TABLE 5

Risk of Bias of Included Studies: Experimental Studies

First Author (Year)DesignaReporting (Maximum 11)External Validity (Maximum 3)Bias (Maximum 7)Confounding (Maximum 6)Power (Maximum 1)
Gadomski (1998)25Prepost without control81440
Gordon (2007)38Prepost without control91520
Liptak (1998)12Quasi-experimental, nonequivalent comparison62420
Palfrey (2004)15Prepost without control91540
  • a Instruments: Experimental designs = Modified Downs and Black Checklist,18 higher scores reflect less bias.