Table 6.
GEE mediation analysis of intervention effects (HIV/STI intervention Vs health promotion intervention) fit to consistent (100 %) condom use 6 and 12 months post-intervention, African American Men, Philadelphia, PA 2008–2012
| Theoretical construct | Alpha path | Beta path | Indirect effect | ||
|---|---|---|---|---|---|
| Mean difference (95 % CI) | p value | Odds Ratio (95 % CI) | p value | Alpha–Beta product (95 % ACI) | |
| Targeted by the HIV/STI intervention | |||||
| Condom-use hedonistic outcome expectancy | 0.19 (0.09, 0.29) | <0.001 | 1.16 (0.83, 1.63) | 0.388 | 0.03 (−0.04, 0.10) |
| Condom-use prevention outcome expectancy | 0.15 (0.04, 0.26) | 0.006 | 0.80 (0.62, 1.04) | 0.095 | −0.03 (−0.08, 0.00) |
| Condom-use self-evaluative outcome expectancy | 0.14 (0.04, 0.24) | 0.007 | 1.33 (0.99, 1.78) | 0.057 | 0.04 (−0.00, 0.10) |
| Condom-use availability self-efficacy | 0.08 (−0.01, 0.17) | 0.076 | 0.83 (0.60, 1.17) | 0.286 | −0.01 (−0.06, 0.01) |
| Condom-use negotiation self-efficacy | 0.05 (−0.04, 0.15) | 0.281 | 1.47 (1.07, 2.02) | 0.017 | 0.02 (−0.01, 0.08) |
| Condom-use technical skills self-efficacy | 0.15 (0.05, 0.25) | 0.004 | 0.94 (0.71, 1.26) | 0.690 | −0.01 (−0.06, 0.04) |
| Condom-use impulse-control self-efficacy | 0.15 (0.02, 0.28) | 0.020 | 1.34 (1.07, 1.69) | 0.011 | 0.04 (0.00, 0.11) |
| HIV risk-reduction knowledge | 1.17 (0.81, 1.54) | <0.001 | 0.99 (0.91, 1.07) | 0.773 | −0.01 (−0.12, 0.09) |
| Condom-use knowledge | 0.19 (0.05, 0.34) | 0.009 | 0.90 (0.73, 1.10) | 0.302 | −0.02 (−0.09, 0.00) |
| Not targeted by the HIV/STI intervention | |||||
| Condom-use subjective norm | 0.04 (−0.06, 0.13) | 0.456 | 1.22 (0.91, 1.64) | 0.189 | 0.01 (−0.01, 0.04) |
| Condom-use descriptive norm | 0.04 (−0.11, 0.18) | 0.620 | 1.46 (1.15, 1.85) | 0.002 | 0.01 (−0.04, 0.08) |
Theoretical constructs are from the immediate post-intervention assessment. Alpha path, the effect of the intervention on the theoretical construct, is adjusted for baseline consistent condom use and the theoretical construct. Beta path, the relation of the theoretical construct to consistent condom use 6 and 12 months post-intervention, is adjusted for intervention and baseline of consistent condom use and mediator. CI is confidence interval. ACI is asymmetric confidence interval based on bootstrap quantile method with 2,000 replicates