目的复核河南省参与ＷＨＯ全球结核病耐药监测病人首次询问既往抗结核药物用药史（以下简称用药史）的准确性，评价用药史复核的作用及影响因素。方法由统一培训的县级医生对３０个耐药监测县首次询问无用药史的１０８０例病人进行用药史复核，通过病史启发、村医生帮助回忆等调查方法，填写用药史调查表。省级随机复核县级复核病例。结果省级复核用药史与县级复核结果一致率为９５７％，其中有用药史的一致率为８４４％，无用药史的一致率为１００％，首次询问无用药史者４７２％有用药史，耐药组用药率为６２０％，敏感组为３１７％，非结核感染用药者多于结核用药者（１１７∶１），既往用药时间＞１月者，耐药组多于敏感组（３７２％，１６８％），结核组多于非结核感染组（８２０％，３５９％），用药种类结核组以ＨＲ为多，非结核感染组以ＳＲ为多。结论用药史复核是耐药监测质量控制的有效方法，使监测结果更趋准确；对既往用药史的询问应规范提纲、内容、方法，人员统一培训，此工作可随病人督导进行，县级医生经培训后复核用药史结果可信。不能忽视非结核感染用药史。ＷＨＯ耐药监测新指南用药史＞１月的标准有待商榷，其对初始耐药的准确性有一定影响。 Objective To examine the accuracy of the history of WHO global tuberculosis drug resistance surveillance in Henan Province for the first time to inquire about the past history of anti-TB drug use (hereinafter referred to as drug history) and to evaluate the role and influencing factors of drug history review. Methods A total of 1,080 patients who first asked for drug-free history in 30 drug-resistant surveillance counties were reviewed by the county-level doctors who were uniformly trained. The medical history questionnaire was filled in by the investigation methods such as medical history inspiration and assistance from village doctors. Provincial random review county-level review of cases. Results The coincidence rate of provincial review drug use and county review was 957%, of which the coincidence rate of drug use history was 844% and that of nonuse medication history was 100%. 2% had a history of drug use, drug-resistant group was 62  0%, sensitive group was 31  7%, non-tuberculosis drug users than tuberculosis drug users (1  17: 1), past medication time> January (372%, 168%) in the drug-resistant group, more in the tuberculosis group than in the non-tuberculosis infection group (820%, 359%). There were more HR in the tuberculosis group than in the susceptible group Tuberculosis infection group to SR as much. Conclusion The review of medication history is an effective method for quality control of drug resistance monitoring, which makes the monitoring results more accurate. The inquiry of the past medication history should be standardized, outline, content, method, personnel training, this work can be carried out with the patient supervision, county doctors The results of the review of the medication history after the training are credible. Can not ignore the history of non-TB infection medication. A new guideline for WHO drug resistance surveillance The history of drug use> January is open to question, which has an impact on the accuracy of initial drug resistance.