Search

Random forests are a powerful machine learning tool that capture complex relationships between independent variables and an outcome of interest. Trees built in a random forest are dependent on several hyperparameters, one of the more critical being the node size. The original algorithm of Breiman, controls for node size by limiting the size of the parent node, so that a node cannot be split if it has less than a specified number of observations. We propose that this hyperparameter should instead be defined as the minimum number of observations in each terminal node. The two existing random forest approaches are compared in the regression context based on estimated generalization error, bias-squared, and variance of resulting predictions in a number of simulated datasets. Additionally the two approaches are applied to type 2 diabetes data obtained from the National Health and Nutrition Examination Survey. We have developed a straightforward method for incorporating weights into the random forest analysis of survey data. Our results demonstrate that generalization error under the proposed approach is competitive to that attained from the original random forest approach when data have large random error variability. The R code created from this work is available and includes an illustration.

Irukandji syndrome (IS) is an extremely painful condition that causes a significant catecholamine surge and sympathetic autonomic response related to the envenomation from certain types of jellyfish. Current management involves intravenous fluids, magnesium sulphate and large doses of opioids for symptom control. Clonidine, a centrally acting alpha-2 agonist, is often used as an analgesic adjunct to reduce opioid requirements in acute pain. The present study explores the safety and efficacy of clonidine in reducing opioid requirements in IS.All patients diagnosed with IS at Cairns Hospital between 1 March 2016 and 30 April 2020, and participants from the Magnesium in Irukandji Study Trial, were included in this retrospective study (n = 114). Cases were separated into two groups depending on whether or not they received clonidine, and subsequently analysed according to pre- and post-intervention opioid requirements, clonidine dose administered and adverse effects.Notably, 39 patients with IS received ≥1 mcg/kg clonidine and the remaining 75 did not. There was no difference in oral morphine equivalent daily dose (oMEDD) between groups before clonidine administration; however, there was a significant reduction in oMEDD required after patients received clonidine (26.1 mg; 95% CI 4.6-47.7) compared with those who did not (66.6 mg; 95% CI 56.9-86.1) (F = 8.722, df = 1 × 224, P = 0.003). One episode of hypotension occurred following the intervention.Patients with IS who received clonidine required significantly lower opioid requirements than those who did not receive clonidine. Clonidine was safe to administer and should be considered early when treating IS. The optimal clonidine dose remains unclear and requires prospective studies to validate our findings.

Venom-induced consumption coagulopathy (VICC) is one of the most dangerous syndromes caused by snake envenomation and can be caused by several snake species worldwide, including the Australian coastal taipan. Rotational thromboelastometry (ROTEM) provides real-time point-of-care information on all stages of clot formation; however, it has yet to be formally evaluated in the assessment of VICC. We report three cases of Taipan envenomation causing VICC and the associated ROTEM results. The implications for future use of ROTEM in the assessment, management, and further research of VICC are discussed.

Anti-glomerular basement membrane (GBM) disease is a rare, aggressive vasculitis with no validated prediction tools to assist its management. We investigated a retrospective multicenter international cohort with the aim to transfer the Renal Risk Score (RRS) and to identify patients that benefit from rescue immunosuppressive therapy. Of a total 191 patients, 174 patients were included in the final analysis (57% female, median age 59 years). Using Cox and Kaplan-Meier methods, the RRS was found to be a strong and effective predictor for end stage kidney disease (ESKD) with a model concordance of C=0.760. The 36-month renal survival was 100%, 62.4%, and 20.7% in the low-, moderate-, and high-risk groups, respectively (P0.001). The need for renal replacement therapy (RRT) at diagnosis and the percentage of normal glomeruli in the biopsy were independent predictors of ESKD (P0.001, P0.001). Considering the 129 patients initially requiring RRT, the best predictor for renal recovery was the percentage of normal glomeruli (C=0.622; P0.001), a split either side of 10% providing good stratification. A model with the predictors RRT and normal glomeruli (N) achieved superior discrimination (C=0.840, P0.001). Dividing patients into four risk groups led to a 36-month renal survival of 96.4% (no RRT, N≥10%), 74.0% (no RRT, N10%), 42.3% (RRT, N≥10%) and 14.1% (RRT, N10%), respectively. In summary, we demonstrate that the RRS concept is transferrable to anti-GBM disease. Stratifying patients according to the need for RRT at diagnosis and renal histology improves prediction, highlighting the importance of normal glomeruli. Here, we propose a stratification to assist in the management of anti-GBM disease.

BACKGROUND AND AIMS Anti-neutrophil cytoplasmic antibody-associated vasculitis (AAV) is a debilitating disease that can have a significant impact on a patient's quality of life. The aim of this study was to assess the longitudinal quality of life amongst those diagnosed with AAV using the EQ-5D instrument, which allows for calculation of quality-adjusted life years (QALYs.) METHOD A total of 343 patients with AAV participated in this study, of which 191 (55.7%) were male, resulting in 2746 episodes. The EQ-5D-5L standardised instrument was used to evaluate health-related quality of life in the domains of mobility, self-care, usual activities, pain/discomfort, anxiety/depression and to generate a summary index score. Overall health was also rated using a visual analogue scale (0–100). EQ-5D questionnaires were completed during routine nephrology clinic attendances and through a vasculitis patient support smartphone app. We used a random effects model to control for multiple entries relating to individual patients. RESULTS A lower quality of life was seen amongst those with AAV (median index value 0.80, overall population average 0.856). The mean visual analogue scale score was 75.6 ± 17.3 (overall population average 82.8, Fig. 1). Patients’ pain and discomfort level (mean 1.95) was most affected while self-care (mean 1.33) was least affected (Fig. 1). An increase in BVAS tightly correlated with a reduction in quality of life. Using the random effects model, the index score was seen to decrease with increasing age with a 2.7% reduction in index score per decade. A 7% reduction in index score was seen during periods of disease activity compared with periods of remission. Patients with end-stage kidney disease requiring dialysis had an 8% reduction in index score. A reduced quality of life was seen following COVID-19 lockdown with a 5% reduction in index score seen. Using a median survival rate of 6.16 years for patients with small vessel vasculitis, we calculated the QALYs for this population as 4.9 years. CONCLUSION We have defined for the first time the EQ-5D index value over the full disease course in patients with AAV. Notably, we have identified a reduction in quality of life during periods of disease activity. Other studies have demonstrated a reduction in quality of life during active disease using the AAV–PRO questionnaire and the Medical Outcomes Study Short Form-36. A decrease in work productivity has also been noted. Previously reported mean index values of 0.72 and 0.76 were lower than our observed values, although both are significantly reduced compared with population norms. In conclusion, this research highlights the negative impact of AAV on patients’ lives.

Presentation We present the case of a 48-year-old man with nasal cellulitis and subsequent oro-naso-sino-orbital-cutaneous fistula from prolonged cocaine use. Diagnosis Initial laboratory investigations reported a raised white cell count (WBC) and C-Reactive Protein (CRP) and subsequently a positive atypical anti-neutrophil cytoplasm antibodies (ANCA) and positive anti-proteinase (PR3). Perihilar lung nodularity on chest imaging raised the possibility of a systemic autoimmune response. His urinalysis was positive for cocaine. Treatment He was commenced on Augmentin, Amphotericin B and Prednisolone. An obturator was created to manage the oro-nasal fistula. A subsequent naso-cutaneous defect was re-approximated. Daily nasal saline douche and abstinence of cocaine were recommended. Discussion Cocaine use in the community is rising and poses a challenge to multiple facets of our health care system.

s from the 20th International Vasculitis and ANCA conference, held in Dublin on 3-6th April, Minor edits

Background: The current coronavirus disease 2019 (COVID-19) pandemic began in Ireland with the first confirmed positive case in March 2020. In the early stages of the pandemic clinicians and researchers in two affiliated Dublin hospitals identified the need for a COVID-19 biobanking initiative to support and enhance research into the disease. Through large scale analysis of clinical, regional, and genetic characteristics of COVID-19 patients, biobanks have helped identify, and so protect, at risk patient groups The STTAR Bioresource has been created to collect and store data and linked biological samples from patients with SARS-CoV-2 infection and healthy and disease controls. Aim: The primary objective of this study is to build a biobank, to understand the clinical characteristics and natural history of COVID-19 infection with the long-term goal of research into improved disease understanding, diagnostic tests and treatments. Methods: This is a prospective dual-site cohort study across two tertiary acute university teaching hospitals. Patients are recruited from inpatient wards or outpatient clinics. Patients with confirmed COVID-19 infection as well as healthy and specific disease control groups are recruited. Biological samples are collected and a case report form detailing demographic and medical background is entered into the bespoke secure online Dendrite database. Impact: The results of this study will be used to inform national and international strategy on health service provision and disease management related to COVID-19. In common with other biobanks, study end points evolve over time as new research questions emerge. They currently include patient survival, occurrence of severe complications of the disease or its therapy, occurrence of persistent symptoms following recovery from the acute illness and vaccine responses.

Rare disease data is often fragmented within multiple heterogeneous siloed regional disease registries, each containing a small number of cases. These data are particularly sensitive, as low subject counts make the identification of patients more likely, meaning registries are not inclined to share subject level data outside their registries. At the same time access to multiple rare disease datasets is important as it will lead to new research opportunities and analysis over larger cohorts. To enable this, two major challenges must therefore be overcome. The first is to integrate data at a semantic level, so that it is possible to query over registries and return results which are comparable. The second is to enable queries which do not take subject level data from the registries. To meet the first challenge, this paper presents the FAIRVASC ontology to manage data related to the rare disease anti-neutrophil cytoplasmic antibody (ANCA) associated vasculitis (AAV), which is based on the harmonisation of terms in seven European data registries. It has been built upon a set of key clinical questions developed by a team of experts in vasculitis selected from the registry sites and makes use of several standard classifications, such as Systematized Nomenclature of Medicine - Clinical Terms (SNOMED-CT) and Orphacode. It also presents the method for adding semantic meaning to AAV data across the registries using the declarative Relational to Resource Description Framework Mapping Language (R2RML). To meet the second challenge a federated querying approach is presented for accessing aggregated and pseudonymized data, and which supports analysis of AAV data in a manner which protects patient privacy. For additional security the federated querying approach is augmented with a method for auditing queries (and the uplift process) using the provenance ontology (PROV-O) to track when queries and changes occur and by whom. The main contribution of this work is the successful application of semantic web technologies and federated queries to provide a novel infrastructure that can readily incorporate additional registries, thus providing access to harmonised data relating to unprecedented numbers of patients with rare disease, while also meeting data privacy and security concerns.