The relative toxicity of medicines detected after poisoning suicide deaths in Australia, 2013–19: a data linkage case series study

Abstract

Objective

To compare the toxicity (relative to population use) and lethality (relative to poisoning events) of medicines involved in poisoning suicides in Australia; to determine the proportions of cases in which the medicines had recently been dispensed to the deceased person.

Study design

Case series study; analysis of linked National Coronial Information System (NCIS) and Pharmaceutical Benefits Scheme (PBS) data.

Setting, participants

Closed coronial cases for deaths of people aged ten years or older deemed to have been medicine poisoning suicides (including multiple cause deaths), Australia, 1 July 2013 – 10 October 2019, with recorded post mortem toxicology findings.

Main outcome measures

Fatal toxicity index (FTI): deaths per million years of use at the defined daily dose in Australia (2013–2015); proportion of FTI attributable to medicines dispensed to the deceased person during the twelve months preceding their death; estimated case fatality: deaths per number of calls to poisons information centres regarding the medicine (based on the number of calls to the NSW Poisons Information Centre, 2013–2017).

Results

During 2013–19, 2132 deaths were classified as medicine poisoning suicide deaths (median age, 51 years [interquartile range, 39–64 years]; 1036 girls or women [49%]). The 5703 detected substances deemed to have contributed to death included 140 medicines. The overall FTI was 32.0 (95% confidence interval [CI], 30.6–33.3) deaths per million years of use; overall estimated case fatality was 1.28% (95% CI, 1.23–1.34%) of poisoning events. FTI and estimated case fatality (each log₁₀ transformed) were moderately correlated (R² = 0.66). Both values were relatively high for most opioids, sedative psychotropics, and tricyclic antidepressants. Specific medicines with high values were phenobarbitone, oxycodone, morphine, clonazepam, nortriptyline, and propranolol; they were relatively low for risperidone and lithium. The proportions of opioids and hypnosedatives that had been recently dispensed to the deceased persons were smaller than for antidepressant, antipsychotic, and antiepileptic medicines.

Conclusions

To reduce the risk of suicide, access to medicines of greater toxicity and lethality should be restricted, including by staged supply (regular supply of medicines in limited quantities), and limiting pack sizes; real-time prescription monitoring could detect and minimise stockpiling.