Adherence to criteria for transvaginal ultrasound imaging and measurement of cervical length

Abstract

Objective: Adherence to published criteria for transvaginal imaging and measurement of cervical length is uncertain. We sought to assess adherence by evaluating images submitted to certify research sonographers for participation in a clinical trial.

Study design: We reviewed qualifying test results of sonographers seeking certification to image and measure cervical length in a clinical trial. Participating sonographers were required to access training materials and submit 15 images, 3 each from 5 pregnant women not enrolled in the trial. One of 2 sonologists reviewed all qualifying images. We recorded the proportion of images that did not meet standard criteria (excess compression, landmarks not seen, improper image size, or full maternal bladder) and the proportion in which the cervical length was measured incorrectly. Failure for a given patient was defined as >1 unacceptable image, or >2 acceptable images with incorrect caliper placement or erroneous choice of the "shortest best" cervical length. Certification required satisfactory images and cervical length measurement from ≥4 patients.

Results: A total of 327 sonographers submitted 4905 images. A total of 271 sonographers (83%) were certified on the first, 41 (13%) on the second, and 2 (0.6%) on the third submission. Thirteen never achieved certification. Of 314 who passed, 196 submitted 15 acceptable images that were appropriately measured for all 5 women. There were 1277 deficient images: 493 were acceptable but incorrectly measured images from sonographers who passed certification because mismeasurement occurred no more than twice. Of 784 deficient images submitted by sonographers who failed the certification, 471 were rejected because of improper measurement (caliper placement and/or failure to identify the shortest best image), and 313 because of failure to obtain a satisfactory image (excessive compression, required landmarks not visible, incorrect image size, brief examination, and/or full maternal bladder).

Conclusion: Although 83% of sonographers were certified on their first submission, >1 in 4 ultrasound images submitted did not meet published quality criteria. Increased attention to standardized education and credentials is warranted for persons who perform ultrasound examinations of the cervix in pregnancy.

Keywords: cervix; preterm birth; ultrasound.

Figure 1

Transvaginal ultrasound image of the cervix demonstrating compression of the anterior cervix and failure to image the external os.

Figure 2

Cervical Ultrasound Images Submitted for Review n = 4905 Eunice Kennedy Shriver NICHD MFMU SCAN Trial