One storm front in Melbourne produced a 672% surge in respiratory emergencies

On the evening of 21 November 2016, a gust front crossed Melbourne at 1800 and the city's emergency system was overwhelmed within hours. Grass pollen concentrations that day were above 100 grains per cubic metre; the front dropped the temperature by 10°C, pushed humidity above 70% and concentrated particulate matter [s1]. Within 30 hours there were 3,365 excess respiratory-related emergency department presentations — 672% above the previous three years — and 476 excess asthma-related hospital admissions, a 992% increase [s1]. Ten people died [s1].

The multidisciplinary investigation that followed is the most detailed account of an epidemic thunderstorm asthma event anywhere, and its most useful finding is about who was affected.

Most of them did not have asthma

Researchers contacted patients who presented during the event at eight metropolitan health services by telephone questionnaire, obtaining data from 1,435 of 2,248 emergency department presentations, a 64% response [s1]. Mean age was 32.0 years (SD 18.6) and 56% were male [s1].

Only 28% had current doctor-diagnosed asthma [s1].

That single figure reorganises the problem. A public health response built around people known to have asthma — targeted messaging, medication reminders, preventer adherence campaigns — would have reached 28% of the people who ended up in an emergency department [s1]. The rest were, as far as the health system knew, not asthma patients at all. The investigation set out to inform both mechanisms and preventive strategies, and it is the second of those that the finding bears on most directly [s1].

The risk was concentrated by ethnicity

Excess presentations and admissions fell disproportionately on individuals of Indian or Sri Lankan birth (10% of admissions versus 1%, p<0.0001) and in south-east Asia (8% versus 1%, p<0.0001) compared with the previous three years [s1]. Among the questionnaire respondents, 39% were of Asian or Indian ethnicity, against 25% of the Melbourne population at the 2016 census — a relative risk of 1.93 (95% CI 1.74 to 2.15, p<0.0001) [s1].

The disparity extended to deaths. Of the ten people who died, six were of Asian or Indian ethnicity, a relative risk of 4.54 (95% CI 1.28 to 16.09, p=0.01) [s1]. With ten deaths the confidence interval is enormous, and the point estimate should not be read as a precise multiplier — but the association is present in both the admissions data and the mortality data.

The study does not establish why. Ethnicity here is a marker for something the analysis did not measure, which could include grass-pollen sensitisation patterns, duration of residence, prior exposure history, healthcare access, or interactions among these.

Who ended up in intensive care

Thirty-five people were admitted to an intensive care unit [s1]. All of them had asthma; 12 were taking inhaled preventers; five died [s1]. So the population that presented was mostly people without an asthma diagnosis, while the population that deteriorated to critical care was entirely people with one [s1].

The authors' summary of the risk factors is that Asian or Indian ethnicity and current doctor-diagnosed asthma portended life-threatening exacerbations such as those requiring ICU admission [s1].

What the event was and was not

The investigation describes convergent environmental factors producing an epidemic of unprecedented magnitude, tempo, geographical range and severity, and setting a new benchmark for emergency and health service escalation [s1]. It is a single event, in one city, and no case series of one can establish how often this happens or under what threshold conditions.

Its lessons are framed as public health ones: event forecasting, coordination of the healthcare response, protection of at-risk populations, and medical management [s1]. Forecasting in particular depends on knowing both the pollen count and the meteorology in advance, and this event supplies the values at which the two combined catastrophically — grass pollen above 100 grains per cubic metre, a 10°C temperature drop, humidity above 70% [s1].

The connection to a lengthening pollen season

Epidemic thunderstorm asthma requires a high grass pollen load and the right storm at the right moment. The first of those is changing measurably. Across 60 North American stations between 1990 and 2018, pollen seasons lengthened by 20 days and pollen concentrations rose 21%, with human forcing of the climate system contributing about 50% of the season trend and about 8% of the concentration trend [s2].

That is a different continent and a different set of taxa, and it says nothing directly about Australian grass pollen. What it establishes is that the exposure side of the equation — how much pollen is in the air, and for how many days of the year it is there — is not stationary. Whether that translates into more epidemic events depends on storm frequency and timing, which this evidence does not address.

Sources

• The Melbourne epidemic thunderstorm asthma event 2016: an investigation of environmental triggers, effect on health services, and patient risk factors — The Lancet Planetary Health, 2018-06-04
• Anthropogenic climate change is worsening North American pollen seasons — Proceedings of the National Academy of Sciences, 2021-02-08