Table 2

 Summary of effects for transition to psychosis

ComparisonTime point (months of treatment)No (%) of trials in analysisNo (%) of participants in analysisRisk ratio (95% CI), random effectsHeterogeneity (I2 (%), χ2 (P))Quality of evidence (GRADE)
CBT v supportive counseling23-25, 28, 290-64 (80)591 (88)0.62 (0.29 to 1.31)17, 3.6 (P=0.31)Low*‡
6-125 (100)645 (71)0.54 (0.34 to 0.86)0, 2.51 (P=0.64) =0.P2)Moderate*
12+4 (80)570 (85)0.63 (0.40 to 0.99)0, 2.50(P=0.48)Low*‡
CBT and risperidone v supportive counselling24, 300-62 (100)130 (100)0.35 (0.13 to 0.95)0, 0.59 (P=0.44)Very low*‡§
6-122 (100)130 (100)0.63 (0.33 to 1.21)0, 0.25 (P=0.61)Very low*‡§
12+1 (50)41 (32)0.59 (0.34 to 1.04)NAVery low*‡§
Integrated psychotherapy v supportive counselling376-121 (100)125 (100)0.19 (0.04 to 0.81)NAVery low*‡¶
12+1 (100)125 (100)0.32 (0.11 to 0.92)NAVery low*‡¶
Integrated psychotherapy v standard care356-121 (100)67 (85)0.24 (0.07 to 0.81)NALow*‡
12+1 (100)65 (82)0.52 (0.26 to 1.02)NALow*‡
CBT and risperidone v CBT and placebo240-61 (100)87 (100)1.02 (0.15 to 6.94)NAVery low*‡§
6-121 (100)87 (100)1.02 (0.39 to 2.67)NAVery low*‡§
Olanzapine v placebo276-121 (100)60 (100)0.43 (0.17 to 1.08)NAVery low*‡§
Omega 3 fatty acids v placebo260-61 (100)76 (94)0.13 (0.02 to 0.95)NALow*§
6-121 (100)81 (100)0.18 (0.04 to 0.75)NALow*§

NA=not applicable; NBI=needs based intervention.

*Reason for downgrading: imprecision.

†Reason for downgrading: inconsistency.

‡Reason for downgrading: risk of bias.

§Reason for downgrading: risk of publication bias.

¶Reason for downgrading: indirectness.