Implement at least 1 session*
Average of 3 waves	0.30	0.06 - 1.46	0.136	0.34	0.09 - 1.26	0.105
At 6-month only	0.23	0.03 - 1.50	0.124	0.23	0.05 - 1.09	0.063
At 12-month only	0.30	0.06 - 1.47	0.137	0.34	0.09 - 1.25	0.104
At 18-month only	0.40	0.06 - 2.57	0.332	0.49	0.11 - 2.22	0.354
*Completed at least 1 cycle*
Average of 3 waves	0.80	0.12 - 5.55	0.819	0.71	0.13 - 3.84	0.691
At 6-month only	1.06	0.05 - 21.1	0.972	0.67	0.05 - 8.31	0.756
At 12-month only	0.82	0.09 - 7.54	0.861	0.66	0.10 - 4.44	0.667
At 18-month only	0.64	0.06 - 7.32	0.717	0.65	0.09 - 4.74	0.666

*p<0.05.
Note:
1. Measurements at the agency level:
a. Number of couples at the agency level = sum(# couple a facilitator reported).
b. Mean number of sessions completed at the agency level = sum(mean number of sessions completed by a facilitator)/(number of facilitator in an agency).
c. Implemented at least 1 session at the agency level = at least one facilitator reported implementing at least one session in an agency.
d. Completed at least 1 cycle at the agency level = at least one facilitator reported completing at least one cycle in an agency.
2. Measurements at the individual level:
a. Number of couples at the individual level = sum(# couple a facilitator reported).
b. Mean number of sessions completed at the individual level = sum(number of sessions completed with a couple)/(number of couples).
c. Implemented at least 1 session at the individual level = facilitators reported implementing at least one session.
d. Completed at least 1 cycle at the individual level = facilitators reported completing at least one cycle.

ISSN: 1748-5908