PurposeThe goal of this study was to characterise adverse drug events (ADE), including both adverse drug reaction (ADR) and accidental poisoning by drugs (AP), considering age, gender, length of stay (LOS), number of deaths and year, during the period 2004–2013. Additionally distributions of the ten’s most frequent ADR and AP were characterized, considering age-group and gender.MethodsA retrospective descriptive nationwide study was conducted, based on the hospital discharges database in Portugal from 2004 to 2013, using ICD-9. Events were identified based on the following codes: from E930 to E949.9 and from E850 to E858.9.ResultsA total of 9 320 076 patients were discharged within this period, with 133 688 patients (1.46%) having at least one ADE, 4% of them related with AP. The mean age of these patients was 63.79 years (SD 21.31), 54.50% were female and the mean LOS was 14.05 days (SD 22.19). Patient with AP had a mean age of 41.06 years (SD 34.05), 54.70% were female and LOS was 7.15 days (SD 19.42). We have identified 10.691 deaths that represent 8.00% from the total of patients with an ADE. The patients above 65 years were more affected by ADR and children below 18 were more affected by AP.ConclusionIn the last decade an increasing trend of ADR were observed and an AP pattern relatively stable. Elderly people and children were the age groups most affected. Antibiotics (in ADR) and benzodiazepine-based tranquilizers (in AP) were the major problems. This is a huge, increasing and challenging problem. Further research, using individual and contextual risk factors should be developed to understand spatiotemporal variability, promoting tailored interventions, within and across countries.

BackgroundThe aim of this study is to identify the distribution by municipalities of adverse drug events (ADE) in Portugal, including adverse drug reactions (ADR) and accidental poisoning by drugs (AP), on municipality/years ADE rate clustering. Also we identify areas with different trends in time.MethodsWe used a national dataset of public hospital discharges in Continental Portugal from 2004 to 2013. Events were identified based on codes: from E930 to E949.9 (ADR) and from E850 to E858.9 (AP). Space-time clustering and spatial variation in temporal trends methods were applied in three different time-periods: globally, by year and grouped in 2 classes (periods of 5 years).ResultsA total of 9,320,076 patients were discharged within this period, with 133,688 patients (1.46%) having at least one ADE, 4% of them related with AP. Critical space-time identified clusters (p < 0.001) were the municipalities from Lisbon metropolitan area and Centro region area. The global rate increased at a 7.8% mean annual percentage change, with high space-time heterogeneity and variation in time trends clusters (p < 0.001). For whole period, 2004–2013, all clusters presented increasing trends. However when analyzed by period of 5 years we identified two clusters with decreasing trends in time in 2004–2008.ConclusionThe impact of ADE is huge, with widely variations within country and in time, and represents an increasing challenge. Future research using individual and contextual risk factors are urgently needed to understand this spatiotemporal variability in order to promote local tailored and updated actions of prevention.