The app, called ResAppDx, has been produced by the company ResApp Health Limited, which focuses on the diagnosis of respiratory disease and medical management solutions to treat such diseases. The company has recently issued an update about a major Australian pediatric clinical study. The study cohort consists of 1,127 children.
The app is made-up of a medical dataset together with machine learning algorithms. The algorithms have been optimized to match the design previous ResApp studies. The algorithms have been fine-tuned to differentially diagnose disease from the child patients recruited to the study.
The study findings have been assessed by Professor Udantha Abeyratne from The University of Queensland. The analysis demonstrates that the ResAppDx app and conventional clinical diagnosis methods correlate. The standard clinical methodologies include clinical presentation, auscultation, imaging and laboratory tests). Based on U.S. Food and Drug Administration (FDA) guidance, the assessment of the health app is made using positive, negative and overall percent agreement instead of the classic analytical method criteria of sensitivity, specificity and accuracy.
Fact check: What is positive and negative agreement?
Positive percent agreement (the substitute for sensitivity) is the proportion of patients with the disease that test positive. Negative percent agreement (specificity) is the proportion of patients without the disease that test negative. Overall percent agreement (accuracy) is a measure of both categories.
Results to date show ResAppDx achieves 90 percent to 100 percent positive agreement and between 89 percent and 96 percent negative percent agreeemnt, in relation to the clinical diagnosis of primary upper respiratory tract infection, croup, lower respiratory tract involvement, asthma (reactive airways disease RAD) and bronchiolitis. Similarly effective levels have been recorded for pneumonia.
Commenting on this, Dr Tony Keating, who is the Managing Director of ResApp Health, told Digital Journal: “This is the first study where ResAppDx has been used to diagnose a full range of respiratory diseases among a group of patients with a variety of ailments, which is a much greater challenge than previously reported. The excellent results achieved reaffirm the performance of ResAppDx and give us a great deal of confidence.” To further evaluate the app, further trials will take place.
