Proprietary Algorithm. Beiswenger and Associates acquired the database of Healthcarelink, an Internet service which offered physicians and hospitals biometric data graphing. Eight years of data was obtained from the database.
While monitoring patients, Healthcarelink had discovered it could identify the onset of a respiratory infection two days or more before symptoms, such as a fever, through graphical analysis of patient data. To confirm their findings, Healthcarelink conducted two research studies, one involving five health clinics in Hong Kong (118 patients participating), and the second at Tianjin Haihe Hospital (50 healthcare workers participating). A later study involved 50 participants in a Pennsylvania community.
In addition to the data from the three Research Studies, we also acquired the data from 400 Infertility patients monitored by Healthcarelink. The total produced over 23,000 data points from which an algorithm was developed by the research team. Next, the data from the three Research Studies were re-analyzed with the new algorithm, and its accuracy was confirmed; Sensitivity: 96.7%, Specificity: 96.8%. We then knew we could reliably detect developing infections days before symptoms. The benefit is dramatic (click on the left graph).
The proprietary algorithm described above will be used in all of the concept monitors with the exception of the blood glucose monitor which uses an algorithm to obtain blood glucose readings.