Fact checked byKristen Dowd

Read more

May 20, 2024
1 min read
Save

AI-based software for spirometry interpretation improves primary care diagnostic accuracy

Fact checked byKristen Dowd
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Key takeaways:

  • Many general practitioners need help interpreting spirometry.
  • An AI-based software had 88% diagnostic accuracy with diagnoses made by a panel of pulmonologists.

SAN DIEGO — Use of an AI-supported software to interpret spirometry curves resulted in more accurate diagnoses made by general practitioners, according to research presented at the American Thoracic Society International Conference.

Elena Smets

“General practitioners often don’t feel comfortable to execute and interpret spirometry due to a lack of training and expertise,” Elena Smets, PhD, product manager at ArtiQ, told Healio. “The ArtiQ.Spiro AI solution investigated in this study is available to be used in everyday practice and can help to remove these barriers, ultimately improving patient outcomes.”

Infographic showing percentage of patients diagnosed correctly by general practitioners.
Data were derived from Maes J, et al. Clinical evaluation of artificial intelligence supported software in the interpretation of spirometry results in a primary care setting. Presented at: American Thoracic Society International Conference; May 17-22, 2024; San Diego.

In this study, Julie Maes, MSc, product manager at ArtiQ, and researchers from the University of Leuven evaluated the use of ArtiQ.Spiro, an AI-supported software, for 3 months in six primary care practices in Belgium to determine the software’s impact on the accuracy of general practitioner spirometry assessments.

To find the value of ArtiQ.Spiro use in this setting, researchers compared the diagnoses general practitioners made using clinical evaluation only and spirometry curves with support from the software.

“ArtiQ.Spiro provides support for execution and interpretation of spirometry,” Smets said.

Further, researchers found the percentage of accurate diagnoses made by the software by looking at them against the diagnoses determined by three pulmonologists who had the same data as the software, as well as relevant patient record information following spirometry.

Diagnostic agreement between general practitioners and ArtiQ.Spiro was common, occurring in 77% of cases.

Without use of spirometry, general practitioners only correctly diagnosed 41% of patients. However, this increased to 68% with use of the AI-supported spirometry software, according to researchers.

Additionally, the AI software had 88% diagnostic accuracy with the diagnoses made by the panel of pulmonologists.

“These results highlight the importance of integrating spirometry into primary care practices,” Smets told Healio.

“Future studies are evaluating the impact of the AI solution on primary care staff’s time and confidence to execute spirometry tests and its impact on patient pathways,” she added.

Reference: