Electroconvulsive therapy is widely used for treatment-resistant psychiatric patients. There is controversy about the minimum seizure duration required for clinical efficacy.

Characterize socio-demographically the electroconvulsive therapy unit of CHPL; find a correlation between improvement on PSAS, PANSS and HAM-D and both mean seizure duration and mean electric charge used; evaluate outcome in patients with seizure < 20s.

Three groups of patients were created according to ICD-10 diagnosis. The scales mentioned were applied at the beginning and end of the treatment: depressed (F31.3-5; F33; F20.4; F25.1) - HAM-D; schizophrenic/ schizoaffective/ bipolar-manic or mixed (F20.X, F25.X, F31.0, F31.6) - PANSS; with any of these diagnosis: PSAS. The correlation between the improvement on each scale and both mean seizure time and mean electric charge was investigated using the Pearson product-moment correlation coefficient. The groups were subdivided by seizure duration (cut-off 20s) and the mean improvement on the different scales was determined for each subgroup.

There is no correlation between any of the variables analyzed (α = 0,05). HAM-D variation (N = 102): seizure duration (r = 0,004) and electric charge (r = 0,106). PANSS variation (N = 43): seizure duration (r = 0,07) and electric charge (r = 0,046). PSAS variation (N = 147): seizure duration (r = −0,043) and electric charge (r = 0,143). There is no difference statistically significant in mean HAM-D, PANSS or PSAS improvement between subgroups >20s and < 20s (11.29 vs 13.94; 30.85 vs 29.19 and 19.07 vs 22.65, respectively).

Mean seizure duration and electric charge didn’t correlate with clinical outcome. Seizure duration below 20 seconds didn’t imply poorer outcome.