Remote Coacervate Robots with Visualized Bacteria Sensing and Removal in Indwelling Catheters for Infection Prevention

Catheter-associated urinary tract infections (CAUTIs) have been a common but challenging disease due to difficulties in the accurate prescribing of antibiotics, while existing clinical diagnostic methods are deficient in their quickness, simplicity, and convenience. Here we report a point-of-care sensing technique using untethered coacervate robots, which demonstrate high controllability and adaptation in confined catheters filled with viscous urine. Under different motion modes, the controlled coacervate robots can achieve a quick, visualized, and selective fluorescence response to Gram-negative bacteria from 102 CFU/mL to 107 CFU/mL in human urine as well as remove the colonized bacteria in the inner wall of catheters within 5 min. Blind tests on clinical urine samples with various bacterial infections reveal that the coacervate robots achieve 90% sensitivity and 100% specificity in identifying Gram-negative bacteriuria. The visualized sensing technique opens up a pathway for quick onsite diagnosis, which helps prevent infections associated with urinary catheters.