Abstract：
Background: Challenges in identifying microsatellite instability (MSI)/mismatch repair (MMR)-tested colorectal carcinoma (CRC) patients in electronic health records have led to gaps in the understanding of MSI-high/deficient mismatch repair prevalence. Methods: An algorithm to identify MSI-/MMR-tested Veterans Affairs patients was developed and an observational study of adult CRC patients with MSI/MMR testing from 2010 to 2018 was undertaken. Results: An optimized model to identify MSI-/MMR-tested patients yielded high positive predictive value (89.0%) and specificity (97.8%). The authors observed MSI-high/deficient mismatch repair CRC in 54 of 291 patients (18.6%); highest frequencies were observed in stages II (25.9%) and III (22.6%) and lowest in stage IV (5.8%). Conclusions: In this real-world study, the authors proposed a novel method of identifying MSI-/MMR-tested patients. Further validation and refinement of this model, and study in a larger CRC cohort, is warranted.
摘要：
背景：在电子健康记录中识别微卫星不稳定性（MSI）/错配修复（MMR）测试的结直肠癌（CRC）患者的挑战导致对MSI高/缺陷错配修复患病率的理解存在差距。方法：开发了一种识别MSI/MMR测试的退伍军人事务患者的算法，并对2010年至2018年进行了MSI/MMR测试的成年CRC患者进行了观察性研究。结果：鉴定MSI/MMR测试患者的优化模型产生了高阳性预测值（89.0％）和特异性（97.8％）。作者在291例患者中的54例（18.6％）中观察到MSI高/缺陷错配修复CRC；在II期（25.9％）和III期（22.6％）中观察到最高频率，在IV期（5.8％）中最低。结论：在这项现实世界的研究中，作者提出了一种鉴定MSI/MMR检测患者的新方法.进一步验证和完善该模型，在一个更大的CRC队列中进行研究，是有保证的。
