Prediction model could prioritize high-risk patients awaiting surveillance colonoscopy
Click Here to Manage Email Alerts
Key takeaways:
- Predictors of colorectal cancer at surveillance colonoscopy included age, smoking status, adenoma detection rate and polyp size.
- The risk prediction model outperformed a guideline-based model.
A multivariable risk prediction model “may be useful” for prioritizing patients at highest risk for colorectal cancer and awaiting surveillance colonoscopy when demand for the procedure is high, according to research in Gastro Hep Advances.
“Despite polyp removal, some patients remain at elevated risk for CRC and are recommended to undergo colonoscopy surveillance,” Theodore R. Levin, MD, of the division of research at Kaiser Permanente Northern California, and colleagues wrote. “Surveillance is a common colonoscopy indication and at times demand for colonoscopy exceeds capacity, necessitating that patient requiring surveillance colonoscopy be placed on a waiting list.
“However, for providers managing these waiting lists, there are no tools available to prioritize those most likely to be diagnosed with CRC among the patients awaiting surveillance colonoscopy.”
In a cross-sectional study, Levin and colleagues developed and validated a multivariable prediction model to identify patients at highest risk for CRC at surveillance, which they compared with a guideline-based model that categorizes patients based on previous polyp findings.
Researchers included 52,848 patients (mean age, 66 years; 41.5% women; 50.9% white) who underwent surveillance colonoscopy between 2014 and 2019 and divided them into model development (n = 36,994) and internal validation (n = 15,854) cohorts. They performed external validation on 30,015 patients who underwent surveillance colonoscopy between 2020 and 2022.
In the development and internal validation cohorts, screening was the indication for 27.7% of index colonoscopies, and adenomas were discovered at 68.6% of those procedures.
According to study results, there were 114 and 43 incident cases of CRC reported in the two cohorts, respectively, with an absolute risk of 5.43 and 5.08 per 10,000 patient-years of follow-up. Seventy-one incident cases were detected in the external validation cohort.
In univariate analysis, researchers identified seven variables that were significantly associated with risk for CRC at surveillance, four of which remained significant in multivariable analysis — increasing patient age, having ever smoked tobacco, endoscopist all-indication adenoma detection rate of less than 32.5% or missing, and polyp size of at least 10 mm.
In addition, the multivariable model “outperformed” the guideline-based model in both development (area under the curve [AUC] = 0.71; 95% CI, 0.67-0.76 vs. AUC = 0.53; 95% CI, 0.49-0.85) and internal validation (AUC = 0.73; 95% CI, 0.66-0.81 vs. AUC = 0.52; 95% CI, 0.45-0.6) cohorts. Although performance declined at external validation (AUC = 0.61; 95% CI, 0.56-0.67 vs. AUC = 0.51; 95% CI, 0.45-0.56), an updated multivariable model yielded an AUC of 0.72 (95% CI, 0.66-0.77).
“The clinical implication of the present study is that at those times when surveillance colonoscopy demand exceeds capacity and patients are on a waiting list for the procedure, risk prediction modeling may be useful for prioritizing patients at highest CRC risk using variables that are routinely collected in clinical practice or for quality reporting,” Levin and colleagues wrote. “These patients could be considered for more intensive efforts to ensure prompt surveillance colonoscopy.”
Collapse
Read more about
Click Here to Manage Email Alerts