Incremental value of hernia-sac CT attenuation beyond symptom duration for predicting bowel resection in incarcerated obturator hernia: a single-site retrospective cohort study

Symptom duration ≥ 72 h increases bowel resection risk in incarcerated obturator hernia, but onset is often imprecise, especially in patients without pain. We investigated whether hernia-sac computed tomography (CT) attenuation improves prediction beyond symptom duration.

In this single-center retrospective cohort study of patients treated between 2006 and 2023, onset was determined from electronic medical records using the best available clinical information. On non-contrast CT, mean hernia-sac attenuation in Hounsfield units (HU) was measured using a circular region of interest on the slice with the largest hernia-sac area. Firth logistic regression was used with bowel resection as primary outcome. In patients with HU available, incremental value of ≥ 20 HU beyond symptom duration ≥ 72 h was tested with likelihood ratios and changes in the area under the receiver operating characteristic curve (AUC); patients without pain were explored separately.

Among 39 patients (median age 88y, median body mass index 16.3 kg/m²), symptom duration ≥ 72 h strongly associated with bowel resection (odds ratio 18.6, 95% confidence interval 4.07-120.7; p < 0.001). HU did not improve model fit in 23 patients (p = 0.99 for ≥ 20 HU, p = 0.82 continuous value) and minimally changed discrimination (ΔAUC ≤ 0.01). In the exploratory subgroup of HU-assessed pain-free patients (n = 7), HU ≥ 20 yielded sensitivity 1.00 and specificity 0.75, with wide exact confidence intervals.

Hernia-sac CT attenuation added limited predictive value beyond symptom duration for bowel resection. In an exploratory subgroup of patients without pain, no bowel resection was observed with HU < 20, but this finding requires validation because of small sample size.