False-positive iFOBT in colorectal cancer screening: Association with prescription drug use in The Malaysian Cohort.

BACKGROUND

Immunochemical faecal occult blood tests (iFOBTs) are preferred for population-based colorectal cancer (CRC) screening because of their improved diagnostic performance over guaiac-based tests, but false-positive (FP) results still pose a concern, often leading to unnecessary colonoscopy. This study investigated whether chronic medications, alongside proton pump inhibitors (PPIs), nonsteroidal anti-inflammatory drugs (NSAIDs), and antiplatelet agents, influence FP iFOBT risk.

METHODS

Participants aged 38-79 who underwent colonoscopy after a positive iFOBT in CRC screening were included. We collected prescription data for statins, antiplatelet agents, alpha-blockers, beta-blockers, calcium channel blockers (CCBs), angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), PPIs, and NSAIDs. Drug intake referred to medication exposure from baseline to the latest visit before screening. An FP result was a positive iFOBT with no CRC, polyps, or other diseases on colonoscopy; a true positive (TP) required such pathology. Logistic regression evaluated associations between these drugs and FP risk.

RESULTS

Of 438 eligible participants, 59.8% were on medication. The FP iFOBT rate was 49.3%. Statins were most used (n = 188), followed by CCBs (n = 102), NSAIDs (n = 55), ARBs (n = 49), ACE inhibitors (n = 46), beta-blockers (n = 43), PPIs (n = 16), antiplatelet (n = 12), and alpha-blockers (n = 8). Statin use was higher in the TP group (56.4% vs. 43.6% in FP; P = 0.039), while the FP group had more alpha-blockers (87.5% vs. 12.5% in TP; P = 0.029). Certain drug combinations, like antiplatelet plus statins (OR 0.658; 95% CI 0.450, 0.962; P = 0.031) and PPIs plus antihypertensives/CVD (OR 0.673; 95% CI 0.458, 0.990; P = 0.044), were initially associated with lower FP risk in unadjusted analyses. However, these associations became non-significant after adjusting for confounders, particularly age and sex (all P > 0.05), indicating that age and sex were more influential factors.

CONCLUSION

Use of statins, CCBs, NSAIDs, PPIs, and antiplatelet agents, alone or in combination, was not independently associated with higher FP iFOBT risk after adjustment. However, inclusion of age and sex in the model attenuated these associations, suggesting that these may act as important confounding factors rather than independent determinants. These findings highlight the importance of considering host factors beyond medication use when interpreting iFOBT in CRC screening.