PTU-063 Miss rates for colorectal cancer investigated with computer tomography scans

    loading  Checking for direct PDF access through Ovid

Abstract

Introduction

Computer tomography (CT) scans are often the initial investigation for patients suspected of colorectal cancer (CRC) because they are better tolerated than colonoscopy. This study aims to evaluate the CRC false negative results (or miss rates) for CT investigations of the colon.

Methods

This is a retrospective review of 773 consecutive CRC cases encountered at Royal Berkshire Hospital between 2014 and 2016. Evidence of CT investigations in the previous 3 years was obtained from computerised health records. Only CT scans with the indication suggestive that it was done for suspected bowel malignancy were labelled as ‘missed cancer’. CT scans done to investigate other abdominal organs were not considered. Statistical analysis was done with a confidence interval of 95%.

Results

5.4% of patients diagnosed with CRC had an unremarkable CT scan in the previous 3 years. The indications included mostly iron deficiency anaemia, change in bowel habit, weight loss, abdominal pain and rectal bleeding. Patients being missed were significantly older than the rest of CRC patients (78.2 years vs. 69.5 years, p=0.000003). A higher proportion of right sided cancers were missed as compared to left sided cancers (7.3% vs. 4.8%) but this was not statistically significant (p=0.36). Average time from scan to diagnosis was 512.9 days (1.4 years). Most CT studies (33 of 42, 79%) were after administration of oral contrast. Two were CT colonographies that missed one caecal cancer and one rectal cancer at 1.2 and 1.8 years before diagnosis.

Results

Concerning the entire cohort of CRC cases, mean age was 70, male:female ratio was 1:1.16, left sided lesions accounted for 60% and right sided lesions 34%. All of these values are comparable to national statistics.

Conclusions

Approximately 1 in 20 patients diagnosed with colorectal cancer had at least one CT scan with no evidence of bowel malignancy in the previous 3 years. Bowel cancer should not be easily excluded by an unremarkable CT scan if there is a high clinical suspicion, especially in a patient older than 70, regardless of the type of scan.

Related Topics

    loading  Loading Related Articles