RESUMEN
Background: Australia's response to the coronavirus disease 2019 (COVID-19) pandemic relies on widespread availability of rapid, accurate testing and reporting of results to facilitate contact tracing. The extensive geographical area of Australia presents a logistical challenge, with many of the population located distant from a laboratory capable of robust severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) detection. A strategy to address this is the deployment of a mobile facility utilizing novel diagnostic platforms. This study aimed to evaluate the feasibility of a fully contained transportable SARS-CoV-2 testing laboratory using a range of rapid point-of-care tests. Method: A 20 ft (6.1 m) shipping container was refurbished (GeneWorks, Adelaide, South Australia) with climate controls, laboratory benches, hand-wash station and a class II biosafety cabinet. Portable marquees situated adjacent to the container served as stations for registration, sample acquisition and personal protective equipment for staff. Specimens were collected and tested on-site utilizing either the Abbott ID NOW or Abbott Panbio rapid tests. SARS-CoV-2 positive results from the rapid platforms or any participants reporting symptoms consistent with COVID-19 were tested on-site by GeneXpert Xpress RT-PCR. All samples were tested in parallel with a standard-of-care RT-PCR test (Panther Fusion SARS-CoV-2 assay) performed at the public health reference laboratory. In-laboratory environmental conditions and data management-related factors were also recorded. Results: Over a 3 week period, 415 participants were recruited for point-of-care SARS-CoV-2 testing. From time of enrolment, the median result turnaround time was 26 min for the Abbott ID NOW, 32 min for the Abbott Panbio and 75 min for the Xpert Xpress. The environmental conditions of the refurbished shipping container were found to be suitable for all platforms tested, although humidity may have produced condensation within the container. Available software enabled turnaround times to be recorded, although technical malfunction resulted in incomplete data capture. Conclusion: Transportable container laboratories can enable rapid COVID-19 results at the point of care and may be useful during outbreak settings, particularly in environments that are physically distant from centralized laboratories. They may also be appropriate in resource-limited settings. The results of this pilot study confirm feasibility, although larger trials to validate individual rapid point-of-care testing platforms in this environment are required.
RESUMEN
INTRODUCTION: To self-monitor asthma symptoms, existing methods (e.g. peak flow metre, smart spirometer) require special equipment and are not always used by the patients. Voice recording has the potential to generate surrogate measures of lung function and this study aims to apply machine learning approaches to predict lung function and severity of abnormal lung function from recorded voice for asthma patients. METHODS: A threshold-based mechanism was designed to separate speech and breathing from 323 recordings. Features extracted from these were combined with biological factors to predict lung function. Three predictive models were developed using Random Forest (RF), Support Vector Machine (SVM), and linear regression algorithms: (a) regression models to predict lung function, (b) multi-class classification models to predict severity of lung function abnormality, and (c) binary classification models to predict lung function abnormality. Training and test samples were separated (70%:30%, using balanced portioning), features were normalised, 10-fold cross-validation was used and model performances were evaluated on the test samples. RESULTS: The RF-based regression model performed better with the lowest root mean square error of 10·86. To predict severity of lung function impairment, the SVM-based model performed best in multi-class classification (accuracy = 73.20%), whereas the RF-based model performed best in binary classification models for predicting abnormal lung function (accuracy = 85%). CONCLUSION: Our machine learning approaches can predict lung function, from recorded voice files, better than published approaches. This technique could be used to develop future telehealth solutions including smartphone-based applications which have potential to aid decision making and self-monitoring in asthma.
RESUMEN
BACKGROUND: In Australia, COVID-19 diagnosis relies on RT-PCR testing which is relatively costly and time-consuming. To date, few studies have assessed the performance and implementation of rapid antigen-based SARS-CoV-2 testing in a setting with a low prevalence of COVID-19 infections, such as Australia. METHODS: This study recruited participants presenting for COVID-19 testing at three Melbourne metropolitan hospitals during a period of low COVID-19 prevalence. The Abbott PanBioTM COVID-19 Ag point-of-care test was performed alongside RT-PCR. In addition, participants with COVID-19 notified to the Victorian Government were invited to provide additional swabs to aid validation. Implementation challenges were also documented. FINDINGS: The specificity of the Abbott PanBioTM COVID-19 Ag test was 99.96% (95% CI 99.73 - 100%). Sensitivity amongst participants with RT-PCR-confirmed infection was dependent upon the duration of symptoms reported, ranging from 77.3% (duration 1 to 33 days) to 100% in those within seven days of symptom onset. A range of implementation challenges were identified which may inform future COVID-19 testing strategies in a low prevalence setting. INTERPRETATION: Given the high specificity, antigen-based tests may be most useful in rapidly triaging public health and hospital resources while expediting confirmatory RT-PCR testing. Considering the limitations in test sensitivity and the potential for rapid transmission in susceptible populations, particularly in hospital settings, careful consideration is required for implementation of antigen testing in a low prevalence setting. FUNDING: This work was funded by the Victorian Department of Health and Human Services. The funder was not involved in data analysis or manuscript preparation.