Skip to main content
. 2022 May 31;19(5):e1003978. doi: 10.1371/journal.pmed.1003978

Table 4. Effect of LifeBuoy relative to LifeBuoy-C on the SIDAS after adjustment for baseline variables associated with attrition.

Difference between arms at T1a,b,c Difference between arms at T2a,b,c Time × condition interactionsb
SIDAS p < 0.001 p = 0.011 T0 to T1: B = −4.98, 95% CI [−7.66, −2.29], t[450.34] = −3.65, p < 0.001
T1 to T2: B = 1.28, 95% CI [−1.21, 3.77], t(215.63) = 1.01, p = 0.312
T0 to T2: B = −3.70, 95% CI [−6.64, −0.75], t[519.54] = −2.47, p = 0.014

aTests for a specific time point were conducted by recoding relevant variables representing time within models.

bAdjusting for variables differentially associated with attrition at T1 and T2: university degree, depression, anxiety, distress, and well-being at baseline.

cIn these adjusted analyses, d = 0.47 at T1 and d = 0.32 at T2 both favoring LifeBuoy (i.e., lower SIDAS scores for those who received the LifeBuoy intervention versus LifeBuoy-C). By comparison, in the original unadjusted analyses, d = 0.45 at T1 and d = 0.34 at T2 both favoring LifeBuoy.

SIDAS, Suicidal Ideation Attributes Scale; T0, baseline; T1, postintervention; T2, 3-month postintervention.