Infectious disease (ID) surveillance plays a crucial role in protecting public health, including detecting disease outbreaks and potential pandemics as well as monitoring ID epidemiology and the impact of interventions.

Aims: 1) To systematically describe all of the ID surveillance systems in Aotearoa New Zealand (NZ) in 2025 according to 13 major functional ID categories and points in the causal path (disease, hazards/protections, interventions, determinants), 2) To compare the current scope of ID surveillance with a review done in 2010 to identify how surveillance has changed over this 15-year period and 3) To identify strengths and gaps in the current scope of ID surveillance where improvements are needed.

Methods and preliminary results: We conducted a review of literature and reports from key agencies responsible for ID surveillance in NZ. Key findings were systematically recorded in tabular form that listed all discrete surveillance systems.
We identified IDs of interest and whether these are currently covered by existing ID surveillance systems and strengths and gaps identified by topic experts and key end users.
We used our novel surveillance sector review framework for reviewing and assessing the performance of these systems and whether they are meeting end user needs.
Our preliminary findings show close to 150 ID surveillance systems are currently functioning in NZ. Over the last 15 years the overall number of ID surveillance systems has increased by more than 30% with increases seen across most categories.

Discussion and initial recommendations: In NZ, ID surveillance is carried out mostly by a range of government agencies. Academic/other groups run around 10% of ID surveillance systems. Some of these systems operate in silos with duplication in some areas and no clear central leadership.
We will present recommendations for achieving a more cohesive approach to ID surveillance that better meets the need of end users.

Background
Respiratory syncytial virus (RSV) is the leading cause of acute lower respiratory infection in young children and a major contributor to paediatric hospitalisations. Comprehensive, population-specific clinical data on Australian children experiencing RSV disease are needed to inform cost-effectiveness analyses, guide policy recommendations, and assess the impact of future interventions. We aimed to describe the burden of RSV-associated hospitalisations and risk factors for severe disease in children admitted to Sydney Children’s Hospitals Network.

Methods
Children under 2 years of age with laboratory confirmed RSV, admitted to The Children’s Hospital at Westmead and Sydney Children’s Hospital, Randwick from 2019 to 2023 inclusive were identified for enrolment. A random sample of 200 children admitted to hospital with RSV but not admitted to ICU were enrolled from each year. All ICU admissions were enrolled. Enhanced clinical data were extracted from hospital medical records. Descriptive statistics, rates and risk indicators were reported and compared between groups.

Results
Between January 2019 and December 2023, 4414 children under 2 years of age were hospitalised with RSV. Of these, 1000 non-ICU and 499 ICU cases were enrolled into the study. ICU admissions represented 11.3% of all RSV-confirmed admissions. Estimated hospitalisation rates among children < 2 years for 2023 were 1504/100,000 and 61/100,000 for non-ICU and ICU respectively. Compared to non-ICU admissions, children in ICU had a lower median age (3 vs 7 months, p<0.001), and a higher proportion children, were born premature (28.0 vs 14.1%, p<0.001), had cardiac (10.0 vs 3.2%, p<0.001) and genetic diseases (6.5 vs 2.4%, p<0.001), or identified as Indigenous (8.6 vs 2.0%, p<0.001).

Conclusions
RSV was responsible for a large number of admissions with a substantial proportion of children under 2 years of age requiring high acuity care. Preventative interventions should be prioritised for children who have characteristics which are overrepresented amongst those with severe disease.

Seasonal influenza and other respiratory viruses such as respiratory syncytial virus (RSV) and SARS-CoV-2 result in substantial annual public health burden worldwide. The magnitude and timing of respiratory virus epidemics is highly variable and difficult to predict. In temperate winter seasons, they can place significant demands on hospital services, particularly if multiple epidemic peaks occur simultaneously. Accurate forecasts of disease burden have the potential to enable a rapid understanding of the status of concurrent epidemics and inform public health responses.
The Australia–Aotearoa Consortium for Epidemic Forecasting and Analytics (ACEFA), ran a pilot multi-team forecasting initiative across the 2024 winter period (9th May 2024 – 12th September 2024) in collaboration with the Department of Health, Victoria. We solicited weekly forecasts from multiple research teams of daily reported cases of influenza, COVID-19, and RSV over a four-week horizon. Four component model forecasts were contributed by three different research groups from across Australia, with a fourth team utilizing the component forecasts to generate ensemble forecasts (making a total of six models, four component models and two ensembles). The performance of each forecast was statistically evaluated against the observed case data.
In this presentation, we first discuss the different approaches each group used to forecast the three diseases. We then present an analysis of the performance of the different models submitted by each team during the winter forecasting period. This analysis includes a discussion of the strengths and weaknesses of the different models throughout the season e.g. if a model performs best during specific phases of an outbreak. Finally, we show how these results can be interpreted, and the implications of the different evaluation metrics used for communicating the results.

Artificial intelligence (AI) presents a transformative approach to public health surveillance, particularly in environments where traditional methods are limited or absent. By leveraging open-source data, AI-driven systems can provide epidemic intelligence, offering early warning signals of infectious disease outbreaks. EPIWATCH is an AI-powered outbreak detection and monitoring system that has demonstrated the ability to identify epidemic signals before official detection by health authorities.

This study evaluates the utility of open-source epidemic intelligence through two case studies. EPIWATCH reports on outbreaks of unspecified influenza-like illness and pneumonia- along with known causes such as influenza A/B, SARS-CoV-2, RSV, pertussis, adenovirus, and Mycoplasma pneumoniae from August to December of 2022 and 2023 were analyzed. In China, EPIWATCH detected an increase in respiratory illness in 2023 compared to 2022, contrasting with a global decline during the same period. Notably, a peak in pneumonia cases was identified from October to early November 2023, preceding the official recognition of Mycoplasma pneumoniae outbreaks by the WHO on 22 November 2023.

To assess EPIWATCH’s utility in conflict zones, we analyzed infectious disease patterns in Ukraine before (November 2021–February 2022) and during the conflict (February–July 2022). Comparison with official sources revealed increased infectious disease reports during wartime, highlighting the system’s potential for real-time epidemic intelligence in crisis settings.

These findings highlight the power of AI-driven surveillance in early outbreak detection, particularly in resource-constrained and conflict-affected regions. By complementing traditional surveillance, AI systems like EPIWATCH can optimize response and preparedness efforts, ultimately improving global health security. Given the acceleration of epidemics in recent years, leveraging open-source intelligence for rapid outbreak detection is essential for timely intervention.

Background: The COVID-19 pandemic led to the unprecedented implementation of self-reported Rapid Antigen Tests (RATs) in Australian disease surveillance systems. In this study we aimed to, compare the epidemiology of RATs-diagnosed versus laboratory-confirmed cases; assess changes in COVID-19 notifications following RAT implementation; and evaluate validity of self-reported demographic data of RAT cases.
Methods: We analysed COVID-19 case notifications from the National Notifiable Disease Surveillance System (NNDSS) between February 26 and October 12, 2022. We stratified cases by sex, age, Indigenous status, state and territory, Socio-Economic Indexes for Areas and remoteness.
Results: Of 7,036,720 COVID-19 notifications to the NNDSS during the study period, 4,314,010 (61%) were RAT-positive cases. Comparable trends between RAT and laboratory-confirmed cases were observed throughout the study period, showing strong concordance and high confidence in the results. Compared with laboratory-confirmed tests, the proportion of RAT usage was higher among those aged 5-19 years (n=983,025; 70%), Indigenous Australians (n=198,557; 72%), those in lower socio-economic areas (n=766,825; 66%) and remote residents (n=59,492; 78%). The proportion of cases diagnosed by RATs varied between states and territories, reflecting potential regional differences in testing strategies and accessibility. Self-reported demographic data demonstrated high completeness across key variables analysed.
Conclusions: RAT implementation successfully expanded testing access, particularly benefiting younger age groups, Indigenous Australians, and residents of disadvantaged and remote areas. The high quality of self-reported data and concordance with laboratory-confirmed cases suggest that RATs and digital self-reporting systems could effectively support future public health surveillance, especially in resource-limited settings.

Context and aim: Respiratory syncytial virus (RSV) is a leading cause of acute lower respiratory tract infections, especially in young children and older adults. Despite its global impact, comprehensive epidemiological data on RSV testing and hospital burden are limited, particularly in regional Australia. Central Queensland (CQ), with its subtropical climate, provides a unique setting to study RSV testing patterns and burden.
Methods & analysis: This study analysed hospital-based RSV surveillance data from CQ, with data collected retrospectively from 2019 to 2021 and prospectively from 2022 to 2024. Individuals admitted to hospitals with laboratory-confirmed RSV or RSV-related diagnoses based on ICD-10-AM codes were included. Descriptive statistics and incidence rate ratios (IRR) were calculated.
Research findings: Between 2019-2024, 43,651 RSV-PCR tests were conducted in CQ, with 5.3%(n=2,344) testing positive. RSV testing and positive cases increased significantly, from 2,633 tests and 130 positive cases in 2019 to 12,389 tests and 658 positive cases in 2024. There were 1,605 RSV-related hospitalisations, with a median age of 1.6 years (IQR: 0.6–46.1 years), 47.9% female, and 34.7% of admissions from infants under 12 months. The hospitalisation rate among infants rose significantly from 15.6 per 1,000 infants in 2019 to 41.2 per 1,000 in 2023 and 2024 (IRR:2.1, 95%CI: 1.8-2.6, p<0.001). The Indigenous population had a significantly higher hospitalisation rate compared to the non-Indigenous population (IRR:2.7, 95%CI: 2.4-3.1, p<0.001). The median length of stay was 2 days, with 20.9% requiring ventilation, 2.1% needing ICU care, and 0.7% resulting in death, mostly among individuals aged 60 and above (83.3%).
Outcomes and future actions: RSV admissions were underreported due to limited testing. Increased awareness and widespread testing during prospective surveillance revealed a significant rise in RSV-related admissions. These findings highlight the need for enhanced RSV testing, better resource allocation, and expanded immunisation efforts to manage the RSV burden effectively.

Background and Aims:
Understanding COVID-19 and influenza severity in children, particularly mortality rates, is essential to inform future public health strategies, including vaccination. Administrative data sources may under- or over-estimate true mortality. To estimate and compare mortality rates attributable to COVID-19 and influenza in the Australian pediatric population.

Methods:
Case series review of children aged <18 years hospitalized with laboratory-confirmed SARS-CoV-2 or influenza infection and recorded as deceased. COVID-19 cases were ascertained by yhe Paediatric Active Enhanced Disease Surveillance (PAEDS) network of eight sentinel hospitals January 2020 – September 2023 and influenza, seasonally, from 2018 – September 2023. Population mortality rates were calculated using deaths notified to Australia’s National Notifiable Diseases Surveillance System (NNDSS). COVID-19 or influenza-laboratory test positive deaths were assessed and categorized for virus role as either: the primary cause; a contributory cause; an unlikely cause; or unable to be determined. Primary and contributory causes of death, deemed “causally related”, were used to calculate an attributable mortality proportion. Crude annual population mortality rates were calculated using NNDSS notifications and Australian Bureau of Statistics (ABS) mid-year population data. Adjusted attributable death rates were estimated using attributable mortality proportions.

Results
In children who died with laboratory confirmed SARS-CoV-2 or influenza infection, attributable in-hospital mortality proportions were 11/19 (58%) and 23/29 (79%) respectively. Among COVID-19 and influenza attributable deaths (age range <1 months – 17 years), 47% (16/34) had no known pre-existing co-morbidity. Across observation periods, the respective crude and adjusted annualized average weighted mortality rates per million population were 1.35 and 0.78 (plausible range: 0.51–1.13) for COVID-19, and 1.34 and 1.06 (plausible range: 0.73–1.15) for influenza.

Conclusions:
In Australia, crude mortality estimates of COVID-19 in children using routine surveillance data may include deaths not attributable to the virus and so overestimate severity. Accurate age-specific attributable mortality rates in children and their association with medical co-morbidity is important to informing vaccination policies and other public health measures for prevention of high burden respiratory viruses in children.

During disease outbreaks, it is crucial to be able to rapidly estimate key disease parameters – such as transmissibility and severity – and understand infection trends to guide decision makers to adopt appropriate countermeasures. Efforts to do this during the COVID-19 pandemic with existing systems highlighted the need to diversify the data streams used in this process in three main ways. Firstly, existing systems relied on estimating disease parameters from First Few Hundred (FFX) studies, but these were slow and difficult to set up. We need to be able to estimate disease parameters from alternative data streams, such as contact tracing data. Secondly, we need to incorporate timeseries data to estimate the underlying infection trends in near-real time. Thirdly, we need the ability to account for behavioural changes in the face of dynamically changing interventions. To address these needs, we have been working with the Department of Defence to build a robust system that fuses different data streams and models to estimate epidemic conditions and parameters, which then flows through to scenario modelling to support decision making. In this talk I will give an overview of this prototype system – called the BioSurveillance Decision Support System (BDSS) – and the different data streams and models it supports. I will illustrate the system pipeline starting from different data streams to disease models, to scenario models and finally to decision outputs, using an example from the COVID-19 Delta outbreak in New South Wales in 2021. Outputs include near-real time estimates of infection prevalence that can be updated frequently, and decision scenarios for longer-term decision making.

During winter months of temperate regions, concurrent epidemics of multiple respiratory pathogens can occur, causing periods of increased clinical burden. Case time series — which are predominantly used to monitor infection levels — can exhibit substantial noise and day-of-the-week effects, limiting the visual interpretation of trends in raw data. However, statistical methods can infer smoothed trends in case time series by quantifying and accounting for different sources of noise. Here we apply statistical models to estimate the epidemic dynamics of SARS-CoV-2, RSV, and influenza subtypes (influenza A H3N2, influenza A H1N1, and influenza B) in Victoria, Australia, over the 2024 winter season. We model trends in daily reported cases and the daily growth rate over time for all pathogens/sub-types. We present: (1) retrospective analyses using the final dataset up to 10 September and (2) weekly real-time analyses from 19 March 2024 to 10 September 2024 using data up to each time-point, including a retrospective performance evaluation. We estimated similar peak timing of SARS-CoV-2 and RSV epidemics in late-May, followed by a H3N2-dominant influenza epidemic, which peaked in early-July. Transient increases in SARS-CoV-2 activity coincided with the emergence of new variants and transient decreases in influenza activity corresponded to the timing of school holidays. Real-time estimates demonstrated good agreement with those produced at the end of the season with significant overlap of the 95% credible intervals. Our findings demonstrate how statistical methods can be implemented in real-time to synthesise noisy case time-series data into interpretable trends (including uncertainty), enabling quantification of the strength of evidence for whether epidemic activity is increasing, stable or declining. Our real-time outputs were reported weekly to the Department of Health, Victoria from June–September 2024, complementing other routine surveillance indicators.

The interim Australian Centre for Disease Control is establishing a National Wastewater Surveillance Program for public health. This program will monitor priority pathogens such as SARS-CoV-2, influenza, polio, and respiratory syncytial virus, and provide capacity for early detection of emerging pathogens with epidemic and pandemic potential such as Japanese encephalitis virus and mpox virus. The program aims to contribute novel surveillance data in near real-time to enhance agile public health response, in line with recommendations from the COVID-19 Response Inquiry Report.
The National Wastewater Surveillance Program commits to three years of national wastewater surveillance and capacity building. During the COVID-19 pandemic, all Australian states and territories implemented wastewater surveillance programs for SARS-CoV-2. The program will collaborate with jurisdictions to develop national capacity and provide opportunity to contribute to international sentinel surveillance networks like the Global Consortium for Wastewater and Environmental Surveillance for Public Health (GLOWACON). Ongoing evaluation of the program will build evidence for the utility of wastewater surveillance data in public health decision-making and policy development.
This session will provide updates on the development of the National Wastewater Surveillance Program, and describe key stages of engagement, design, analysis, and evaluation of the program.
By the end of the session, participants will understand how the National Wastewater Surveillance Program is being developed and recognise how the program will contribute to surveillance of infectious diseases in Australia.

Gonorrhoea, caused by the bacterial etiological agent Neisseria gonorrhoeae, is routinely treated with a single dose of azithromycin and ceftriaxone as single- or dual-therapy. However, since the introduction of antibiotics, N. gonorrhoeae strains have developed resistance to nearly all antimicrobials used to treat gonorrhoea. SA Pathology began genomic surveillance of all culturable clinical N. gonorrhoeae in October 2021, just prior to the end of the COVID-19 social and travel restrictions. This study describes the genomic epidemiology of N. gonorrhoeae in South Australia (SA) during and immediately following the cessation of the COVID-19 public health restrictions.

Approximately 1,208 N. gonorrhoeae isolates from three specimen types (ie. oropharygeal, genital, rectal swabs) were referred to SA Pathology for sequencing between October 2021 and December 2024. Genomes that passed the quality check were further subtyped and characterised. Phenotypic susceptibility testing, performed on all isolates, was compared to genotypic results.

As the state emerged from the pandemic-directed restrictions in November 2021, the N. gonorrhoeae population in SA was dominated by three main sequence types (STs) of susceptible N. gonorrhoeae: ST-7363, ST-7359 and ST-8156. After overseas travelling resumed and increased in 2023, the diversity of N. gonorrhoeae STs also increased. There was a quick expansion followed by the disappearance of N. gonorrhoeae ST-7827 (NG-STAR ST-38) with decreased susceptibility to ceftriaxone, mostly isolated from cases associated with increased risk for infection. In addition, the rpsJ:V57M mutation was associated with a large range of N. gonorrhoeae STs, whereas tetM gene was only identified in N. gonorrhoeae ST-7363 and ST-16676. The change in the genomic epidemiology of N. gonorrhoeae within SA during the study period provides valuable insights to the treating physicians and public health officers and highlights the risk of the emergence and spread of antimicrobial resistance in gonorrhoea in Australia.