The Development and Validation of a Machine Learning Model to Predict Bacteremia and Fungemia in Hospitalized Patients Using Electronic Health Record Data.

OBJECTIVES: Bacteremia and fungemia can cause life-threatening illness with high mortality rates, which increase with delays in antimicrobial therapy. The objective of this study is to develop machine learning models to predict blood culture results at the time of the blood culture order using routine data in the electronic health record. DESIGN: Retrospective analysis of a large, multicenter inpatient data. SETTING: Two academic tertiary medical centers between the years 2007 and 2018. SUBJECTS: All hospitalized patients who received a blood culture during hospitalization. INTERVENTIONS: The dataset was partitioned temporally into development and validation cohorts: the logistic regression and gradient boosting machine models were trained on the earliest 80% of hospital admissions and validated on the most recent 20%. MEASUREMENTS AND MAIN RESULTS: There were 252,569 blood culture days-defined as nonoverlapping 24-hour periods in which one or more blood cultures were ordered. In the validation cohort, there were 50,514 blood culture days, with 3,762 cases of bacteremia (7.5%) and 370 cases of fungemia (0.7%). The gradient boosting machine model for bacteremia had significantly higher area under the receiver operating characteristic curve (0.78 [95% CI 0.77-0.78]) than the logistic regression model (0.73 [0.72-0.74]) (p < 0.001). The model identified a high-risk group with over 30 times the occurrence rate of bacteremia in the low-risk group (27.4% vs 0.9%; p < 0.001). Using the low-risk cut-off, the model identifies bacteremia with 98.7% sensitivity. The gradient boosting machine model for fungemia had high discrimination (area under the receiver operating characteristic curve 0.88 [95% CI 0.86-0.90]). The high-risk fungemia group had 252 fungemic cultures compared with one fungemic culture in the low-risk group (5.0% vs 0.02%; p < 0.001). Further, the high-risk group had a mortality rate 60 times higher than the low-risk group (28.2% vs 0.4%; p < 0.001). CONCLUSIONS: Our novel models identified patients at low and high-risk for bacteremia and fungemia using routinely collected electronic health record data. Further research is needed to evaluate the cost-effectiveness and impact of model implementation in clinical practice.

Identification of a new gene required for the biosynthesis of rhodoquinone in Rhodospirillum rubrum.

Rhodoquinone (RQ) is a required cofactor for anaerobic respiration in Rhodospirillum rubrum, and it is also found in several helminth parasites that utilize a fumarate reductase pathway. RQ is an aminoquinone that is structurally similar to ubiquinone (Q), a polyprenylated benzoquinone used in the aerobic respiratory chain. RQ is not found in humans or other mammals, and therefore, the inhibition of its biosynthesis may provide a novel antiparasitic drug target. To identify a gene specifically required for RQ biosynthesis, we determined the complete genome sequence of a mutant strain of R. rubrum (F11), which cannot grow anaerobically and does not synthesize RQ, and compared it with that of a spontaneous revertant (RF111). RF111 can grow anaerobically and has recovered the ability to synthesize RQ. The two strains differ by a single base pair, which causes a nonsense mutation in the putative methyltransferase gene rquA. To test whether this mutation is important for the F11 phenotype, the wild-type rquA gene was cloned into the pRK404E1 vector and conjugated into F11. Complementation of the anaerobic growth defect in F11 was observed, and liquid chromatography-time of flight mass spectrometry (LC-TOF-MS) analysis of lipid extracts confirmed that plasmid-complemented F11 was able to synthesize RQ. To further validate the requirement of rquA for RQ biosynthesis, we generated a deletion mutant from wild-type R. rubrum by the targeted replacement of rquA with a gentamicin resistance cassette. The ΔrquA mutant exhibited the same phenotype as that of F11. These results are significant because rquA is the first gene to be discovered that is required for RQ biosynthesis.

Evidence that ubiquinone is a required intermediate for rhodoquinone biosynthesis in Rhodospirillum rubrum.

Rhodoquinone (RQ) is an important cofactor used in the anaerobic energy metabolism of Rhodospirillum rubrum. RQ is structurally similar to ubiquinone (coenzyme Q or Q), a polyprenylated benzoquinone used in the aerobic respiratory chain. RQ is also found in several eukaryotic species that utilize a fumarate reductase pathway for anaerobic respiration, an important example being the parasitic helminths. RQ is not found in humans or other mammals, and therefore inhibition of its biosynthesis may provide a parasite-specific drug target. In this report, we describe several in vivo feeding experiments with R. rubrum used for the identification of RQ biosynthetic intermediates. Cultures of R. rubrum were grown in the presence of synthetic analogs of ubiquinone and the known Q biosynthetic precursors demethylubiquinone, demethoxyubiquinone, and demethyldemethoxyubiquinone, and assays were monitored for the formation of RQ(3). Data from time course experiments and S-adenosyl-l-methionine-dependent O-methyltransferase inhibition studies are discussed. Based on the results presented, we have demonstrated that Q is a required intermediate for the biosynthesis of RQ in R. rubrum.