Table 3

Results from the trim-and-fill method for publication bias in the 17 haloperidol trials for schizophrenia

Filled meta-analysis
ModelSummary logRR95% CIz Valuep ValueNumber of studies
Fixed effect0.2740.0840.4652.8250.00526
Random effects0.331−0.0440.7051.7310.08326
  • RR, risk ratio.