Abstract
Purpose Ovarian torsion (OT) is a true surgical emergency that requires prompt diagnosis
and therapy. There are conflicting reports on the accuracy of different imaging modalities
for OT. In this study, we performed a meta-analysis of all the published studies on
B-mode ultrasound (US, morphological criteria), Doppler US (DUS, flow criteria), and
computed tomography (CT) for the diagnosis of OT.
Methods The medical literature from 1987 to 2013 was searched for studies that evaluated
US, DUS, CT, or combination of these techniques to diagnose OT in children using PubMed/MEDLINE
database. The studies were screened and included if the quality criteria were met.
Data were extracted using a standardized form. Reported sensitivities and specificities
were pooled with 95% confidence intervals using a RevMan version 5.1 software (The
Cochrane Collaboration, Oxford, United Kingdom).
Results A total of 18 US studies with 716 combined patients were included, along with 15
DUS studies comprising of 1,021 patients, and 5 CT studies with a total of 107 patients.
All studies reported sensitivities, while only few selected studies also reported
specificities. Using morphological criteria on B-mode US (including absolute or relative
ovarian size, echotexture, location, and configuration) yielded high sensitivity and
specificity (average 92 and 96%, respectively); DUS was highly specific, but lacked
sensitivity, while CT was the least sensitive test. CT specificities could not be
calculated from the available data.
Conclusion This study found considerable variability of the reported sensitivities and specificities
for the diagnosis of OT across all evaluated imaging modalities. Interestingly, B-mode
US was the most sensitive and specific examination to detect OT. Some authors reported
high diagnostic accuracy of DUS as well, but these findings have not been universally
reproduced. CT had low overall sensitivity and is not recommended for the workup of
suspected OT.
Keywords
ovarian torsion - ultrasound - Doppler ultrasound - computed tomography - meta-analysis
