Purpose To assess the value of the administration of positive rectal contrast at CT in patients referred for suspected diverticular disease (DD) of the colon.
Materials and Methods 460 patients (253 male, 207 female; median age 62 years; interquartile range 24) with clinical suspicion of DD of the colon were included in this retrospective IRB-approved study. CT was performed with i. v. contrast only (n = 328, group M1), i. v. + positive rectal contrast (n = 82, group M2), neither i. v. nor rectal contrast (n = 32, group S1), or positive rectal contrast only (n = 19, group S2). Two readers in consensus evaluated all CT datasets concerning diagnosis of DD (yes/no) and categorized findings (classification of diverticular disease (CDD)). Sensitivity, specificity, positive (PPV) and negative (NPV) predictive values for the diagnosis of DD were calculated for all groups, using either clinical follow-up (n = 335) or intraoperative findings (n = 125) as the reference standard. In patients undergoing surgery, radiological staging of DD was correlated with the histopathology (weighted Cohen-k).
Results 224 patients (48.7 %) were diagnosed with DD. The sensitivity, specificity, PPV, and NPV were as follows. Group M1 / M2: 92 %/92 %, 97 %/94 %, 96 %/96 %, 94 %/89 %, respectively; group S1 / S2: 94 %/86 %, 93 %/80 %, 94 %/92 %, 93 %/67 %, respectively. Radiological staging and histopathology correlated substantially in all groups (k = 0.748–0.861).
Conclusion Abdominal CT had a high sensitivity and specificity for the diagnosis of DD. Disease staging correlated well with the findings at surgery. Additional positive rectal contrast administration did not have a significant advantage and may therefore be omitted in patients with suspected DD.
Key Points:
Citation Format