Wildfire smoke and emergency department surveillance

Wildfire smoke can change demand for urgent care before the exposure is fully characterised. Monitoring wildfire smoke impact using emergency department respiratory complaint counts gives public-health teams an early view of how communities are responding to fine particles, ozone and other irritants. Instead of waiting for laboratory confirmation, teams can watch cough, wheeze, breathlessness, chest tightness and asthma or COPD exacerbations as they appear in triage notes and clinical encounters. The signal is not proof that smoke caused every visit; it is a timely indication that symptoms, medication demand and community exposure may be moving together.

Australian smoke episodes often cross borders and affect both dense cities and regional towns. Smoke from a northern NSW fire can reach Sydney, while weather and bushland conditions create different patterns in Melbourne, Brisbane, Perth, Adelaide, Canberra and Hobart. Residents may close windows, run air conditioners, avoid outdoor exercise, keep children indoors, or buy reliever medicines and masks. These habits alter exposure and care-seeking, so respiratory counts work best alongside air-quality measurements, pharmacy data, ambulance calls, school absenteeism and aged-care reports.

Why respiratory counts matter

Although smoke is not an infectious agent, syndromic surveillance can reveal population stress before a specific cause is confirmed. Smoke exposure is not limited to visible haze. PM2.5 can penetrate deep into the lungs, while ozone and volatile organic compounds may irritate airways. Symptoms can appear within hours or days. Children, older adults, pregnant people, outdoor workers, and people with asthma or COPD may be more vulnerable, as may some Aboriginal communities facing high smoke exposure and barriers to timely care. An increase in acute respiratory presentations can therefore warn of population stress before every case receives a specific diagnosis.

An ED respiratory count can combine the stated reason for attendance, triage category, clinician-assessed symptoms and discharge diagnoses. Analysts might group records under cough, wheeze, shortness of breath, asthma exacerbation or respiratory infection, while retaining the original terms. This syndromic approach is flexible because smoke symptoms overlap with viral illness, pollen, dust and heat. It should be treated as a signal for investigation and prevention, not as a replacement for clinical diagnosis or laboratory testing.

The most useful measure is often excess visits: observed presentations minus the number expected from the same weekday, season and recent baseline. Teams can calculate rates per 10,000 residents, rolling seven-day trends, age-specific changes and neighbourhood clusters. A sustained excess is more informative than a one-day spike, although a sharp rise may still require early protection. Each chart should show the denominator, baseline period, data completeness and uncertainty.

Building an Australian smoke signal

Emergency departments are one part of a wider system. A network can combine hospital and clinic data with ambulance dispatch, pharmacy sales, school absences, residential aged-care notifications and laboratory results. Japan's multi-channel approach illustrates the value of combining partial signals rather than trusting a single feed. In Australia, the design must bridge public and private providers, state reporting systems, different coding practices and delays between an event, an attendance and a data extract.

Data lineage should be explicit. An unrelated resource such as online progressive pokies can sit on a broad information site, but it should not enter a respiratory dashboard as an epidemiological feed. More generally, every input needs a named owner, a health-data definition, a refresh schedule and a documented reason for inclusion. This provenance check prevents an unrelated editorial page, duplicate extract or commercial feed from being mistaken for a smoke-related observation.

Australian geography makes linkage especially important. The northern dry season and southern spring and summer produce different smoke patterns, and a plume can travel across several state borders. Hospitals in Sydney, Melbourne, Brisbane, Perth and regional centres may see the same event at different times. Mapping counts against PM2.5 monitors, satellite observations, wind records and fire warnings improves context, while local knowledge is needed because one monitor cannot represent every neighbourhood. State and territory environmental-protection and emergency-management rules also shape who issues alerts and how warnings move between agencies.

Reading an increase without overclaiming

A rise in respiratory complaints can have several explanations: smoke, influenza or another virus, pollen, cold weather, heat, changed access to primary care, or altered coding. During severe smoke, some people may delay GP visits and seek emergency care instead, increasing urgent presentations without proving a new disease outbreak. Surveillance should describe this pattern and test competing explanations rather than assign a cause from timing alone.

Build the baseline from previous weeks, months and years, adjusting for day of week, school holidays, public holidays and seasonal patterns. Compare observed and expected counts, then examine age, sex, postcode, triage acuity and symptom group. Pool small areas or use minimum-count rules to avoid noisy rates. A predefined alert threshold, such as a sustained excess above a baseline percentage, is more defensible than choosing a threshold after seeing the data.

Data quality belongs beside the estimate. A new extraction, delayed transmission, missing free-text fields or a change from one coding system can create an artificial jump. Publish the event date, arrival date, completeness and known limitations. The signal gains weight when ED counts move with ambulance dispatches, pharmacy sales and school absences, and when clinicians confirm that the symptoms are consistent with smoke exposure.

Escalation during extreme smoke days

An alert system should turn rising counts into proportionate action. Public-health teams can intensify checks on high-risk households, share advice through state health departments, Healthdirect 1800 022 222 and local councils, and coordinate with ambulance services, pharmacies, schools and aged-care providers. Hospitals can review staffing and respiratory-care capacity, while primary-care services receive a clear summary of exposure and referral pathways. Escalation should follow a documented threshold and a named decision-maker.

Event-based surveillance offers a useful model for unusual exposure periods. The Tokyo Olympics example described in enhanced monitoring in action shows how temporary increases in testing, reporting and analysis can reveal changes routine systems might miss. Australia can adapt that model for severe smoke episodes or major events in Sydney, Melbourne or other cities: refresh data more often, add relevant symptom categories and issue short situation reports to agencies that need to act.

Public messages should be practical and locally credible. Residents can reduce strenuous outdoor activity, keep windows closed when outdoor air is worse, use available filtration, and follow official advice about well-fitting P2/N95 masks. People using reliever medicines should follow their prescribed plan and seek urgent help for severe breathlessness. Schools, workplaces, community organisations and multilingual media can reach families who may never see a surveillance dashboard, including older neighbours and people with limited transport.

Coding, privacy and data quality

Emergency departments use different fields for the same illness. Triage notes may contain free text, while formal records use ICD-10-AM, SNOMED CT-AU or local diagnosis codes. Analysts should count both structured and narrative signals, document mapping rules, and retain an other respiratory category rather than forcing uncertain cases into asthma or infection. A code is useful for consistency, but it can miss symptoms and can change when hospitals update their systems.

Experience with other syndromic conditions shows why coding must be interpreted carefully. The discussion of opioid mental status codes illustrates how emergency-department mental-status fields can support opioid-poisoning detection while exposing the limits of nonspecific labels. Smoke surveillance should learn from that lesson: publish the code set, test sensitivity, retain uncertainty and never infer a toxic exposure from a single administrative field.

Privacy and governance are especially important when small geographic cells, free text or linkage across agencies are involved. The Privacy Act 1988 and Australian Privacy Principles provide a national framework for handling personal information, while states and territories add health-record, emergency-management and research requirements. Data should be minimised, de-identified where possible, access-controlled and retained only as long as needed. Communities affected by smoke, including Aboriginal communities, should be involved in decisions about how small-area results are communicated.

From detection to public health action

A workable response follows a short chain: detect an excess, verify data quality, compare air-quality and clinical sources, investigate the affected areas, communicate risk, and review whether actions reduced harm. Each step needs a time stamp and responsible agency. A dashboard alone will not protect people; it must connect the signal to stock checks, outreach, clinical guidance and a clear process for escalating concerns.

Local coordination can make the response faster. Community pharmacies—including Chemist Warehouse, Priceline and independent chemists—can report unusual demand within privacy-preserving limits, while schools and aged-care services can flag clusters of absence or respiratory symptoms. Ambulance services, public hospitals, councils, state laboratories and Aboriginal health services may all hold a different piece of the picture. Sharing de-identified summaries through established channels is more useful than creating a new reporting burden for every provider.

Evaluation closes the loop. After a smoke event, teams can compare the timing of alerts with outdoor activity, ambulance calls, pharmacy sales, hospital admissions and later respiratory diagnoses. They can document false alarms, missed cases and community feedback, then adjust thresholds and messaging. This review supports trust and makes the next season safer, especially when smoke crosses jurisdictions or coincides with another respiratory outbreak.

Public-health teams, hospitals, pharmacies and community organisations can act on this approach now. Establish a baseline respiratory signal, connect it with air-quality and social-care information, set a clear escalation threshold, and rehearse who will contact whom when PM2.5 and urgent respiratory presentations rise together. Keep public messages simple, protect privacy, and make advice usable for households, schools and workplaces across Australia. When the next smoke day arrives, a coordinated count can reveal pressure on the health system early enough to support people before illness becomes severe. Use the signal to trigger prevention and verification, not to label communities or attribute every respiratory presentation to smoke without evidence. Build one trusted, cross-sector response around respiratory complaints, cleaner air and timely care.