A systematic search (Medline, Cochrane library, Embase, bibliographies, to
5.2000, no language restriction) was performed for published reports of ran
domized comparisons of propofol and methohexital for anesthesia during elec
troconvulsive therapy. We analyzed 15 trials with data on 706 patients. The
duration of motor seizure was shorter with propofol (range, 18-39 seconds)
than with methohexital (range, 26-48 seconds, weighted mean difference 8.4
seconds [95% CI, 6.6-10.0]). With both propofol and methohexital, there wa
s little evidence of an association between dose and duration of motor seiz
ure (for propofol: r(2) = 0.25, P = .08: for methohexital: r(2) = 0.11, P =
.27). Two small trials investigated clinical outcome; results were inconcl
usive. Data on adverse effects were sparse. Duration of seizure was not pro
ven to be a useful measure of treatment success in the study of electroronv
ulsive therapy with propofol or methohexital. The impact of the technique o
f anesthesia on the underlying disease needs to be established.