Increased availability of paracetamol in Sweden and incidence of paracetamol poisoning: using laboratory data to increase validity of a population-based registry study.

Abstract

To estimate the incidence trend and outcome of paracetamol poisoning, in relation to increased availability of paracetamol from non-pharmacy outlets in 2009. Patients' serum paracetamol results over 14years (2000-2013) from 20 (out of 21) regions in Sweden were linked to national registers of hospital care, cause of death, and prescriptions. Paracetamol poisonings were defined by serum paracetamol levels, hospital diagnoses, or cause of death. The change in incidence of poisonings following increased availability of paracetamol was analysed by using segmental regression of time series. Of the 12068 paracetamol poisonings, 85% were classified as intentional self-harm. Following increased availability from non-pharmacy outlets, there was a 40.5% increase in the incidence of paracetamol poisoning, from 11.5/100000 in 2009 to 16.2/100000 in 2013. Regression analyses indicated a change in the trend (p<0.0001) but not an immediate jump in the incidence (p=0.5991) following the increased availability. Adjusting for trends in hospital episodes for self-harm, suicides, and the sales volume of paracetamol did not influence the result. All-cause mortality at 30days (3.2%) did not change over time. The incidence of paracetamol poisoning in Sweden has increased since 2009, contrasting the decreased incidence in the period of 2007-2009. The change in trend was temporally associated with the introduction of availability of paracetamol from non-pharmacy outlets but did not appear to be related to sales volume of paracetamol or general trends in self-harm or suicides. © 2017 Commonwealth of Australia. Pharmacoepidemiology and Drug Safety © 2017 John Wiley & Sons, Ltd.