We describe the epidemiology of 1918 cases of childhood poisoning refe
rred to the emergency room in Trieste, Italy, from 1915 to 1994. The i
ncidence race of emergency room referral and subsequent hospitalizatio
n was calculated on the basis of the distribution of children resident
in Trieste by calendar year. The occurrence of childhood poisoning wa
s described according to time trends, age and gender of the child, rou
te of exposure, symptoms at presentation to the emergency room, role o
f the child or others, intention, and sub stance involved in the poiso
ning. The association between presence of symptoms and characteristics
of referral, host factors and substances involved was evaluated by es
timating the odds ratio in multivariate models. Possible determinants
of the clinical decision to treat certain cases were evaluated using l
ogistic regression. Despite an increasing incidence of referral (from
155 per 100,000 persons per year in 1975-79 to 352 per 100,000 in 1990
-94), hospital admission races showed a two-fold decrease. Younger chi
ld-en (age 0-4 years) were more likely to be asymptomatic and required
treatment and hospitalization less often than older children (age gre
ater than or equal to 10 years). Trends show a decrease in pharmaceuti
cal poisonings due probably to the introduction of child-resistant con
tainers and an increase in domestic poisons. We also observed a steady
increase in carbon monoxide inhalation and alcohol poisonings, mostly
among teenagers. (C) 1998 Elsevier Science Inc.