A custom fidget spinner that uses centrifugal forces to concentrate bacterial pathogens in urine samples enables the rapid on-device colorimetric detection of urinary tract infections and the testing of the pathogen’s susceptibility to antibiotics.
The detection of bacterial infections has considerable limitations. On one hand, it often requires trained technicians and bulky equipment that is not available in rural communities in resource-limited settings; on the other hand, conventional culture-based methods for the detection of bacterial infections take three to five days. The lack of rapid tools for the detection of bacterial pathogens increases the usage of broad-spectrum antibiotics, leading to unnecessary treatment, to the perturbation of the body’s gut microbiota and to ineffective antibiotic selection, which is associated with poor clinical outcomes and, critically, promotes the emergence of multidrug-resistant bacteria. Therefore, the fast determination of the presence or absence of bacterial pathogens at a clinically significant concentration is an important unmet clinical need, which ideally should be complemented with the ability to rapidly characterize the antimicrobial resistance profile of the bacterial pathogen to assist in selecting the most effective and specific antibiotics for the patient. Reporting in Nature Biomedical Engineering, Yoon-Kyoung Cho and colleagues now describe a palm-sized point-of-care testing device resembling a fidget spinner for the rapid diagnosis of bacterial infections1.
Cho and co-authors’ device consists of a multi-lobed plate which serves as a hand-powered centrifuge that is spun with the help of a ball bearing to enrich bacteria from biological samples (such as urine) on a filtration membrane (Fig. 1a). Manually spinning the device for 1–3 minutes enables over a 100-fold reduction in volume, effectively concentrating the bacteria in suspension and therefore bypassing the time-consuming bacteria culturing step. The spinning step (only one or two spins are required to process one cubic millilitre of liquid) can be confirmed visually by inspecting the loading chamber for any remaining volume. A colorimetric viability dye is then loaded into the device for determining the abundance of viable bacteria by the naked eye, or with the support of a mobile phone camera or a digital single-lens reflex camera. The whole process takes approximately 50 minutes, and the authors show that it can quantify bacteria cells in urine in the range of 103–106 colony-forming units (c.f.u.) per millilitre (Fig. 1b).
To address variabilities in actual device usage — a fundamental consideration of point-of-care diagnostic systems — Cho and colleagues optimized the flow dynamics in the device by using fluid-assisted separation technology previously developed for tumour-cell separation from whole blood2. In particular, the device was constructed so that the centrifugal force remains perpendicular to the filtration membrane; also, the pressure balance in the device adjusts itself during operation so that the back of the filtration membrane is always filled with liquid. These design features ensure a uniform pressure across the membrane which maximizes the membrane’s usable area, prevents it from clogging and improves the overall isolation efficiency. Such optimizations should help standardize device use, as operator variability (particularly in terms of the spinning speed achieved) may introduce human error and inconsistency in the results, especially in resource-limited environments where users may not have received training. Also, the relatively low spinning speed that is needed (with respect to electric-powered and other hand-powered centrifuges3) may exacerbate differences in user operation.
Cho and co-authors also show that the diagnostic fidget spinner can be used for antimicrobial susceptibility testing (AST). The workflow was designed for samples containing the bacterium Escherichia coli, the most common cause of urinary tract infection. Once samples positive for the bacterium were identified via recombinase polymerase amplification followed by a lateral-flow assay, or with an immunoassay featuring anti-E. coli antibody-labelled gold nanoparticles, the authors performed on-device phenotypic AST by first exposing the sample containing the bacteria to an antibiotic for 20 minutes, followed by in-device enrichment and colorimetric detection. The workflow took less than two hours, whereas conventional culture-based AST takes two to three days. In a pilot AST study with 30 patient samples, the device led to the discrimination of samples with bacteria that were either resistant or susceptible to the antibiotics ciprofloxacin or cefazolin (Fig. 1c). Moreover, the authors tested the applicability of the fidget spinner for the diagnosis of urinary tract infection in the Kauvery Hospital in Tiruchirappalli, India, with samples from 39 individuals suspected of having community-acquired urinary tract infections. Culture assays performed in the hospital’s microbiology laboratory using a clinical cut-off measurement of 103 c.f.u. ml–1 (as per the European urinalysis guidelines for urinary tract infection) indicated that over half of the samples were negative for infection; however, two of the culture-negative samples had high numbers of pus cells and red blood cells in subsequent urinalysis, thus indicating infection. The diagnostic fidget spinner identified the two false negatives from the culture test as being infected samples. On the basis of recommendations given by the local medical doctors, the authors estimated that over half of patients suspected to have urinary tract infection may be overtreated with antibiotics, and that 5% of the patients may be undertreated. This suggests that antibiotic misuse could be significantly reduced with faster and accurate point-of-care bacterial detection systems.
However, for hand-powered diagnostic devices to be useful on a larger scale, temperature control, reagent storage, easy-to-use user–device interfaces and the potential for contamination are essential elements and considerations that need to be optimized during system integration and implementation. And although time-consuming, a culture step with selective and differential media is an efficient enrichment strategy for the detection of bacteria down to single-cell resolution. Colony formation eliminates the influence of viable organisms that cannot be cultured, minimizes interference from sample components (known as ‘the matrix effect’), helps detect polymicrobial infections and standardizes the initial concentration for subsequent microbiological analysis. These advantages of culture-based methods are not readily available with centrifugation-based enrichment. Still, microfluidics, smartphone-based imaging, automated image analysis and single-cell manipulation techniques4,5 may improve the performance of point-of-care diagnostic systems so that they also work with blood and swab samples.
The rapid identification of pathogens is also beneficial during the optimization of an assay’s protocol, in the selection of a pathogen-specific or narrow-spectrum antimicrobial treatment (once the infection aetiology has been identified) and in antimicrobial susceptibility testing, when quantifying bacterial load for the standardization of the initial concentration and for the minimization of the inoculum effect (that is, an increase in the lowest inhibitory concentration of an antibiotic when increasing the amount of bacteria inoculated). Ultimately, the ideal point-of-care device for infectious disease diagnostics should perform comprehensive microbiological analyses that rapidly identify pathogens and determine their antibiotic-resistant profiles. In resource-limited settings, however, improving the speed and accessibility of diagnostic devices for infectious diseases would reduce misdiagnoses and the over-prescription of antibiotics.
Michael, I. et al. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-020-0557-2 (2020).
Kim, T.-H. et al. Anal. Chem. 89, 1155–1162 (2017).
Bhamla, M. S. et al. Nat. Biomed. Eng. 1, 0009 (2017).
Li, H. et al. Proc. Natl Acad. Sci. USA 116, 10270–10279 (2019).
Davenport, M. et al. Nat. Rev. Urol. 14, 296–310 (2017).
About this article
Cite this article
Li, H., Torab, P. & Wong, P.K. Detection of bacterial infection via a fidget spinner. Nat Biomed Eng 4, 577–578 (2020). https://doi.org/10.1038/s41551-020-0571-4