RESEARCH ARTICLE Factors affecting sedimentational separation of bacteria from blood William G. Pitt 1 | Mahsa Alizadeh 1 | Rae Blanco 1 | Alex K. Hunter 1 | Colin G. Bledsoe 1 | Daniel S. McClellan 1 | Madison E. Wood 2 | Ryan L. Wood 1 | Tanner V. Ravsten 1 | Caroline L. Hickey 1 | William Cameron Beard 3 | Jacob R. Stepan 3 | Alexandra Carter 1 | Ghaleb A. Husseini 4 | Richard A. Robison 2 | Evelyn Welling 1 | Rebekah N. Torgesen 1 | Clifton M. Anderson 1 1 Chemical Engineering Department, Brigham Young University, Provo, Utah 2 Department of Microbiology and Molecular Biology, Brigham Young University, Provo, Utah 3 Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 4 Chemical Engineering Department, American University of Sharjah, Sharjah, United Arab Emirates Correspondence William G. Pitt, Chemical Engineering Department, Brigham Young University, 330W Engineering Building, Provo, UT 84602. Email: pitt@byu.edu Funding information National Institutes of Health, Grant/Award Number: R01AI116989 Abstract Rapid diagnosis of blood infections requires fast and efficient separation of bacteria from blood. We have developed spinning hollow disks that separate bacteria from blood cells via the differences in sedimentation velocities of these particles. Factors affecting sepa- ration included the spinning speed and duration, and disk size. These factors were varied in dozens of experiments for which the volume of separated plasma, and the concentra- tion of bacteria and red blood cells (RBCs) in separated plasma were measured. Data were correlated by a parameter of characteristic sedimentation length, which is the dis- tance that an idealized RBC would travel during the entire spin. Results show that char- acteristic sedimentation length of 20 to 25 mm produces an optimal separation and collection of bacteria in plasma. This corresponds to spinning a 12-cm-diameter disk at 3,000 rpm for 13 s. Following the spin, a careful deceleration preserves the separation of cells from plasma and provides a bacterial recovery of about 61 ± 5%. KEYWORDS bacterial bloodstream infection, bacterial separation, centrifugation, disk design, E. coli, human blood, sedimentation 1 | INTRODUCTION There is increasing need to rapidly identify the species and antibiotic resistance profile of bacteria causing bloodstream infections (BSI) so proper treatment commences quickly. Empirical applications of broad- spectrum antibiotics often work satisfactorily, but the general use of these agents may promote the development of resistant bacterial strains. 1,2 Unfortunately, when undiagnosed BSIs of carbapenem- resistant organisms produce a state of sepsis in a patient, the survival rate decreases by 7.6% per hour of ineffective treatment, 3 eventually producing a 50% mortality rate. 4,5 Thus rapid identification of species and antibiotic resistance are essential. Because the bacterial concentration in BSIs is extremely low (often between 1 and 100 CFU/ml 6,7 ), growth amplification is required for most diagnostics. 8 While PCR and other genomic amplifi- cation techniques are rapid and work well for the high concentrations found in urine, 9-11 they are often ineffective in blood samples due to interfering agents and degradative enzymes. 12 Our proposed diagnostic process described herein requires only 7 ml of blood, a typical volume collected in vacutainer tubes. The challenge is to rapidly separate a maximal amount of bacteria from ~5 × 10 9 red blood cells (RBCs), ~5 × 10 6 white blood cells (WBCs) and ~3 × 10 8 platelets per ml of blood. Table 1 and Table S1 (Supporting Information) tabulate some important characteristics of blood components. Received: 14 January 2019 Revised: 19 June 2019 Accepted: 1 August 2019 DOI: 10.1002/btpr.2892 Biotechnology Progress. 2019;e2829. wileyonlinelibrary.com/journal/btpr © 2019 American Institute of Chemical Engineers 1 of 9 https://doi.org/10.1002/btpr.2892