Back

Clinical Diagnosis Codes Identify Patients Unlikely To Receive Orders For Fecal Immunochemical Tests

Li, A. T.; Burns, S. T.; Martinez, D.; Wang, P.; Aggarwal, S. L.; Potter, M.; Sarkar, U.; Somsouk, M.

2025-06-09 gastroenterology
10.1101/2025.06.07.25327375 medRxiv
Show abstract

IntroductionOrganized screening programs automate fecal immunochemical test (FIT) outreach based on chronological age. However, providers and patients may defer screening due to coexisting conditions, which can be captured by International Classification of Disease, 10th revision (ICD-10) codes. We sought to identify codes associated with not having a FIT order. MethodsWe included screen-eligible patients with a primary care visit between June 2022 and June 2023. We enumerated and compared the frequency of each ICD-10 code between patients with and without a FIT ordered. We conducted a subgroup analysis including only those with a Charlson comorbidity index (CCI) <5. ResultsWe identified 15,020 screen-eligible patients, with 10,187 (67.8%) patients having had a FIT order and 4,833 (32.3%) patients without a FIT order. Of the 1,215 ICD-10 codes examined, 96 were significantly associated with not having a FIT order. One broad category of codes pertained to digestive diseases such as benign and malignant colonic neoplasms (ICD-10 code C18, odds ratio (OR) 0.06, 95% CI [0.001-0.44]) and diverticular disease (K57, OR 0.26 [0.18-0.36]). Another category was comorbid conditions which included frailty (R54, OR 0.14 [0.03-0.55]) and paralysis (G82, OR 0.14 [0.03-0.55]). Those with acute conditions such as cervical fractures (S12, OR 0.23 [0.10-0.48]) and cryptococcosis (B45, OR 0.05 [0.001-0.38]) were also less likely to have a FIT order. Among the patients with CCI<5, 41 codes were significantly associated with not having a FIT order, including heart failure (I50, OR 0.55 [0.43-0.71]) and chronic kidney disease (N18, OR 0.58 [0.43-0.77]). ConclusionPatients deferred from screening were more likely to have ICD-10 codes signifying digestive diseases, comorbidities, and acute conditions. Even among patients with fewer co-morbid conditions, we identified health conditions that were associated with screening deferral. Future work should consider whether screening programs could incorporate ICD-10 codes to align FIT outreach more closely with provider and patient preferences.

Published in Cancer Control · not in our set (fewer than 10 published preprints to learn from) · training set

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.