TABLE 4.
Results of Sensitivity Analyses for Selected Outcomes Measures Used to Compare Men Randomized to the Intervention Versus Control Groups in a Risk-Reduction Intervention Evaluation of African American Men Aged 18 to 29 Years Newly Diagnosed With an STD: Southern United States, September 2004 Through May 2006
| Univariate |
Multivariable |
|||
| Measure of Effect, Estimate (95% CI) | P | Measure of Effect, Estimate (95% CI) | P | |
| Observations with extreme outlying values excludeda | ||||
| Partners in past 3 monthscd | −1.52 (−2.37, −0.67) | <.001 | −1.37 (−2.18, −0.55) | .001 |
| Partners in past 3 monthsc,e | −1.28 (−2.14, −0.43) | .004 | −1.19 (−2.01, −0.36) | .006 |
| Unprotected acts of sexual intercourse, past 3 monthsc,f | −6.1 (−13.7, 1.4) | .11 | −1.0 (−9.8, 7.9) | .83 |
| Unprotected acts of sexual intercourse, past 3 monthsd,f | −4.6 (−12.0, 2.8) | .21 | −3.3 (−11.9, 5.4) | .43 |
| Response variable transformed logarithmically to reduce the impact of outlying valuesb | ||||
| Partners in past 3 monthsc,d | −0.33 (−0.49, −0.17) | <.001 | −0.32 (−0.47, −0.17) | <.001 |
| Partners in past 3 monthsc,e | −0.29 (−0.48, −0.10) | .004 | −0.29 (−0.48, −0.11) | .004 |
| Unprotected acts of sexual intercourse, past 3 monthsc,f | −0.53 (−1.00, −0.07) | .03 | −0.32 (−0.88, 0.24) | .26 |
| Unprotected acts of sexual intercourse, past 3 monthsd,f | −0.50 (−0.99, 0.00) | .051 | −0.42 (−1.00, 0.17) | .15 |
Note. STD = sexually transmitted disease.
Four participants in the control group and 3 participants in the intervention group were excluded who claimed more than 100 unprotected acts at baseline or follow-up or who claimed more than 25 partners at baseline or follow-up.
The transformed value is the natural logarithm of 1 plus the original value. Point and 95% confidence interval estimates for measures of effect are not directly comparable to those obtained in the absence of a logarithmic transformation; the main feature of interest is whether the P value is in qualitative agreement with the corresponding P value in Table 2 (i.e., both < .05 or both > .05).
Multivariable analysis controls for income, mixed STDs, and the baseline value for number of female sexual partners in the past 3 months.
Multivariable analysis controls for income, mixed STDs, and baseline values for skill, unprotected sexual intercourse, and condom use at last act of sexual intercourse.