INTRODUCTION: It is well-established that heatwaves increase demand for emergency transport in metropolitan areas; however, little is known about the impact of heat events on demand for prehospital retrieval services in rural and remote areas, or how heatwaves are defined in this context. INCLUSION CRITERIA: Papers were eligible for inclusion if they reported on the impact of a heat event on the activity of a prehospital and retrieval service in a rural or remote area. METHODS: A search of PubMed, Cochrane, Science Direct, CINAHL, and Google Scholar databases was undertaken on August 18, 2020 using search terms related to emergency medical transport, extreme heat, and rural or remote. Data relevant to the impact of heat on retrieval service activity were extracted, as well as definitions of extreme heat. RESULTS: Two papers were identified, both from Australia. Both found that heat events increased the number of road ambulance call-outs. Both studies used the Excess Heat Factor (EHF) to define heatwave periods of interest. CONCLUSIONS: This review found almost no primary literature on demand for prehospital retrieval services in rural and remote areas, and no data specifically related to aeromedical transport. The research did recognize the disproportionate impact of heat-related increase in service demand on Australian rural and regional health services. With the effects of climate change already being felt, there is an urgent need for more research and action in this area.

Rural and remote patients have reduced access to palliative care, often resulting in interhospital transfers and death a long way from home and family. Katherine Hospital, a 50-bed hospital services a population with high Aboriginality who experience this issue. To characterise trends in mortality and transfers at a remote hospital in reference to increasing capacity to provide palliative care. Retrospective analysis of deaths in patients over 18 years of age, admitted between 2008-2018 at Katherine Hospital, Northern Territory. Outcome measures include number of deaths, aeromedical transfers to tertiary facility, palliative care episodes, demographics including Aboriginality, admission data and comorbidity. Statistical analysis included unpaired t-test, chi-square test and regression analysis. The number of deaths in Katherine Hospital increased from 23 (0.88% of inpatient admissions) in 2011 to 52 in 2018 (1.7%). During the same period, the proportion of all deaths classified as palliative increased from 51.4% to 66.0% (p=0.001), with fewer deaths occurring in the emergency department (17.2% to 1.4% for the last three years, R=0.75, p=0.008). The number of aeromedical transfers of patients from Katherine Hospital to tertiary centres decreased from 769 (10.4% of all admissions) in 2011 to 434 (3.4%) in 2018 (p=0.006). Increasing the capacity of a remote hospital to provide palliative care allowed more patients to die closer to home and decreased inappropriate aeromedical retrievals. An increased in-hospital mortality rate should not be misinterpreted as reflecting suboptimal care if palliative intent, patients' wishes and non-clinical risk factors have not been ascertained. This article is protected by copyright. All rights reserved.

To determine the clinical and environmental variables associated with frequent presentations by adult patients to a remote Australian hospital emergency department (ED) for reasons other than chronic health conditions. Unmatched case-control study of all adult patients attending Katherine Hospital ED between 1 January and 31 December 2012. Cases were defined as frequent attenders (FAs) without a chronic health condition who presented to the ED six or more times during the 12-month period. A single presentation was randomly selected for data collection. Controls were patients who presented on only one occasion. Basic demographic data were collected, including clinical outcomes, Indigenous status, living arrangements, and whether alcohol and violence contributed to the presentation. Environmental variables were extracted from the Bureau of Meteorology database and mapped to each presentation. FAs were much more likely to be homeless (odds ratio [OR], 16.4; P < 0.001) and to be Aboriginal (OR, 2.16; P < 0.001); alcohol as a contributing factor was also more likely (OR, 2.77; P = 0.001). FAs were more likely to present in hotter, wetter weather, although the association was statistically weak. Clinical presentations by cases and controls were similar; the annual death rates for both groups were high (3.6% and 1.5%, respectively). There was a strong association between FA and Aboriginal status, homelessness and the involvement of alcohol, but alcohol was more likely to contribute to presentation by non-Aboriginal FAs who had stable living conditions. FAs and non-FAs had similar needs for emergency medical care, with strikingly higher death rates than the national average in both groups. As a result of this study, Katherine Hospital has initiated a Frequent Attender Pathway that automatically triggers a dedicated ED service for those at greatest clinical risk. Homelessness is a serious problem in the Northern Territory, and is associated with poor health outcomes.

Classifying free-text from historical databases into research-compatible formats is a barrier for clinicians undertaking audit and research projects. The aim of this study was to (a) develop interactive active machine-learning model training methodology using readily available software that was (b) easily adaptable to a wide range of natural language databases and allowed customised researcher-defined categories, and then (c) evaluate the accuracy and speed of this model for classifying free text from two unique and unrelated clinical notes into coded data. A user interface for medical experts to train and evaluate the algorithm was created. Data requiring coding in the form of two independent databases of free-text clinical notes, each of unique natural language structure. Medical experts defined categories relevant to research projects and performed 'label-train-evaluate' loops on the training data set. A separate dataset was used for validation, with the medical experts blinded to the label given by the algorithm. The first dataset was 32,034 death certificate records from Northern Territory Births Deaths and Marriages, which were coded into 3 categories: haemorrhagic stroke, ischaemic stroke or no stroke. The second dataset was 12,039 recorded episodes of aeromedical retrieval from two prehospital and retrieval services in Northern Territory, Australia, which were coded into 5 categories: medical, surgical, trauma, obstetric or psychiatric. For the first dataset, macro-accuracy of the algorithm was 94.7%. For the second dataset, macro-accuracy was 92.4%. The time taken to develop and train the algorithm was 124 minutes for the death certificate coding, and 144 minutes for the aeromedical retrieval coding. This machine-learning training method was able to classify free-text clinical notes quickly and accurately from two different health datasets into categories of relevance to clinicians undertaking health service research.