Table 6.
Analysis of rural vs urban effects on psychosocial measures*
| Outcome measures | Urban-rural effect averaged across intervention | ||
|---|---|---|---|
| 1 week after pretest counseling | 1 week after post-test counseling (tested women only) | 6 months after last counseling | |
| Effect (95% CI) | Effect (95% CI) | Effect (95% CI) | |
| Anxiety, change from baseline | −0.87 (−1.94 to 0.31) | 0.55 (−1.52 to 2.67) | 0.47 (−0.78 to 1.83) |
| Cancer-related distress, change from baseline | 0.36 (−1.84 to 2.30) | 0.26 (−3.10 to 3.72) | 1.00 (−1.15 to 3.20) |
| Decisional conflict | - | - | −0.43 (−2.36 to 1.47) |
| Decision regret | - | - | 2.15 (−1.31 to 5.72) |
| Knowledge, change from baseline | 0.27 (−0.04 to 0.64) | −0.05 (−0.61 to 0.47) | - |
| Quality of life-physical, change from baseline | - | 0.09 (−1.73 to 2.18) | 1.06 (−0.25 to 2.59) |
| Quality of life-mental, change from baseline | - | 0.30 (−1.80 to 2.42) | −1.37 (−3.06 to 0.12) |
| Patient centeredness | |||
| Informativeness | 0.12 (−.37 to 0.59) | - | - |
| Interpersonal sensitivity | 0.00 (−0.29 to 0.29) | - | - |
| Partnership building | 0.1 (−0.41 to 0.65) | - | - |
* Estimated effect of rural vs urban from linear regression models adjusted for intervention. Models included a random family effect at one week after pretest counseling and six months after last counseling, but not at one week after post-test counseling because cluster sizes were not the same as at randomization. Associated 95% confidence intervals are bootstrap estimates. Analyses varied in the number of participants with complete data at each time point (one week after pretest counseling, the maximum n = 865; one week after post-test counseling, the maximum n = 239; six months after last counseling, the maximum n = 751) (see Table 5 for ns in each analysis). Empty cells are time points when these measures were not collected. CI = confidence interval.