Novel Risk Score for 30-Day Adverse Events Following Colonoscopy in Older Adults
Authors
Min-Jae Kim ; Seoyoon Choi ; Jihye Park ; Soo Jung Park ; Jae Jun Park ; Jae Hee Cheon ; Tae Il Kim ; Yuna Kim ; Jie-Hyun Kim ; Young Hoon Youn ; Hyojin Park ; Jaeyoung Chun
Background/Aims: Physicians are challenged with balancing benefits and risks of performing colonoscopies in older adults. We identified adverse event risk factors in this population and developed a predictive risk score for colonoscopy-related adverse events.
Methods: From August 2017 to August 2022, 8,154 patients aged ≥60 years who underwent screening or diagnostic colonoscopies were enrolled at Gangnam Severance Hospital. The primary outcome was 30-day adverse events, defined as emergency room visits or unplanned hospitalizations post-colonoscopy. The frailty index calculated via laboratory findings (FI-LAB) was derived from blood test results and vital signs. A risk score was developed and categorized to predict colonoscopy-related adverse events. Data from 9,154 colonoscopies from September 2022 to December 2023 at two tertiary referral hospitals were used for internal and external validation.
Results: The mean age was 67.9 years (range, 60 to 94 years). The 30-day adverse event rate was 1.4%. Adverse events were independently associated with the use of aspirin (adjusted odds ratio [aOR], 2.24), P2Y12 inhibitors (aOR, 1.79), and anticoagulants (aOR, 2.47) and with moderate (aOR, 4.54) and high (aOR, 11.40) FI-LABs. The incidence of adverse events in the low-, moderate-, and high-risk groups were 0.3%, 2.2%, and 10.7%, respectively (p<0.001). The area under the receiver operating characteristic curve for the risk scores were 0.821, 0.856, and 0.757 for the derivation, internal, and external cohorts, respectively.
Conclusions: Colonoscopy-related adverse events in older adults were linked to frailty and medication use and were not dependent on age. This novel risk score supports personalized decision-making when performing colonoscopies in older adults.