Gastrostomy tube (GT) complications are often managed in the Emergency Department (ED). We aimed to characterize and compare the pattern of ED presentations of GT complications in adults and children. A retrospective chart review of patients with GT complications presenting to 3 Australian EDs in 2 years was undertaken. ED visits for GT complications occurred in 70 GT patients (36 adults, 34 children) with 122 presentations. When comparing adults to children, infections occurred in 21% versus 36%, respectively; P = 0.08, mechanical issues in 48% versus 52%; P = 0.86, vomiting in 23% versus 8%; P = 0.02, and other issues in 7% versus 5%; P = 0.7. Presentation to ED within 28 days of initial GT insertion occurred in 3 (8%) adults and 3 (9%) children, predominantly with tube dislodgement. GT complications seen in ED are predominantly infectious and mechanical in nature, with an increased frequency of vomiting in adults when compared with children.

Background: The Victorian Audit of Surgical Mortality (VASM) investigates all surgically related deaths in Victoria, Australia, as a surgical educational activity aimed to make surgery safer. Whilst data collected within the audit are regularly reviewed for accuracy, there has never been a review of the data provided from health services. Methods: Two-year death data provided by one Victorian health service were reviewed. Hospital notes for 4 months of each year were analysed to assess patients dying under surgical care. These data were compared to referrals to the VASM over the same period. Results: Of the 3907 patient deaths recorded, 35.1% were reviewed. During their final admission, 178 (13%) patients underwent a procedure (93 medical and 85 surgical). Only 29.2% of these were recorded in the health service data set. Eighteen patients died under the care of a surgeon without a procedure, meaning that 103 deaths should have been reported to the VASM of which only 55.3% (57/103) were reported. Conclusion: There were major errors in the health service database resulting in underreporting of deaths to the VASM which could have education and policy repercussions. For improvements to the safety and quality of health services, it is critical that all deaths are accurately recorded by health services and reported to the relevant bodies with internal verification processes.© 2020 Royal Australasian College of Surgeons ANZ J Surg 90 (2020) 725-727ANZJSurg.com

IntroductionEpileptic and psychogenic non-epileptic seizures (PNES) are common diagnostic problems encountered in hospital practice. This study explores the use of unsupervised machine learning in discovering themes in medical records of patients presenting with PNES. We hypothesised that themes generated by machine learning are comparable with the classification by human experts.MethodsThis is a retrospective analysis of the medical records in the emergency department of patients (age >18 years) with PNES who underwent inpatient video-electroencephalography monitoring from May 2009 to June 2014 and received a final diagnosis of PNES. Prior to machine learning of written text, we applied a standardised approach in natural language processing to create a document-term matrix (removal of numbers, stop-words and punctuations, transforming fonts to lower case). The words were separated into tokens and treated as if existing within a bag-of-words. A probability of each word existing within a topic (theme) was modelled on multivariate Dirichlet distribution (R Foundation, V.3.5.0). Next, we asked four experts to independently provide a clinical interpretation of the generated topics. When the majority of (≥3) experts agreed, it was regarded as highly congruent. Interactive data are available on the web at (https://gntem2.github.io/PNES/%23topic=1&lambda=0.6&term=).ResultsThere were 39 patients (74.4% women, median age 35 years with range 20–82). A total of 121 documents were converted to text files for text mining. There were 15 generated topics with 12/15 topics rated as highly congruent. The main themes were about descriptors of seizures and medication use.ConclusionsThe findings from machine learning on PNES-related documentation provides evidence for the feasibility of applying machine-learning methodology to analyse large volumes of medical records. The topics generated by machine learning were congruent with interpretations by clinicians indicating this method can be used for screening of medical conditions among large volumes of medical records.