目的 構(gòu)建并驗(yàn)證環(huán)孢素血藥濃度超限列線圖預(yù)測(cè)模型,篩選血藥濃度超限的高危人群,提高環(huán)孢素治療的安全性。方法 回顧性分析2019年10月-2022年10月西安國(guó)際醫(yī)學(xué)中心醫(yī)院服用環(huán)孢素并接受血藥濃度監(jiān)測(cè)的207例住院患者的臨床資料,采用Logistic回歸方程分析環(huán)孢素血藥濃度超限的影響因素,構(gòu)建列線圖預(yù)測(cè)模型,并對(duì)該模型進(jìn)行評(píng)估。結(jié)果 日劑量、合用唑類(lèi)抗真菌藥是環(huán)孢素血藥濃度低于正常值下限的獨(dú)立影響因素;日劑量、腎功能不全、合用達(dá)那唑、合用糖皮質(zhì)激素、合用唑類(lèi)抗真菌藥是環(huán)孢素血藥濃度超出正常值上限的獨(dú)立影響因素。根據(jù)所篩選出的5項(xiàng)環(huán)孢素血藥濃度超出正常值上限的獨(dú)立影響因素,建立了環(huán)孢素血藥濃度超限的列線圖預(yù)測(cè)模型,該模型的C-index指數(shù)為0.841,表明該列線圖模型具有較好的區(qū)分度,對(duì)該模型進(jìn)行驗(yàn)證后發(fā)現(xiàn)預(yù)測(cè)值和觀察值基本一致,表明該列線圖預(yù)測(cè)模型具有較好的一致性。結(jié)論 構(gòu)建的環(huán)孢素血藥濃度超限列線圖預(yù)測(cè)模型有良好的區(qū)分度及一致性,具有較高的臨床應(yīng)用價(jià)值,可甄別易發(fā)生環(huán)孢素血藥濃度超限的高危人群,有目的性地開(kāi)展血藥濃度監(jiān)測(cè),提高藥物治療安全性。

Objective To construct and verify the prediction model of the blood concentration of cyclosporine exceeding the limit, screen the high-risk groups of the blood concentration exceeding the limit, and improve the safety of the treatment of cyclosporine. Methods The clinical data of 207 inpatients who took cyclosporine and received blood concentration monitoring in Xi'an International Medical Center Hospital from October 2019 to October 2022 were retrospectively analyzed. The influencing factors of excessive blood concentration of cyclosporine were analyzed by using Logistic regression equation, and the prediction model was built and evaluated. Results Daily dose and the combination of azole antifungal drugs were the independent factors influencing the blood concentration of cyclosporine below the lower limit of normal value. Daily dose, renal insufficiency, combination of danazol, glucocorticoid, and azole antifungal drugs were independent factors influencing the concentration of cyclosporine beyond the upper limit of normal value. Based on the five selected independent factors affecting the concentration of cyclosporine exceeding the upper limit of normal value, a column-line prediction model for the concentration of cyclosporine exceeding the upper limit of normal value was established. The C-index of the model was 0.841, indicating that the column-line model had a good degree of differentiation. After verifying the model, it was found that the predicted value was basically consistent with the observed value. The results show that the prediction model has good consistency. Conclusion The prediction model built in this study has good differentiation and consistency, and has high clinical application value, which can identify high-risk groups prone to excessive blood concentration of cyclosporine, carry out purposeful blood concentration monitoring, and improve the safety of drug treatment.

R978.1