Performance Data
In the most recent study, 104 women attending the colposcopy clinic were tested using the device alongside the normal colposcopic examination (an inherently subjective process). The results showed that, when used to identify women who needed treatment for high grade CIN, the referral device, Zilico's first product had:
- a sensitivity of 88%
- a specificity of 64%
- a positive predictive value (PPV) of 72%
The important metric for colposcopic impression is the predictive value because this is a measure of how accurately the colposcopist has been able to discriminate between women who require treatment and those who do not. The referral device's PPV of 72% compares favourably with the performance requirement set by the National Cervical Screening Programme for England which is a PPV of 65%, especially considering that the device removes the subjectivity and variability from the procedure.
A detailed review of the above results also showed that within the 104 women that there were a total of 22 biopsies of which 18 were high-grade CIN and 4 were negative (non-HG). All the biopsy sites were identified correctly by device demonstrating that the device can discriminate between HG CIN and other tissue types and provide clinically-important information.
The current Zilico referral product is the third generation device developed from research carried out at the Hallamshire Hospital in Sheffield. The first generation device consisted of a probe that was hard-wired to a PC which had to be sterilised between patients. The second generation device was wireless and included the electronics on the handset; however it still required sterilisation between patients. These earlier versions have been used in four studies carried out in Sheffield involving over 500 women in total and have been published in peer reviewed journals.