Evaluating sensitivity and specificity of handheld point-of-care ultrasound testing for gynecologic pathology: a pilot study for use in low resource settings.

Background: Point-of-care ultrasound (POC-US) is a diagnostic test conducted at the site of patient care with direct interpretation by the clinician, providing immediate results. POC-US for gynecologic application is not well characterized by current literature yet has the potential to increase access in limited resource settings. We compared the diagnostics of three POC-US devices for gynecologic (GYN) pathology and then performed evaluation of sensitivity and specificity of a single best POC-US device for intended use in a low resource setting.

Methods: This is prospective, pilot descriptive study of 60 subjects. In part 1, comparison of three POC-US devices was performed. Twenty subjects underwent POC-US with three test units [GE Vscan (Vscan), Sonosite Iviz (Iviz), Philips Lumify (Lumify)] followed by diagnostic ultrasound (Dx-US) for reference imaging. Image quality and correlation for devices was scored by blinded reviewers and quantitative measurements of GYN pathology were compared. In part 2, forty subjects underwent POC-US validation with the highest scoring device (Lumify) and Dx-US for reference imaging. Concordance of POC-US operator-interpreted diagnosis with reference imaging interpretation were assessed by Cohen's unweighted kappa coefficient. Accuracy and agreement of POC-US were assessed by linear regression and Bland-Altman plot analysis. Sensitivity and specificity of POC-US for gynecologic pathologies were calculated.

Results: In aggregate qualitative measurements, Lumify and Iviz units performed superiorly to Vscan. There was no statistically significant difference in quantitative measurements between devices, but a trend towards lower mean error was seen for Lumify and Iviz as compared to Vscan. Lumify device had highest overall scoring and was selected for further testing. In validation comparison of Lumify to Dx-US, no statistically significant differences were found for measurements of endometrium, uterus, ovaries, adnexal pathology, or leiomyomata, (P < 0.02) with excellent agreement in operator-interpreted diagnosis (Kappa > 0.7). Sensitivity and specificity of detecting pathology was 80-100% with PPV and NPV 76-100%.

Conclusion: Among three POC-US devices, Lumify and Iviz devices show highest potential for successful application to clinical gynecologic ultrasound. Clinician-performed POC-US has high diagnostic accuracy, sensitivity, and specificity for basic GYN anatomy and pathology. POC-US is an acceptable and feasible diagnostic tool with potential for future application in a low resource setting to increase access to ultrasound.