The optimal configuration of backpropagation (BP) neural networks was deter
mined after 35 trials with different BP configurations evaluating the total
detection rate. Ten different training and testing sets were used to ident
ify optimal samples. All trials included sample files of patients with medi
cally serious suicidal attempts (MSSA) and those of non-suicidal patients.
Fifty files were used in each group for training and 49 files for testing w
ith no overlap between the samples. The target variable for training was se
riousness of suicide attempt (0 = non-suicidal, 1 = MSSA). The input set in
cluded 44 demographic, clinical and patient-history variables. The optimal
results showed that 93.8% of MSSA and 89.8% of the non-suicidal patient fil
es were detected. Total success rate (TSR) was 91.8% and positive and negat
ive prediction values (PPV, NPV) were 92% and 95.6%, respectively. Living a
lone (6.76), treatment compliance (5.86), drug abuse or dependence (2.8), g
lobal assessment of functioning (GAF) score (1.49), non-paranoid delusions
(1.22) and suicide of first degree relative (1.1) were highly associated wi
th MSSA according to the Garson calculation. However, logistic regression a
ttributed high importance to hallucinations (p < 0.0001), diagnosis (p < 0.
002), number of children (p < 0.006), GAF score (p < 0.006), employment sta
tus(p < 0.02) and stressors(p < 0.03). It was shown that: backpropagation n
eural networks are very successful in identifying records of MSSA patients;
a high GAF score is associated with high risk of MSSA and is the only comm
on variable identified by both methods; and backpropagation identified two
non-specific factors (living alone and treatment compliance) whereas statis
tics found specific factors (hallucinations and diagnosis) highly associate
d with MSSA.