A smartphone-based algorithm has been developed that can diagnose community-acquired pneumonia without the need for chest x-rays, research reveals today.
The Australian study, conducted by Joondalup Health Campus, Curtin University, The University of Queensland and ResApp Health, shows the app can assess the sounds generated during five cough events plus four simple patient-reported symptoms with more than 85% accuracy.
Between January 2016 and March 2019, the study recruited two cohorts from a large, general hospital in Western Australia. Data from the first cohort were used to develop and optimise the software algorithm, while the second cohort was used to test the optimised algorithm.
Cough recordings were collected using iPhone 6 phones that were held 25-50cm from the mouth and tests among the patients, who were all aged 12 and over, showed that it was also able to accurately assess mild cases of CAP, which tend to be more difficult to diagnose, and among those over 65.
The app takes less than one minute to provide a result, is easy to use and can be performed in a non-medical setting if required.
Because it does not require a clinical examination or radiology, the researchers say the app offers a new avenue for diagnosing pneumonia where in-person attendance at a medical facility is not possible.
Lead author Paul Porter said: “We have developed a method that agrees with expert diagnosis of community acquired pneumonia while not requiring auscultation or radiology.
“Diagnosing respiratory diseases can be challenging, particularly in circumstances where an examination is not possible such as in telehealth.
“This technology solves a major obstacle in the provision of respiratory healthcare during telehealth and digital consultations.”
British Journal of General Practice 9 February 2021
Leave a Reply