Rawson Timothy M, Hernandez Bernard, Wilson Richard C, Ming Damien, Herrero Pau, Ranganathan Nisha, Skolimowska Keira, Gilchrist Mark, Satta Giovanni, Georgiou Pantelis, Holmes Alison H
National Institute for Health Research Health Protection Research Unit in Healthcare Associated Infections and Antimicrobial Resistance, Imperial College London, Hammersmith Campus, Du Cane Road, London W12 0NN, UK.
Centre for Antimicrobial Optimisation, Hammersmith Hospital, Imperial College London, Du Cane Road, London W12 0NN, UK.
JAC Antimicrob Resist. 2021 Feb 3;3(1):dlab002. doi: 10.1093/jacamr/dlab002. eCollection 2021 Mar.
Bacterial infection has been challenging to diagnose in patients with COVID-19. We developed and evaluated supervised machine learning algorithms to support the diagnosis of secondary bacterial infection in hospitalized patients during the COVID-19 pandemic.
Inpatient data at three London hospitals for the first COVD-19 wave in March and April 2020 were extracted. Demographic, blood test and microbiology data for individuals with and without SARS-CoV-2-positive PCR were obtained. A Gaussian Naive Bayes, Support Vector Machine (SVM) and Artificial Neural Network were trained and compared using the area under the receiver operating characteristic curve (AUCROC). The best performing algorithm (SVM with 21 blood test variables) was prospectively piloted in July 2020. AUCROC was calculated for the prediction of a positive microbiological sample within 48 h of admission.
A total of 15 599 daily blood profiles for 1186 individual patients were identified to train the algorithms; 771/1186 (65%) individuals were SARS-CoV-2 PCR positive. Clinically significant microbiology results were present for 166/1186 (14%) patients during admission. An SVM algorithm trained with 21 routine blood test variables and over 8000 individual profiles had the best performance. AUCROC was 0.913, sensitivity 0.801 and specificity 0.890. Prospective testing on 54 patients on admission (28/54, 52% SARS-CoV-2 PCR positive) demonstrated an AUCROC of 0.960 (95% CI: 0.90-1.00).
An SVM using 21 routine blood test variables had excellent performance at inferring the likelihood of positive microbiology. Further prospective evaluation of the algorithms ability to support decision making for the diagnosis of bacterial infection in COVID-19 cohorts is underway.
新冠病毒病（COVID-19）患者的细菌感染诊断具有挑战性。我们开发并评估了监督式机器学习算法，以支持对COVID-19大流行期间住院患者继发性细菌感染的诊断。
提取了伦敦三家医院在2020年3月和4月第一波COVID-19疫情期间的住院患者数据。获取了SARS-CoV-2聚合酶链反应（PCR）检测呈阳性和阴性个体的人口统计学、血液检测及微生物学数据。使用接受者操作特征曲线下面积（AUCROC）对高斯朴素贝叶斯、支持向量机（SVM）和人工神经网络进行训练并比较。性能最佳的算法（具有21个血液检测变量的SVM）于2020年7月进行了前瞻性试点。计算入院48小时内微生物样本呈阳性预测的AUCROC。
共识别出1186例个体患者的15599份每日血液检测结果用于训练算法；771/1186（65%）个体SARS-CoV-2 PCR检测呈阳性。166/1186（14%）患者在住院期间有具有临床意义的微生物学检测结果。用21个常规血液检测变量和8000多个个体检测结果训练的SVM算法表现最佳。AUCROC为0.913，敏感性为0.801，特异性为0.890。对54例入院患者（28/54，52% SARS-CoV-2 PCR阳性）进行的前瞻性检测显示AUCROC为0.960（95%置信区间：0.90 - 1.00）。
使用21个常规血液检测变量的SVM在推断微生物学检测呈阳性的可能性方面表现出色。目前正在对这些算法支持COVID-19队列中细菌感染诊断决策的能力进行进一步的前瞻性评估。
