Mark A. Stevenson

The Medical journal of Australia • October 18, 2024

Paul Beggs, Alistair Woodward, Stefan Trueck, Martina Linnenluecke, Hilary Bambrick, Anthony Capon, Zerina Lokmic Tomkins, Jacqueline Peel, Kathryn Bowen, Ivan Hanigan, Nicolas Arriagada, Troy Cross, Sharon Friel, Donna Green, Maddie Heenan, Ollie Jay, Harry Kennard, Arunima Malik, Celia Mcmichael, Mark Stevenson, Sotiris Vardoulakis, Aditya Vyas, Marina Romanello, Maria Walawender, Ying Zhang

The MJA-Lancet Countdown on health and climate change in Australia was established in 2017 and produced its first national assessment in 2018 and annual updates in 2019, 2020, 2021, 2022 and 2023. It examines five broad domains: health hazards, exposures and impacts; adaptation, planning and resilience for health; mitigation actions and health co-benefits; economics and finance; and public and political engagement. In this, the seventh report of the MJA-Lancet Countdown, we track progress on an extensive suite of indicators across these five domains, accessing and presenting the latest data and further refining and developing our analyses. We also examine selected indicators of trends in health and climate change in New Zealand. Our analyses show the exposure to heatwaves is growing in Australia, increasing the risk of heat stress and other health threats such as bushfires and drought. Our analyses also highlight continuing deficiencies in Australia's response to the health and climate change threat. A key component of Australia's capacity to respond to bushfires, its number of firefighting volunteers, is in decline, dropping by 38 442 people (17%) in just seven years. Australia's total energy supply remains dominated by fossil fuels (coal, oil and natural gas), and although energy from coal decreased from 2021 to 2023, energy from oil increased, and transport energy from petrol grew substantially in 2021-22 (the most recent year for which data are available). Greenhouse gas emissions from Australia's health care sector in 2021 rose to their highest level since 2010. In other areas some progress is being made. The Australian Government completed the first pass of the National Climate Risk Assessment, which included health and social support as one of the eleven priority risks, based in part on the assessed severity of impact. Renewable sources such as wind and solar now provide almost 40% of Australia's electricity, with growth in both large-scale and small-scale (eg, household) renewable generation and battery storage systems. The sale of electric vehicles reached an all-time high in 2023 of 98 436, accounting for 8.47% of all new vehicle sales. Although Australia had a reprieve from major catastrophic climate events in 2023, New Zealand experienced cyclone Gabrielle and unprecedented floods, which contributed to the highest displacement of people and insured economic losses over the period of our analyses (ie, since the year 2010 and 2000 respectively). Nationally, regionally and globally, the next five years are pivotal in reducing greenhouse gas emissions and transitioning energy production to renewables. Australia is now making progress in this direction. This progress must continue and accelerate, and the remaining deficiencies in Australia's response to the health and climate change threat must be addressed. There are strong signs that Australians are increasingly engaged and acting on health and climate change, and our new indicator on health and climate change litigation in Australia demonstrates the legal system is active on this issue in this country. Our 2022 and 2023 reports signalled our intentions to introduce indicators on Aboriginal and Torres Strait Islander health and climate change, and mental health and climate change in Australia. Although the development of appropriate indicators is challenging, these are key areas and we expect our reporting on them will commence in our next report.

Research Integrity And Peer Review • January 31, 2025

Aaron H Fletcher, Mark Stevenson

Background: Retractions undermine the scientific record's reliability and can lead to the continued propagation of flawed research. This study aimed to (1) create a dataset aggregating retraction information with bibliographic metadata, (2) train and evaluate various machine learning approaches to predict article retractions, and (3) assess each feature's contribution to feature-based classifier performance using ablation studies. Methods: An open-access dataset was developed by combining information from the Retraction Watch database and the OpenAlex API. Using a case-controlled design, retracted research articles were paired with non-retracted articles published in the same period. Traditional feature-based classifiers and models leveraging contextual language representations were then trained and evaluated. Model performance was assessed using accuracy, precision, recall, and the F1-score. Results: The Llama 3.2 base model achieved the highest overall accuracy. The Random Forest classifier achieved a precision of 0.687 for identifying non-retracted articles, while the Llama 3.2 base model reached a precision of 0.683 for identifying retracted articles. Traditional feature-based classifiers generally outperformed most contextual language models, except for the Llama 3.2 base model, which showed competitive performance across several metrics. Conclusions: Although no single model excelled across all metrics, our findings indicate that machine learning techniques can effectively support the identification of retracted research. These results provide a foundation for developing automated tools to assist publishers and reviewers in detecting potentially problematic publications. Further research should focus on refining these models and investigating additional features to improve predictive performance. Background: Not applicable.

Preventive Veterinary Medicine • February 19, 2025

P Chowdhury, P Hemsworth, A Fisher, M Rice, R Galea, P Taylor, M Stevenson

In intensively managed poultry production systems the term 'smothering' refers to deaths from suffocation that occur as a consequence of piling behaviour where birds crowd together into densely packed groups. Smothering is a non-negligible source of loss in free-range layer hens, having both negative welfare and economic effects. Smothering events are rarely observed and are usually detected by the discovery of groups of dead hens. The aim of this study was to identify risk factors for smothering deaths in three commercial free-range layer poultry farms in Australia. This was a prospective cohort study of poultry flocks managed by three commercial free-range layer farms in eastern Australia. Flocks were enrolled into the study from 1 January 2019-29 March 2021 and were followed until the end of lay or until the end of the study on 31 March 2022, whichever occurred first. Throughout the follow-up period of the study, daily production and weather data, details of flock management and details of the place and time of smothering events were recorded. Time to event (survival) analyses were used to quantify the association between hypothesised risk factors and the number of days in lay at the time of smothering. Shed and bird level characteristics associated with time to event were quantified using a stratified Cox proportional hazards model which included a frailty term to account for birds clustered within sheds within farm. Across the three farms, for every 100 birds placed into a shed, there were 12 deaths over the duration of the production period. Of the 12 deaths per 100 birds, 2 were due to smothering. Our Cox proportional hazards regression analyses showed that the daily hazard of smothering was increased for birds housed in aviary sheds compared with flat-deck sheds (HR 4.0, 95 % CI 1.7-9.7). The daily hazard of smothering mortality was increased on warm, humid and rainy days, and in birds with low fear of humans and high fear of novel objects. Rainy days on which outdoor daily average humidity was greater than or equal to 70 % were associated with a 3.7 (95 % CI 3.5-3.9) fold increase in the daily hazard of indoor smothering deaths, compared with days when outdoor daily average humidity was less than 70 % and no rain. This study provides useful insight into the determinants of smothering in Australian free-range layer hens, in particular risk factors that do not change over time (e.g., shed type) and those that change daily (e.g., weather conditions). This information allows flock management strategies to be adapted accordingly.

Veterinary Journal (London, England : 1997) • May 01, 2025

Ghazanfar Abbas, Hannah Mitchell, Mark Stevenson, Yuet Lee, Abdul Ghafar, Charles Gauci, Hagan Brodziak, Leah Starick, Regina Fogarty, Bernie Gleeson, Dianne Phillips, Tao Wang, Robin Gasser, Abdul Jabbar

This study employed Bayesian latent class analyses to estimate the diagnostic accuracy of faecal egg count (FEC), milk spot detection and the ELISA for detection of Ascaris suum using matched samples from individual pigs in Australia. A total of 251 blood, faecal and liver samples were collected from finisher pigs from four Victorian pig farms. Matched samples (n = 189) were used to compare the three diagnostic tests. The ELISA detected a higher proportion of positive samples (56 %; 95 % CI 48-62 %) compared to milk spot detection (42 %; 95 % CI 35-49 %) and FEC (17 %; 95 % CI 12-23 %). Only the ELISA detected A. suum infections on two of the four farms, with 14 % and 41 % within-farm prevalence estimates. Agreement between diagnostic tests was moderate for FEC and milk spot detection (Cohen's kappa 0.42; 95 % CI 0.30-0.53) and ELISA and milk spot detection (0.52; 0.41-0.64), while fair agreement was observed between FEC and ELISA (0.28; 0.19-0.37). Our latent class analyses identified a higher diagnostic sensitivity for the ELISA (0.92; 95 % CrI [credible interval] 0.86-0.96) than FEC (0.43; 0.34-0.53) and milk spot detection (0.86; 0.79-0.92). A strong association was observed between ELISA outcomes (optical density [OD] and OD ratio [ODr]) and milk spot grades (low, medium, high), with higher OD and ODr values corresponding to an increased number of milk spots on the liver. This study highlights the limitations of conventional A. suum detection methods. Quantitative estimates of the diagnostic sensitivity of the ELISA facilitate its use as a tool for assessing A. suum exposure in pig herds.

Parasites & Vectors • September 27, 2024

Endris Ali, Ghazanfar Abbas, Ian Beveridge, Sandra Baxendell, Berwyn Squire, Mark Stevenson, Abdul Ghafar, Abdul Jabbar

Background: Gastrointestinal parasites such as nematodes and coccidia are responsible for significant economic losses in the goat industry globally. An indiscriminate use of antiparasitic drugs, primarily registered for use in sheep and cattle, in goats has resulted in drug-resistant gastrointestinal parasites. Very little is known about the gastrointestinal parasite control practices used by Australian dairy goat farmers that are pivotal for achieving sustainable control of economically important parasites. The study reported here provides insights into gastrointestinal parasite control practices of Australian dairy goat farmers based on responses to an online survey. Methods: The questionnaire comprised 58 questions on farm demography, husbandry and grazing management, knowledge of gastrointestinal parasites and their importance in dairy goats, diagnosis of infections, antiparasitic drugs and alternate control options. After a pilot survey (n = 15 respondents), a link to the questionnaire was available to all (n = 456) registered members of the Dairy Goat Society of Australia Ltd from 17 April to 16 June 2023. Multiple correspondence analyses (MCA) were performed to explore the association between selected parasite control practices. Results: A total of 66 (14%) respondents completed the questionnaire. Of these, 74% (49/66) observed parasite-related illnesses in their goats; two-thirds of them assessed worms burden using faecal egg counts (FECs), with 26% (39/149) deworming their goats based on the results of the FECs. Most respondents (97%; 183/188) perceived that gastrointestinal parasites caused production losses and ranked Haemonchus contortus as the most important parasite. Anitparasitic drugs were used by 94% (62/66) of respondents, with the most frequently used anthelmintics being a commercial combination of four anthelmintics (levamisole, closantel, albendazole and abamectin), benzimidazoles and macrocyclic lactones. Most respondents (77%; 51/66) were unaware of anthelmintic resistance on their property. MCA results delineated two clusters of gastrointestinal parasites management. Conclusions: This study provides insights into the demography of Australian dairy goat farms, the husbandry and grazing practices used by dairy goat farmers, their knowledge regarding gastrointestinal parasites and their practices for internal parasite control, thereby paving the way for tackling drug resistance in gastrointestinal parasites in dairy goats.