Using machine learning models to predict falls in hospitalised adults.

BACKGROUND: Identifying patients at high risk of falling is crucial in implementing effective fall prevention programs. While the integration of information systems is becoming more widespread in the healthcare industry, it poses a significant challenge in analysing vast amounts of data to identify factors that could enhance patient safety. OBJECTIVE: To determine fall-associated factors and develop high-performance prediction tools for at-risk patients in acute and sub-acute care services in Australia. METHODS: A retrospective study of 672,400 patients admitted to acute and sub-acute care services within a large metropolitan tertiary health service in Victoria, Australia, between January 1, 2019, and December 31, 2021. Data were obtained from four sources: the Department of Health Victorian Admitted Episodes Dataset, RiskManTM, electronic health records, and the health workforce dataset. Machine learning techniques, including Random Forest and Deep Neural Network models, were used to analyse the data, predict patient falls, and identify the most important risk factors for falls in this population. Model performance was evaluated using accuracy, F1-score, precision, recall, specificity, Matthew's correlation coefficient, and the area under the receiver operating characteristic curve (AUC). RESULTS: The deep neural network and random forest models were highly accurate in predicting hospital patient falls. The deep neural network model achieved an accuracy of 0.988 and a specificity of 0.999, while the RF achieved an accuracy of 0.989 and a specificity of 1.000. The top 20 variables impacting falls were compared across both models, and 12 common factors were identified. These factors can be broadly classified into three categories: patient-related factors, staffing-related factors, and admission-related factors. Although not all factors are modifiable, they must be considered when planning fall prevention interventions. CONCLUSION: The study demonstrated machine learning's potential to predict falls and identify key risk factors. Further validation across diverse populations and settings is essential for broader applicability.

Family experiences and perceptions of intensive care unit care and communication during the COVID-19 pandemic.

INTRODUCTION: In 2020, during the first wave of the COVID-19 pandemic in Melbourne, visitor access to acute hospitals including intensive care units (ICUs) was initially barred, followed by a limit of one person per patient for one hour per day. This study explores the care and communication experienced by family members of ICU patients during this time. METHODS: This qualitative descriptive study was conducted at an Australian quaternary hospital. Semistructured phone interviews were conducted using an aide-memoire designed to understand participants' experiences as family of a patient during this time. Interviews were recorded, transcribed, and thematically analysed. FINDINGS: Twenty family members of patients in the ICU participated. Three major themes were identified: 'impact of restricting visiting procedures', 'family experiences of communication', and 'care and support'. Inflexible visiting restrictions had a momentous impact on families. Participants objected to having to nominate only two people to visit during the admission and the short visiting time limit. Some family members suffered extreme stress and anxiety during their absence from the bedside. Additional challenges were experienced by rural families, visitors with disabilities, and the young children of patients who were excluded. Communication with clinicians varied. Telehealth was valued by some but not universally embraced. The relationship between staff members and families and involvement in decision-making were unaffected. CONCLUSION: Families experienced significant psychological distress from being separated from their critically ill relatives. Patient care and involvement in decision-making appeared to be unchanged, but communication with staff felt to be lacking. Better alternatives to face-to-face communication must be sought to limit the impact of family separation on mental health. Families are a key link between the patient and clinicians and often play a major role in patient support and recovery after discharge. There is an urgent need to support them and facilitate meaningful engagement despite the obstacles.

Predicting medical emergency team calls, cardiac arrest calls and re-admission after intensive care discharge: creation of a tool to identify at-risk patients.

We aimed to develop a predictive model for intensive care unit (ICU)-discharged patients at risk of post-ICU deterioration. We performed a retrospective, single-centre cohort observational study by linking the hospital admission, patient pathology, ICU, and medical emergency team (MET) databases. All patients discharged from the Alfred Hospital ICU to wards between July 2012 and June 2014 were included. The primary outcome was a composite endpoint of any MET call, cardiac arrest call or ICU re-admission. Multivariable logistic regression analysis was used to identify predictors of outcome and develop a risk-stratification model. Four thousand, six hundred and thirty-two patients were included in the study. Of these, 878 (19%) patients had a MET call, 51 (1.1%) patients had cardiac arrest calls, 304 (6.5%) were re-admitted to ICU during the same hospital stay, and 964 (21%) had MET calls, cardiac arrest calls or ICU re-admission. A discriminatory predictive model was developed (area under the receiver operating characteristic curve 0.72 [95% confidence intervals {CI} 0.70 to 0.73]) which identified the following factors: increasing age (odds ratio [OR] 1.012 [95% CI 1.007 to 1.017] P <0.001), ICU admission with subarachnoid haemorrhage (OR 2.26 [95% CI 1.22 to 4.16] P=0.009), admission to ICU from a ward (OR 1.67 [95% CI 1.31 to 2.13] P <0.001), Acute Physiology and Chronic Health Evaluation (APACHE) III score without the age component (OR 1.005 [95% CI 1.001 to 1.010] P=0.025), tracheostomy on ICU discharge (OR 4.32 [95% CI 2.9 to 6.42] P <0.001) and discharge to cardiothoracic (OR 2.43 [95%CI 1.49 to 3.96] P <0.001) or oncology wards (OR 2.27 [95% CI 1.05 to 4.89] P=0.036). Over the two-year period, 361 patients were identified as having a greater than 50% chance of having post-ICU deterioration. Factors are identifiable to predict patients at risk of post-ICU deterioration. This knowledge could be used to guide patient follow-up after ICU discharge, optimise healthcare resources, and improve patient outcomes and service delivery.

Influence of task properties and subjectivity on consistency of triage: a simulation study.

AIM: This paper reports a study to determine nurses' levels of agreement using a standard 5-point triage scale and to explore the influence of task properties and subjectivity on decision-making consistency. BACKGROUND: Triage scales are used to define time-to-treatment in hospital emergency departments. Studies of the inter-rater reliability of these scales using paper-based simulation methods report varying levels of consistency. Understanding how various components of the decision task and individual perceptions of the case influence agreement is critical to the development of strategies to improve consistency of triage. METHOD: Simulations were constructed from naturalistic observation, cue types and frequencies were classified. Data collection was conducted in 2002, and the final response rate was 41 x 3%. Participants were asked to allocate an urgency code for 12 scenarios using the Australasian Triage Scale, and provide estimates of case complexity, levels of certainty and available information. Data were analysed descriptively, agreement between raters was calculated using kappa. The influence of task properties and participants' subjective estimates of case complexity, levels of certainty and available information on agreement were explored using a general linear model. FINDINGS: Agreement among raters varied from moderate to poor (kappa=0 x 18-0 x 64). Participants' subjective estimates of levels of available information were found to influence consistency of triage by statistically significant amounts (F 5 x 68;

Triage nurses' clinical decision making. An observational study of urgency assessment.

BACKGROUND: Researchers have described both the various decision tasks performed by triage nurses using self-report methods and identified time as a factor influencing the quality of triage decisions. However, little is known about the decision tasks performed by triage nurses when making acuity assessments, or the factors influencing triage duration in the real world. AIMS: The aims of this study were to: describe the data triage nurses collect from patients in order to allocate a triage priority using the Australasian Triage Scale (ATS); describe the duration of nurses' decision making for ATS categories 2-5; and to explore the impact of patient and nurse variables on the duration of the triage nurses' decision making in the clinical setting. DESIGN: A structured observational study was employed to address the research aims. Observational data was collected in one adult emergency department located in metropolitan Melbourne, Australia. A total of 26 triage nurses consented and were observed performing 404 occasions of triage. Data was collected by a single observer using a 20-item instrument that recorded the performance frequencies of a range of decision tasks and a number of observable patient, nurse and environmental variables. Additionally, the nurse-patient interaction was recorded as time in minutes. RESULTS: It was found that there was limited use of objective physiological data collected by the nurses' in order to decide patient acuity, and large variability in the duration of triage decisions observed. In addition, analysis of variance indicated strong evidence of a true difference between triage duration and a range of nurse, patient and environmental variables. CONCLUSION: These findings have implications for the development of practice standards and triage education. In particular, it is argued that practice standards should include routine measurement of physiological parameters in all but the collapsed or obviously unwell patient, where further delay may impede the delivery of time-critical intervention. Furthermore, the inclusion of arbitrary time frames for triage assessment in practice standards are not an appropriate method of evaluating triage decision making in the real world.

Critical care nurses' decision-making activities in the natural clinical setting.

This article reports on observation of 18 nurses in urban and rural based critical care settings. The purpose of the study was to observe and describe the decision-making activities of critical care nurses within natural clinical settings. During the 2-hour observation, the researcher dictated a detailed commentary on to audio-tape of each nurse's actions. Tapes were transcribed and subjected to content analysis. Findings indicated three main categories of decisions. Decision frequencies were linked to nurses' critical care experience, appointment level, and location, as well as nursing shifts. The findings are discussed in relation to previous empirical evidence and the implications for practice. The author concludes that future research should be directed towards measuring the contextual influences on nurses' decision-making on the outcome of patient care.

Why we do the things we do: applying clinical decision-making frameworks to triage practice.

This paper discusses major themes presented in the published literature concerning clinical decision making and links these to the practice of emergency department nurse triage. Themes discussed include: approaches to decision research in nursing and medicine; decision autonomy in nursing practice and clinical decision making under conditions of uncertainty. Some assumptions underpinning clinical decision-making frameworks are explored and the use of triage scales, algorithms and intuitive thought processes are discussed in terms of clinical practice. In addition, the strengths and limitations of each approach are outlined. It is argued that naturalistic research methods are necessary in order to describe the often uncertain and frequently chaotic environment in which triage decisions are made. This research must occur in order to evaluate and improve both the triage process and the outcomes of these decisions in practice.

From abstract to acclaim.

Increasing specialisation in the health-care sector requires nurses to be proficient communicators both within their own profession and among other health-care disciplines. Although the work environment provides numerous opportunities for oral communication, traditionally we have been taught written communication. Like scientific writing, an oral presentation demands a sense of planning, audience awareness and attention to detail. Yet, because of the different medium, other important factors must also be considered. While a reader may select passages or reread a report, the audience must listen in a linear way. Your voice and body, aided by visual displays, will have a different effect on the listener than the printed pages have on the reader. Therefore, it is important to concentrate on both the content and the method of presentation if the audience is to be left feeling inspired. This article outlines the planning of a presentation, development of visual aids and delivery of a polished performance.