Investigating healthcare contacts of Dialysis patients by age and gender.

BACKGROUND: The objective of this paper is to utilise a clinical costing system to investigate differences in the patient journey, defined as the sequence and timing of contacts with the Gold Coast Hospital and Health Services (GCHHS), for four dialysis patient groups defined based on age and gender. It is hypothesised that frequency of contact and form of contact will differ based on both gender and age. METHODS: Data were provided for 393 patients discharged from the GCHHS facility with dialysis treatment between the 1st of January 2015 and the 31st of December 2016. Features extracted from the data included the number and type of contacts (inpatient admissions, outpatient appointments, and emergency department presentations), the likelihood of subsequent contact types, and time spent in and between contact types. Likelihoods of subsequent contact types were estimated by treating the sequence of contacts observed for each patient as a Markov chain and estimating transition probabilities. RESULTS: Differences in patient journey were most prominent when considering age differences, with older patients being characterised by a greater volume of average contacts over the two-year period. The larger volume of average contacts was attributable to shorter times between all types of contacts with the GCHHS as well as an increased volume of inpatient admissions for older patients. Patient journeys did not consistently differ by gender, though some isolated differences were noted for older female patients relative to older male patients. CONCLUSIONS: Different patient groups are characterised by different patient journeys, and better understanding these differences will facilitate improved management of the resources required to service these patients. Clinical costing systems represent a valuable and easily accessible source of data for formulating institution-specific expectations of healthcare utilisation for different groups.

Empirical validation of ELM trained neural networks for financial modelling.

The purpose of this work is to compare predictive performance of neural networks trained using the relatively novel technique of training single hidden layer feedforward neural networks (SFNN), called Extreme Learning Machine (ELM), with commonly used backpropagation-trained recurrent neural networks (RNN) as applied to the task of financial market prediction. Evaluated on a set of large capitalisation stocks on the Australian market, specifically the components of the ASX20, ELM-trained SFNNs showed superior performance over RNNs for individual stock price prediction. While this conclusion of efficacy holds generally, long short-term memory (LSTM) RNNs were found to outperform for a small subset of stocks. Subsequent analysis identified several areas of performance deviations which we highlight as potentially fruitful areas for further research and performance improvement.

Methods of performance analysis in women's Australian football: a scoping review.

Background: The first women's Australian football (AF) professional competition was established in 2017, resulting in advancement in performance analysis capabilities within the sport. Given the specific constraints of women's AF, it is currently unclear what match-play performance analysis methods and techniques are implemented. Therefore, the aim of this scoping review was to describe and critically appraise physical, technical, and tactical performance analysis methods that have been employed in women's and girls' AF match-play. Methodology: A systematic search was conducted on the 27th of June 2022 through five databases. Eligibility criteria were derived from the PCC framework with the population (P) of women and girls AF players, of any level of play; concepts (C) of interest which were measures, data, and methods related to the sport's physical, technical, and tactical performance; and the context (C) of methods that analysed any match-play performance. A narrative synthesis was conducted using extracted study characteristic data such as sample size, population, time period, collection standards, evaluation metrics for results, and application of thematic categorisations of previous sports performance reviews. Critical appraisal of eligible studies' methodologies was conducted to investigate research quality and identify methodological issues. Results: From 183 studies screened, twelve eligible studies were included, which examined match-play through physical (9/12, 75%), technical (4/12, 33%), and tactical analysis (2/12, 17%). Running demands and game actions analysis were the most researched in senior women's AF. Research into junior girls' AF match-play performance has not been investigated. No research has been conducted on non-running physical demands, contact demands, acceleration, and tactical aspects of women's AF. All studies utilised either inferential statistics or basic predictive models. Critical appraisal deemed most studies as low risk of bias (11/12, 92%), with the remaining study having satisfactory risk. Conclusions: Future research utilising increased longitudinal and greater contextual data is needed to combat the prominent issue of data representativeness to better characterise performance within women's and girls' AF. Additionally, research involving junior and sub-elite AF players across the talent pathways is important to conduct, as it provides greater context and insight regarding development to support the evolving elite women's AF competition. Women's AF has been constrained by its resource environment. As such, suggestions are provided for better utilisation of existing data, as well as for the creation of new data for appropriate future research. Greater data generation enables the use of detailed machine learning predictions, neural networks, and network analysis to better represent the intertwined nature of match-play performance from technical, physical, and tactical data.

Improving mortality models in the ICU with high-frequency data.

BACKGROUND: Assessment of the performance of Intensive Care Units (ICU) is of vital importance for an effective healthcare system. Such assessment ensures that the limited resources of the healthcare system are allocated where they are most needed. Severity scoring systems are employed for this purpose and improving these systems is a continuing area of research which has focused on the use of more complex techniques and new variables. OBJECTIVES: This paper investigates whether scoring systems could be improved through use of metrics which better summarise the high frequency data collected by automated systems for patients in the ICU. METHODS AND DATA: 3128 admissions to the Gold Coast University Hospital ICU are used to construct three logistic regressions based on the most widely used scoring system (APACHE III) to compare performance with and without predictors leveraging available high frequency information. Performance is assessed based on model accuracy, calibration, and discrimination. High frequency information was considered for existing pulse and mean arterial pressure physiology fields and resulting models compared against a baseline logistic regression using only APACHE III physiology variables. RESULTS: Model discrimination and accuracy were better for models which included high frequency predictors, with calibration remaining good in all cases. The most influential high frequency summaries were the number of turning points in a patient's mean arterial pressure or pulse in the first 24 h of ICU admission. CONCLUSIONS: The findings indicate that scoring systems can be improved by better accounting for high frequency data.

Text Mining and Automation for Processing of Patient Referrals.

BACKGROUND: Various tasks within health care processes are repetitive and time-consuming, requiring personnel who could be better utilized elsewhere. The task of assigning clinical urgency categories to internal patient referrals is one such case of a time-consuming process, which may be amenable to automation through the application of text mining and natural language processing (NLP) techniques. OBJECTIVE: This article aims to trial and evaluate a pilot study for the first component of the task-determining reasons for referrals. METHODS: Text is extracted from scanned patient referrals before being processed to remove nonsensical symbols and identify key information. The processed data are compared against a list of conditions that represent possible reasons for referral. Similarity scores are used as a measure of overlap in terms used in the processed data and the condition list. RESULTS: This pilot study was successful, and results indicate that it would be valuable for future research to develop a more sophisticated classification model for determining reasons for referrals. Issues encountered in the pilot study and methods of addressing them were outlined and should be of use to researchers working on similar problems. CONCLUSION: This pilot study successfully demonstrated that there is potential for automating the assignment of reasons for referrals and provides a foundation for further work to build on. This study also outlined a potential application of text mining and NLP to automating a manual task in hospitals to save time of human resources.