Increasing anaphylaxis events in Western Australia identified using four linked administrative datasets

Sandra M. Salter, Ross J. Marriott, Kevin Murray, Samantha L. Stiles, Paul Bailey, Raymond J. Mullins, Frank M. Sanfilippo

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


Background: Anaphylaxis events are increasing worldwide, based on studies of single administrative datasets including hospital admissions, emergency room presentations, and prescription and medical claims data. Linking multiple administrative datasets may provide better epidemiological estimates, by capturing a greater number of anaphylaxis events occurring at the individual level. In this linked data study in Western Australia, we combined 4 population-based datasets to identify anaphylaxis events, factors influencing occurrence, and change in event rates from 2002 to 2013. Methods: Four linked administrative datasets from the Western Australian Data Linkage System were used, representing ambulance attendances, emergency department presentations, hospital inpatient admissions and death registrations. An anaphylaxis cohort was identified using ICD-9-CM, ICD-10-AM and additional anaphylaxis diagnosis codes, with event rates calculated. We explored the impact of age, gender, cause, Indigenous status and socioeconomic index on event rates. Standard Poisson regression models were used to examine the significance of the change in anaphylaxis event rates over time. Results: A total 12,637 individuals (mean age 31.8 years, 49.6% female) experienced 15,462 anaphylaxis events between 2002 and 2013 (97.5% in non-Indigenous patients and 59.5% residing in the area of greatest socioeconomic advantage). Anaphylaxis event rates increased from 15.4 to 82.5/105 population between 2002 and 2013. The greatest increase in anaphylaxis events was seen in those coded as unspecified anaphylaxis (all ages, males and females combined, p < 0.001), with the highest rates of unspecified anaphylaxis in males 0–4 years (171.9/105 population in 2013), and females 15–19 years (104.0/105 in 2013). The average annual percent increase (95% CI) for food-related anaphylaxis was 9.2% (6.6–12.0); for medication-related anaphylaxis was 5.8% (4.5–7.1); and for unspecified anaphylaxis was 10.4% (9.8–11.0); all p < 0.001. There was a significant increase in ambulance attendance, emergency presentations and inpatient admissions for anaphylaxis between 2002 and 2013, with emergency presentations (56.0/105 population), inpatient admissions (43.2/105), and ambulance attendance (21.6/105) highest in 2013. Only 25 anaphylaxis-related deaths were recorded in the mortality register with no significant change in rates over time. Conclusion: Using multiple linked administrative datasets, we identified significantly higher rates of total anaphylaxis than previously reported, with more than 5-fold increases in anaphylaxis events between 2002 and 2013. While the combination of 4 population-level datasets provides a more comprehensive capture of cases, even at the individual dataset level, admission rates for anaphylaxis in Western Australia are substantially higher than those previously reported for similar time periods, both in Australia and worldwide.

Original languageEnglish
Article number100480
JournalWorld Allergy Organization Journal
Issue number11
Publication statusPublished - Nov 2020


Dive into the research topics of 'Increasing anaphylaxis events in Western Australia identified using four linked administrative datasets'. Together they form a unique fingerprint.

Cite this