Nurse‐Reported Patient Safety Incidents in an Italian Teaching Hospital: A 10‐Year Retrospective Study of Trends, Contributory Factors, and Implications for Nursing Practice

ABSTRACT Aims To examine 10‐year trends in nurse‐submitted patient safety incident reports and to identify report characteristics associated with perceived preventability, contributory factors, and harm severity in a large Italian teaching hospital. Background Incident‐report counts reflect both reporting behaviour and underlying safety events. Because voluntary systems are subject to under‐reporting and reporting‐culture effects, they are best interpreted as learning signals rather than direct measures of harm. Nurse‐specific longitudinal analyses may clarify profession‐specific patterns in reporting‐system use and highlight priorities for local safety learning. Design Retrospective observational study reported following the STROBE guideline. Methods We analysed all 216 nurse‐submitted incident reports recorded between July 2015 and June 2025 using descriptive statistics, negative binomial regression for annual counts, and logistic and ordinal models. Denominator‐adjusted reporting‐intensity metrics were calculated for 2020–2024 using ordinary admissions, total hospitalisations, and patient‐days. Models were interpreted as exploratory because staffing, workload, case‐mix, and independent outcome denominators were unavailable. Results Nurse‐submitted reports increased over the period. Aggression and violence (44.0%) and falls (15.7%) were the most frequent categories. Most reports described no harm (48.6%) or mild harm (22.7%). Perceived preventability and system‐factor attribution increased over time, and denominator‐adjusted reporting intensity also rose between 2020 and 2024, suggesting the increase was not explained only by hospital activity volume; true incident incidence could not be inferred. Conclusions The study demonstrates longitudinal changes in nurse reporting patterns, not direct changes in patient harm. Incident‐report analytics can inform local safety learning when interpreted cautiously and triangulated with independent indicators. Relevance to Clinical Practice Nurse managers may use feedback loops, safety huddles, and action‐tracking to convert reports into learning. Violence and falls prevention emerged as nurse‐reported priorities, but interventions should be evaluated prospectively.