A sensor‐based system for rapid on‐site testing of microbial contamination in meat samples and carcasses

• Published in: Journal of applied microbiology 2022-02, Vol.132 (2), p.1210-1220
• Main Authors: Santovito, Elisa, Elisseeva, Sophia, Smyth, Conor, Cruz‐Romero, Malco, Kerry, Joseph P., Duffy, Geraldine, Papkovsky, Dmitri B.
• Format: Article
• Language: English
• Subjects: Animals; Automation; Calibration; carcass swabs; Carcasses; Cattle; Colony Count, Microbial; Correlation coefficient; Correlation coefficients; Food contamination; Food Contamination - analysis; Food Microbiology; Food safety; Meat; Meat industry; Meat processing industry; Meat Products; Microbial contamination; Microorganisms; optical oxygen sensing; Oxygen probes; Phosphorescence; Product safety; rapid microbial testing; raw meat spoilage; respirometric assays; Sample preparation; Sensors; total viable counts
• ISSN: 1364-5072; 1365-2672
• DOI: 10.1111/jam.15274

Aims To develop an oxygen sensor‐based method for testing total aerobic viable counts (TVC) in raw meat samples and cattle carcass swabs, which is rapid, simple, affordable, provides good sensitivity and analytical performance and allows on‐site use. Methods and Results The test uses the same sample preparation procedure as the established plate counting TVC method for meat samples and carcasses, ISO4833‐1:2013. After this liquid samples are transferred into standard 25‐ml vials with built‐in phosphorescent O2 sensors and incubated on a block heater with hourly readings of sensor signals with a handheld reader, to determine signal threshold time (TT, hours) for each sample. The method is demonstrated with the quantification of TVC in industrial cuts of raw beef meat (CFU per g) and carcass swabs (CFU per cm2). Calibration curves were generated, which give the following analytical equations for calculating the TVC load in unknown samples from measured TT values: TVC [Log(CFU per cm2)] = 7.83–0.73*TT(h) and TVC [Log(CFU per g)] = 8.74–0.70*TT(h) for the carcass swabs and meat samples respectively. The new tests show good correlation with the ISO methods, with correlation coefficients 0.85 and 0.83 respectively. The testing requires no dilutions, covers the ranges 2–7 Log(CFU per g) for the meat samples and 1–7 Log(CFU per cm2) for carcass swabs, and has time to result 1–10 h with faster detection of more contaminated samples. Conclusions The sensor‐based testing demonstrates simplicity, high speed, sample throughput and automation. It can provide a straightforward replacement for the conventional TVC tests, which are time consuming, laborious and have time to result of 48–72 h. Significance and Impact of the Study The method(s) can be adopted by the meat industry and research labs, and used to improve microbial quality and safety of meat products and processes.