Ultrasound measurements of gestational sac diameter and crown-rump length in intrauterine pregnancy of uncertain prognosis: A systematic review

Background: Pelvic ultrasound examination is the diagnostic modality of choice for the determination of early pregnancy loss. Objective: To evaluate the diagnostic performance of mean gestational sac diameter and crown-rump length for identifying early pregnancy loss in patients with an intrauterine pregnancy of uncertain prognosis. Methods: We searched MEDLINE, Embase, Cochrane Library, Scopus, and other databases to identify studies of first-trimester pregnant patients with intrauterine pregnancy of uncertain prognosis that reported diagnostic accuracy for mean gestational sac diameter and crown-rump length. Two reviewers separately screened studies, extracted data, and assessed quality using modified QUADAS-2. We reported operating characteristics (sensitivity, specificity, likelihood ratios) of mean gestational sac diameter and crown-rump length for predicting early pregnancy loss. Results: Fourteen studies met the inclusion criteria. Study methods, cutoff values, and outcome standards varied widely. Across all clinical categories, studies consistently demonstrated high specificity for early pregnancy loss using mean gestational sac diameter and crown-rump length thresholds, ranging from 88% to 100%. In contrast, sensitivity was highly variable and generally low, ranging from 2% to 100%. The positive likelihood ratio was typically high (median = 30; interquartile range = 19–76), indicating that meeting diagnostic thresholds substantially increased the probability of early pregnancy loss. However, the negative likelihood ratio was less informative (median = 0.64; interquartile range = 0.46–0.85), limiting the ability of these criteria to reliably exclude early pregnancy loss. Conclusion: Mean gestational sac diameter and crown-rump length measurements have high specificity and high positive likelihood ratio for identifying early pregnancy loss. Their low sensitivity and higher negative likelihood ratio limit their use in ruling out early pregnancy loss. Conservative cutoff values are useful for confirming early pregnancy loss when clearly met; however, early ultrasound findings alone are often insufficient to rule out pregnancy loss with confidence.