Increases in anaphylaxis fatalities in Australia from 1997 to 2013

Recent epidemiological studies indicate increases in Australian, UK and US hospital anaphylaxis admission rates.

The aim of this study was to determine whether Australian anaphylaxis fatalities are increasing in parallel and to examine the characteristics of fatalities recorded in the National Coronial Information System (NCIS).

Time trends in Australian anaphylaxis fatalities were examined using data derived from the Australian Bureau of Statistics (ABS) 1997-2013 and the NCIS 2000-2013, the latter providing additional information to verify cause and identify risk factors.

The ABS recorded 324 anaphylaxis fatalities by cause: unspecified (n = 205); medication (n = 52); insect stings/tick bites (n = 41); food (n = 23); and blood products (n = 3). From 1997 to 2013, all-cause fatal anaphylaxis rates increased by 6.2%/year (95% CI: 3.8-8.6%, P < 0.0001) or from 0.054% to 0.099/10(5) population. Fatal food anaphylaxis increased by 9.7%/year (95% CI: 0.25-20%, P = 0.04) and unspecified anaphylaxis deaths by 7.8% (95% CI: 4.6-11.0, P < 0.0001). There was an insignificant change in medication-related fatalities (5.6% increase/year; 95% CI: 0.3% decrease to 11.8% increase, P = 0.06), and sting/bite fatalities remained unchanged. Hospital anaphylaxis admission rates for all-cause, food, unspecified and medication anaphylaxis increased at rates of 8%, 10%, 4.4% and 6.8%/year, respectively. A total of 147 verified NCIS deaths were examined in detail: medication- and sting/bite-related fatalities occurred predominantly in older individuals with multiple comorbidities. Upright posture after anaphylaxis was associated with risk of sudden death (all causes). Seafood (not nuts) was the most common trigger for food-related anaphylaxis deaths.

Australian anaphylaxis fatality rates (most causes) have increased over the last 16 years, contrasting with UK- and US-based studies that describe overall lower and static overall anaphylaxis fatality rates (0.047-0.069/10(5) population).