Edwina Mead

Background and objective: The unprecedented growth in medical literature limits the feasibility of traditional systematic review methods, particularly in high-volume clinical fields like maternity care. To address this meta-scientific challenge, this paper presents a novel methodological approach using a hybrid workflow that combines Large Language Model (LLM) semantic data extraction with programmatic filtering to automate title and abstract screening. Methods: To provide a critical evaluation of this LLM-assisted approach, the hybrid workflow was deployed on a high-volume clinical dataset (51,281 records retrieved from six databases) as a validation case study. Google Gemini 1.5-pro and 2.5-flash were used to extract 12 structured semantic variables from abstracts, followed by a seven-stage rule-based filtering pipeline to determine eligibility. To ensure transparent reporting, the workflow’s diagnostic accuracy was validated against a large-scale manual audit to calculate sensitivity, specificity, and error rates. Results: The hybrid workflow reduced the human manual screening burden by 96% while maintaining accuracy comparable to dual-human review. Validation against a demonstrated an estimated sensitivity of 89.9% (95% confidence interval; CI 87.3-92.1%) and specificity of 96.4% (95% CI 96.2-96.5%), with a false negative rate of 0.18% (95% CI 0.09-0.35%). This automated phase was completed at a total API cost of US$458.73. Subgroup analysis found no change in screening accuracy after changing LLM. Discussion: This approach offers a scalable, transparent, cost-effective, and model-agnostic solution for evidence synthesis in high-volume medical fields. By decoupling data extraction from inclusion logic, this method allows researchers to maintain granular control over complex eligibility criteria, facilitating the development of living reviews as global research output continues to expand.

Issue Addressed: Learning to swim reduces drowning risk and facilitates access to lifetime health benefits of aquatic activity. Amid cost‐of‐living pressures and disparities in swimming lesson access, further exacerbated by the COVID‐19 pandemic, governments have invested in strategies to foster access and participation. This study examines the swim school industry's perceived impacts of the NSW government First Lap voucher program to subsidise swimming lessons for young children. Methods: A mixed‐methods approach analysed anonymous survey responses of 100 swim schools. Quantitative analysis calculated frequencies with 95% confidence intervals (95% CI), and thematic analysis was applied to open‐ended qualitative responses. Results: Most providers (63%) reported increased enrolment (95% CI 53%–72%) and 58% noted boosted income (95% CI 48%–67%) since program commencement. Sixty‐nine percent observed more new enrolments (95% CI 60%–78%), leading to additional classes (57%; 95% CI 47%–67%) and teaching staff hours (59%; 95% CI 49%–68%). Increased pool space usage (36%; 95% CI 27%–46%) and non‐teaching staff hours (36%; 95% CI 26%–45%) were noted. Key themes derived from open‐ended question responses included swim schools' operational impacts (such as promotional opportunities, student retention and increased administration) and family effects, such as improved affordability and earlier enrolment. Conclusions: Findings highlight benefits such as increased enrolment and income, alongside industry challenges in staffing and resourcing. Future research should explore voucher programs' impact on participant retention and long‐term enrolment outcomes. So What?: This study details impacts of a population‐level voucher program on the aquatic industry, addressing a knowledge gap around the role and views of industry, which can inform future health promotion strategies.

Background Aquatic incident location is vital for determining drowning risk and therefore, prevention. We investigated a method to determine drowning incident location using open-source data, addressing a gap often present in existing data systems. Methods Global Positioning System (GPS) data for drowning-related ambulance attendances in New South Wales from 2010 to 2022 were obtained. OpenStreet Map (OSM) was used to overlay mapped water bodies like beaches, pools and natural waterways. We tested our method for matching water features from OSM, considering location differences, duplicate and overlapping entities and using drowning fatality data from 2002 to 2022 to develop a method to prioritise location when multiple locations were identified. Results Among 4161 drowning-related ambulance attendances, 3588 (86.2%) had a valid GPS location. We searched for water features within 200 m for major cities/inner regional areas, and within 500 m for outer regional/remote areas, identifying a water feature for 2579 (71.9%) locations. For most (1221; 47.3%), a single water feature was found, while 574 (22.3%) had three or more, to a maximum of eight. The aquatic location of 6.6% of incidents changed after applying the prioritisation method. Discussion Our method supplemented ambulance data with water body. This study highlights the need for a consistent approach to recording GPS data for ambulance attendances. While based on drowning fatalities, our prioritisation method resulted in few changes, emphasising the value of capturing detailed location data in other sources. Conclusions Using GPS and open-source data, this study demonstrated the feasibility of data linkage to improve insights to inform targeted drowning prevention strategies.

Background Water safety awareness and swimming lesson participation can reduce drowning risk among children in Australia. Providing financial support for swimming lesson participation could additionally impact knowledge and awareness among parents/carers of the importance of learning to swim as a water safety strategy. Between 2021 and 2023, the New South Wales First Lap voucher programme provided preschool-aged children (ages 3–6 years) with two $A100 vouchers, one per financial year, for swimming lessons. This study aimed to examine parent/carer changes in water safety knowledge and awareness during this programme implementation period. Methods This cohort study used survey responses from parents/carers who applied for First Lap vouchers and completed the survey in two consecutive years of the programme. Parents/carers were asked about their knowledge and awareness of five evidence-based strategies to help keep children safe around water, selecting ‘yes’ or ‘no’ in response to each strategy. Multinomial regressions examined changes in knowledge and awareness of parents, adjusted for sociodemographic variables. Results Parents/carers who spoke a non-English language at home were less likely to have or acquire knowledge on water safety strategies. Conversely, parents/carers living in regional/remote areas, having younger children or whose children had previously participated in swimming lessons were more likely to have or acquire water safety knowledge. Conclusion The increases in knowledge of supervision as a water safety strategy in some population subgroups in this cohort of parents/carers who applied for First Lap programme vouchers suggest that the programme may have contributed to this favourable change.

Background Child swimming lessons are a key intervention for drowning prevention; however, participation has been severely affected by COVID-19 restrictions and cost-of-living challenges. To encourage re-engagement, the New South Wales government began providing A$100 swimming lesson vouchers. Parent/carers of eligible preschool-aged children were asked to complete a survey during voucher creation. This study aimed to examine pre-existing challenges to lesson participation among families who had not participated in the preceding 12 months. Methods Cross-sectional data from parent/carer surveys completed between December 2021 and June 2022 were analysed. Binary logistic models analysed associations between sociodemographic factors and answered affirmatively to one or more challenges to participation from a predefined list. Results Of 221 218 vouchers created, 79 553 parent/carers (36%) indicated that their child had not participated in swimming lessons in the last 12 months and responded to the question about participation challenges. Parent/carers of children with disabilities or residing in low socioeconomic areas had higher odds of indicating cost was a challenge while regional/remote families had over five times higher odds of indicating swim school availability difficulties. Families speaking a non-English language at home had over 2.5 times higher odds of thinking that their child was too young for swimming lessons, higher odds of thinking swimming lessons were not important and higher odds of indicating that COVID-19 had challenged participation. Conclusion These findings highlight how different population groups experience challenges to participation in swimming lessons. Alleviating costs and increasing lesson availability is important.