Table 2.
Treatment comparison | Treatment effect in z-values (95% CI) | Treatment effect in PCL-4 scale | Between-group difference in percent change from baseline | Between-group difference for percent recovereda (95% CI) | Odds ratio for recoverya (95% CI) |
---|---|---|---|---|---|
CPT initiator v. non-EBP initiator | −0.36 (−0.44 to −0.28) | −4.8 (−5.8, 3.7) | −7.3 (−10.1, −4.5) | 7.4 (5.4–9.5) | 2.6 (2.0–3.4) |
Complete 8 + CPT v. Complete 8 + non-EBP | −0.49 (−0.75 to −0.23) | −6.4 (−10.0, −3.1) | −17.0 (−22.9, −11.1) | 11.9 (7.7–16.2) | 8.7 (3.7–20.8) |
PE initiator v. Non-EBP initiator | −0.23 (−0.36 to −0.10) | −3.0 (−4.8, −1.3) | −5.2 (−9.0, −1.4) | 5.5 (2.5–8.5) | 2.2 (1.4–3.3) |
Complete PE 8 + Complete 8 + non-EBP | −0.73 (−1.04 to −0.41) | −9.7 (−13.8, −5.4) | −14.5 (−21.5, −7.6) | 2.8 (−2.1–7.6) | 1.9 (0.6–6.2) |
PE initiator v. CPT initiator | 0.05 (−0.11 to 0.22) | 0.6 (−1.5, 2.9) | 0.4 (−4.1, 4.8) | 1.8 (−2.1–5.8) | 1.2 (0.8–1.8) |
Complete 8 + PE Complete 8 + CPT | −0.10 (−0.32 to 0.11) | −1.3 (−4.2, 1.4) | −4.1 (−11.4, 3.3) | 5.1 (−3.0–13.2) | 1.6 (0.8–3.0) |
Recovery defined as having experienced a 10-point or greater improvement from baseline and follow-up PCL-4 score <39.
bEffect is the difference between propensity-score matched treatment and control group in the change in z-scored PCL (i.e. difference-in-difference), weighted by inverse probability of having follow-up PCL. The complete adherence effect additionally accounted for treatment switching or lack of adherence with inverse probability of censoring weights. A negative value means the treatment group had greater improvement (decrease) in symptoms than the control group, e.g. CPT completers had a 0.49 greater decrease in z-score (or, 6.4-points on the PCL-4 scale) compared to the non-EBP group. Note. Initiators are those receiving any individual therapy and completers are those receiving eight or more individual therapy sessions.