The aim of the paper is to describe cases of violent death in children up to five years of age, where the abuse of addictive substances by their mother or other caregiver was detected or could have been presumed. Based on the results, the aim is to discuss the possibilities of preventive protection of children under 5 years in families at risk. Theoretical base: A problematic lifestyle in childcare decreases the quality of care and can lead to the death of the child. Effective measures are especially necessary for children from families with risky behavior. Methods: Retrospective analysis of autopsy files (at 2 forensic medicine departments) of children under 5 years of age who died suddenly, unexpectedly, and violently (n = 192) between 2007 and 2016. Outcomes: There were 41 cases of violent deaths. In all cases there were families at risk and socially weak families. Control of the families by the Department of Social and Legal Protection of Children had been demonstrably conducted in 10 cases, which is 24% of the cases in the sample. Alcohol was present in 29% of cases, other illegal substances in 29% of cases. In total, the use of alcohol or illegal substances was found in 44% of violent child deaths. Conclusions: Early detection of risks, dispensarization of endangered children and rehabilitation of their families is advisable preventively. The prevention of families with risky behavior and the protection of their children should take both paths -preventive and restrictive. The necessary condition for the protection of children is functional multidiscipline cooperation of all participating institutions.

The validity of infant mortality data is essential in assessing health care quality and in the setting of preventive measures. This study explores different diagnostic procedures used to determine the cause of death across forensic settings and thus the issue of the reduced validity of data. All records from three forensic medical departments that conducted autopsies on children aged 12 months or younger (n = 204) who died during the years 2007–2016 in Moravia were included. Differences in diagnostic procedures were found to be statistically significant. Each department works with a different set of risk factors and places different emphasis on different types of examination. The most significant differences could be observed in sudden infant death syndrome and suffocation diagnosis frequency. The validity of statistical data on the causes of infant mortality is thus significantly reduced. Therefore, the possibilities of public health and social policy interventions toward preventing sudden and unexpected infant death are extraordinarily complicated by this lack of data validity.