Accuracy of magnetic resonance imaging for acute appendicitis in pregnant women: an updated diagnostic systematic review and meta-analysis.
Lucas Habiro Alves, Guilherme Franceschini Machado, Giovanna Cristina de Castro Martin et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and diagnostic meta-analysis assessed MRI accuracy for suspected acute appendicitis in pregnant women, including 34 studies with surgical or histopathological confirmation. Pooled sensitivity was 0.95 (95% CI: 0.91-0.98; I² = 0%) and specificity was 0.97 (95% CI: 0.96-0.98; I² = 55.3%). A positive MRI increased the estimated post-test probability from 30% to 93%, while a negative result reduced it to 2.2%. The authors conclude that non-contrast MRI performs well and recommend it after inconclusive ultrasound.
Key findings
- Pooled MRI sensitivity was 0.95 (95% CI: 0.91-0.98; I² = 0%).
- Pooled MRI specificity was 0.97 (95% CI: 0.96-0.98; I² = 55.3%).
- A positive MRI raised post-test probability from 30% to 93%; a negative result lowered it to 2.2%.
- Patient selection and index-test interpretation were common sources of bias; Deeks' test suggested potential publication bias (p = 0.0024).
Why it matters
Appendicitis assessment in pregnancy must balance diagnostic certainty with concern about fetal radiation exposure. These pooled estimates indicate that MRI can provide useful diagnostic information in this setting, with a negative result associated with a low post-test probability in the review’s analysis. The findings support considering MRI within an imaging pathway while accounting for study bias and heterogeneity.
What to keep in mind
The review found moderate heterogeneity in specificity, common risk-of-bias concerns in patient selection and test interpretation, and potential publication bias.