Emergency department violence and self-harm visits fell sharply during Italy’s first lockdown
Emergency department surveillance captured a blunt shift in the burden of self-harm and interpersonal violence when the first COVID-19 lockdown altered daily life across a North Italian region. Across 5,409,509 emergency department visits recorded between January 2016 and September 2025, 53,086 were related to violence or self-harm, representing 0.98% of the total. The pattern was not random. Visits clustered unevenly across provinces, rose more often in males, and peaked during evening and nighttime hours, weekends, and summer months.
One result stood out immediately: during lockdown, emergency visits tied to interpersonal violence among Italian residents dropped by about 55%.
Lockdown restrictions changed exposure opportunities and emergency department demand
The study used regional healthcare data warehouse records and an interrupted time series approach built on negative binomial regression to examine how lockdown measures changed emergency department use. For Italian residents, interpersonal violence-related visits fell to an incidence rate ratio of 0.449, with a 95% confidence interval of 0.365 to 0.552 and a p value below 0.001. Self-harm-related visits also declined, though less steeply, with an incidence rate ratio of 0.750, a 95% confidence interval of 0.571 to 0.985, and a p value of 0.039.
Those declines fit a plausible behavioral mechanism. Stay-at-home orders, reduced crowding, restricted nightlife, and fewer opportunities for street, workplace, and social conflict likely lowered the number of situations in which violence could occur and require acute care. The emergency department records do not measure every incident in the community, but they do show when injured or distressed patients actually crossed the threshold into hospital care. In that sense, the data reflect both injury incidence and the health system’s contact point for acute psychological crisis.
Foreign nationals accounted for a large share of visits and suggest persistent inequality
Foreign nationals accounted for roughly one-third of all violence- and self-harm-related emergency department visits, and their share increased over time. The distribution raises a clear equity concern. Differences in housing stability, occupational risk, access to prevention services, language barriers, legal insecurity, and help-seeking behavior can all shape who reaches emergency care and when. The study does not assign a single cause to this gradient, but the consistency of the trend makes it difficult to dismiss as a short-term fluctuation.
Among foreign residents, the lockdown-associated reductions moved in the same direction as those seen in Italian residents. For self-harm, however, the change did not reach statistical significance, likely because the subgroup included fewer events. Small counts can blunt statistical certainty even when the underlying trajectory looks similar, leaving an important gap in precision rather than a contradiction in direction.
Evening and nighttime visits were more common than daytime presentations.
Temporal clustering points to social context, not just clinical severity
The timing of presentations adds another layer of interpretation. Weekend clustering suggests that acute violence and self-harm episodes are embedded in social rhythms, not only in individual pathology. Summer months also carried a higher load, consistent with patterns often seen in injury, alcohol-related conflict, and other forms of acute behavioral dysregulation. Emergency departments absorb the downstream consequences of these exposures, but the seasonality hints that prevention cannot be confined to hospital walls.
The regional variation across provinces matters as well. Unequal rates across neighboring areas imply differences in local social conditions, service availability, or community risk profiles. A single regional average can hide these gaps, and the study’s geographic breakdown makes clear that prevention strategies will need to be tailored rather than uniform. Surveillance systems are most useful when they identify where the burden concentrates and which populations remain systematically overrepresented.
What the pandemic period revealed about injury prevention and surveillance
The lockdown era functioned as an unplanned natural experiment. When social contact narrowed and movement patterns changed, emergency department visits for violence and self-harm fell substantially, especially for interpersonal violence. The finding does not mean the underlying risk vanished. It shows that access to potential perpetrators, places of conflict, and the routines that sustain harm changed enough to alter the flow of patients into emergency care.
For health systems, the lesson is practical. Routinely collected emergency department data can serve as a real-time surveillance tool for violence and self-harm, helping public health teams track shifts by province, sex, nationality, season, and time of day. Because these events are tightly linked to social structure, the records can also expose inequality rather than obscure it. A rise in foreign-national representation, for example, may indicate unmet preventive need long before broader population trends are visible.
The study was based on fully anonymized administrative data and did not involve direct intervention or identifiable personal information, so ethical committee approval was not required under applicable regulations. It received no external funding and declared no competing interests.
The strongest contribution here is not a single statistic, but the combination of a large regional dataset, a long observation window, and a sharp policy-defined interruption. Emergency department visits for self-harm and interpersonal violence fell when lockdown restrictions changed daily exposure patterns, yet the burden remained socially patterned and unevenly distributed. Future surveillance will need to determine whether those inequalities persist once mobility, work, and social mixing return to baseline.