Enrique Oyarzún, Harumi Tsunekawa, José Badía, Beltrán Mena, Alfredo Hernández, Galo Andrade, Andres Poblete, Pedro González, Braulio Herrera and the Telemedicine Project Group.
Department of Obstetrics and Gynecology, Catholic University School of Medicine, Santiago, Chile.
Purpose: To evaluate distant ultrasound fetal measurements using a desktop computer video conferencing system over a dedicated Ethernet network, assessing simultaneously the interobserver variability. Methods: Ultrasound examinations (USE) were performed by a non specialist physician directed by an expert from a distant centre. Later, a second specialist performed a new USE on the patient, not aware of the transmitted results. A compound video signal from a Phillips P700 ultasound unit was input to a Macintosh PowerPC 8500/180. A video-conference was established through modified Apple Videophone software over the network with a similar distant computer. The transmitted digital video was 320x240, 24bits, 15fps and used AppleVideo compression. Biparietal diameter (BPD), abdominal circunference (AC) and femur lenght (FL) were registered in 55 pregnant women (20 to 40 weeks gestation), choosen randomly from our outpatient clinic. The measurements obtained were compared using correlation coefficient, variation coefficient and limits of agreement. Results: All the correlation coefficients are statistically significant (p=0.0001).
|
|
|
|
|
|
Correlation Coefficient |
0.99 |
0.97 |
0.97 |
|
Coefficient of Variation (%) |
-1.855 |
-74.116 |
5.871 |
|
Difference Mean |
-1.127 |
-0.164 |
0.309 |
|
SD |
(2.091) |
(12.155) |
(1.814) |
|
Limits of Agreement |
-6.2 to 4 |
-29.9 to 29.5 |
-4.1 to 4.7 |
|
Systematic Bias |
0.0002 |
NS |
NS |
Conclusion: This method is safe and reproducible. Significant differences were found in the BPD measurement in the limits of agreement analysis. This cannot be attributed to the transmission displaying qualities of the video-conference system, but to the standard interobserver variability described for standard BPD measurement.