Annie Conway

Background: Frailty is prevalent among hospitalized patients and is associated with increased risk of complications, prolonged hospitalization, higher mortality, and health care costs. Existing frailty screening tools, including the Clinical Frailty Scale (CFS) and Hospital Frailty Risk Score (HFRS), have limitations related to subjectivity, resource intensity, or reliance on retrospective data, limiting their use in routine acute care. Early identification of frailty is critical to support targeted interventions and resource allocation, yet scalable approaches for predicting frailty early during hospital admission remain limited. Objective: This paper evaluates whether routinely collected electronic medical record (EMR) data available within the first 72 hours of admission can be used to predict frailty in hospitalized patients. Methods: A retrospective cohort study was conducted using routinely collected EMR data from the Royal Adelaide Hospital (RAH), including overnight admissions with a length of stay of at least 72 hours between July 2023 and December 2024. Frailty was defined using the HFRS (score >4). Predictors were derived from EMR data available within 72 hours of admission, including demographics, medications, laboratory results, nursing assessments, and medical and allied health consult orders. Data were split into training (21,436/26,824, 80%) and test (5388/26,824, 20%) sets. Multiple machine learning models were trained, with the F ₂ -score as the primary optimization metric. Model performance was evaluated using the F ₂ -score, recall, area under the receiver operating characteristic curve (AUROC), calibration, and clinical utility analyses. Results: This study included 26,824 admissions, of whom 9427 (35%) were classified as frail. The final XGBoost (XGB) model achieved an F ₂ -score of 0.755 (95% CI 0.741-0.768), an AUROC of 0.777 (95% CI 0.764-0.790), and a recall of 0.873 (95% CI 0.857-0.887) on the test dataset, with good calibration and an expected-to-observed ratio close to 1 (0.996, 95% CI 0.963-1.028). Key predictors included polypharmacy, laboratory results (including creatinine, albumin, and sodium), allied health consult orders, and nursing assessment data. Model performance was lower at 24 hours postadmission and improved by discharge. Conclusions: An EMR-based frailty prediction model using data available at 72 hours postadmission can identify patients with frailty early in their hospital stay, supporting targeted interventions and operational planning. Reduced performance in Indigenous patients highlights the need for culturally validated frailty indices and ongoing evaluation of equity in predictive modeling.

Background: The proportion of dialysis patients performing PD in Australia is decreasing over time. ¹ Understanding the trends in population, the rates and causes of peritonitis and haemodialysis (HD) transfer over time is critical towards understanding this trend. Methods: Study Design: Registry-based cohort study using Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry data. Inclusion criteria: All Australian patients starting PD older than 18 years, starting PD within 90 days of KRT entry from 1 January 2013 through 31 December 2022. Analysis: Peritonitis rates were analysed using Poisson regression, while time to first peritonitis and HD transfer were assessed using Cox Proportional Hazards and competing-risks regression models. Results: The demographic profile of PD patients has remained relatively stable. Peritonitis rates declined from 0.40 to 0.33 infections per patient-year between 2013 and 2022. Over time, peritonitis-associated HD transfer reduced, HR 0.96 [95% CI: 0.93–0.98], p  < . 001. Despite this, there was a deterioration in overall HD transfer, HR 1.03 [95% CI: 1.02–1.04] per year, p  < . 001 and worse adjusted non-infective HD transfer HR 1.02 [95% CI: 1.01–1.04] per year, p = .006. Conclusion: PD utilisation in Australia has declined over the past decade, despite improvements in peritonitis rates and peritonitis-associated HD transfer. The demographic profile of PD patients has remained relatively stable, suggesting systemic rather than patient-level factors are driving these trends. The apparent disconnect between infection control improvements and overall HD transfer highlights the growing importance of non-infective contributors such as access, psychosocial factors, and clinical support.

Introduction: Therapy alerts during automated peritoneal dialysis (APD) can cause significant disruptions to patients’ sleep and quality of life and may portend poorer outcomes. Understanding the relationship between alert frequency during this early period and longer-term PD outcomes is important. Methods: Following the probabilistic linkage of Vantive's Sharesource database to the Australian and New Zealand Dialysis and Transplant (ANZDATA) Registry, we examined the relationship between alert frequency in the first 30 days of APD and PD discontinuation. We included adult patients in Australia and New Zealand who commenced APD with the Vantive Homechoice Claria cycler over 2019–2023 and continued for at least 30 days. The average alerts per treatment in the first 30 days were divided into quartiles and time to PD discontinuation (inclusive of HD transfer and death), HD transfer only, and infective and non-infective HD transfer were modelled as outcomes. Results: The cohort was 1880 patients, 65% male, and median age at PD commencement of 58 years. Overall PD continuation at 1,2, and 3 years was 78%, 56% and 41%, with HD transfer rates at 14%, 23% and 27%. Higher rates of HD transfer in the first 12 months were seen in the groups with a higher average alert number. Within 12 months, there was a progressive risk of non-infective HD transfer with increasing 30-day alert quartile. Conclusion: Alert burden in the first 30 days is a risk factor for HD transfer in the first 12 months, and resolving underlying issues early may help to improve PD continuation.