Search

Abstract Background Lung auscultation is fundamental to the clinical diagnosis of respiratory disease. However, auscultation is a subjective practice and interpretations vary widely between users. The digitization of auscultation acquisition and interpretation is a particularly promising strategy for diagnosing and monitoring infectious diseases such as Coronavirus-19 disease (COVID-19) where automated analyses could help decentralise care and better inform decision-making in telemedicine. This protocol describes the standardised collection of lung auscultations in COVID-19 triage sites and a deep learning approach to diagnostic and prognostic modelling for future incorporation into an intelligent autonomous stethoscope benchmarked against human expert interpretation. Methods A total of 1000 consecutive, patients aged ≥ 16 years and meeting COVID-19 testing criteria will be recruited at screening sites and amongst inpatients of the internal medicine department at the Geneva University Hospitals, starting from October 2020. COVID-19 is diagnosed by RT-PCR on a nasopharyngeal swab and COVID-positive patients are followed up until outcome (i.e., discharge, hospitalisation, intubation and/or death). At inclusion, demographic and clinical data are collected, such as age, sex, medical history, and signs and symptoms of the current episode. Additionally, lung auscultation will be recorded with a digital stethoscope at 6 thoracic sites in each patient. A deep learning algorithm (DeepBreath) using a Convolutional Neural Network (CNN) and Support Vector Machine classifier will be trained on these audio recordings to derive an automated prediction of diagnostic (COVID positive vs negative) and risk stratification categories (mild to severe). The performance of this model will be compared to a human prediction baseline on a random subset of lung sounds, where blinded physicians are asked to classify the audios into the same categories. Discussion This approach has broad potential to standardise the evaluation of lung auscultation in COVID-19 at various levels of healthcare, especially in the context of decentralised triage and monitoring. Trial registration: PB_2016-00500, SwissEthics. Registered on 6 April 2020.

Abstract Background Despite vaccination programs, Streptococcus pneumoniae remains among the main microorganisms involved in bacterial pneumonia, notably in terms of severity. The prognosis of pneumococcal infections is conditioned in part by the precocity of the diagnosis. The aim of this study was to evaluate the impact of a Rapid Diagnostic Test (RDT) targeting cell wall polysaccharide of Streptococcus pneumoniae and performed directly in respiratory samples, on the strategy of diagnosis of respiratory pneumococcal infections in children. Results Upper-respiratory tract samples from 196 children consulting at hospital for respiratory infection were tested for detecting S. pneumoniae using a newly-designed RDT (PneumoResp, Biospeedia), a semi-quantitative culture and two PCR assays. If positive on fluidized undiluted specimen, the RDT was repeated on 1:100-diluted sample. The RDT was found highly specific when tested on non-S. pneumoniae strains. By comparison to culture and PCR assays, the RDT on undiluted secretions exhibited a sensitivity (Se) and negative predictive value (NPV) of more than 98%. By comparison to criteria of S. pneumoniae pneumonia combining typical symptoms, X-ray image, and culture ≥107 CFU/ml, the Se and NPV of RDT on diluted specimens were 100% in both cases. Conclusions In case of negative result, the excellent NPV of RDT on undiluted secretions allows excluding S. pneumoniae pneumonia. In case of positive result, the excellent sensitivity of RDT on diluted secretions for the diagnosis of S. pneumoniae pneumonia allows proposing a suitable antimicrobial treatment at day 0.