Table 2.
PROPr | Intervention (n=298) | Usual care (n=287) | Total (n=585) | Estimate (95% CI)*; P value | ||
---|---|---|---|---|---|---|
ITT | CACE (full)† | CACE (full+partial)† | ||||
3 months | ||||||
No of participants | 237 | 248 | 485 | 0.03 (0.01 to 0.05); 0.02 | 0.05 (0.01 to 0.09); 0.01 | 0.03 (0.01 to 0.05); 0.01 |
Mean (SD) | 0.27 (0.18) | 0.23 (0.18) | 0.25 (0.18) | |||
Median (IQR) | 0.24 (0.13-0.37) | 0.21 (0.10-0.34) | 0.22 (0.11-0.36) | |||
6 months | ||||||
No of participants | 225 | 237 | 462‡ | 0.02 (−0.003 to 0.05); 0.08 | 0.04 (−0.004 to 0.08); 0.08 | 0.02 (−0.003 to 0.05); 0.08 |
Mean (SD) | 0.27 (0.20) | 0.24 (0.20) | 0.26 (0.20) | |||
Median (IQR) | 0.24 (0.12-0.39) | 0.21 (0.11-0.34) | 0.225 (0.11-0.37) | |||
12 months | ||||||
No of participants | 217 | 227 | 444§ | 0.03 (0.01 to 0.06); 0.02 | 0.06 (0.01 to 0.10); 0.02 | 0.04 (0.01 to 0.07); 0.02 |
Mean (SD) | 0.29 (0.22) | 0.25 (0.20) | 0.27 (0.22) | |||
Median (IQR) | 0.25 (0.12-0.43) | 0.21 (0.09-0.37) | 0.23 (0.10-0.40) |
ITT=intention to treat; CACE=complier average causal effect; SD=standard deviation; IQR=interquartile range.
Based on a partially nested heteroscedastic model adjusted for baseline overall health and stratification variables (age, level of hospital care, and level of mental health disorder); the therapist effect was included as a random effect to account for partial clustering.
Based on a single equation instrumental variable regression model with outcome adjusted for baseline overall health and stratification variables (age, level of hospital care, and level of mental health disorder).
Includes one participant who died at this time point.
Includes four participants who died at this time point.