RESUMO
Little progress has been made in decreasing the incidence rate of salmonellosis in the US over the past decade. Mitigating the contribution of contaminated raw meat to the salmonellosis incidence rate requires rapid methods for quantifying Salmonella, so that highly contaminated products can be removed before entering the food chain. Here we evaluated the use of Time-to-Positivity (TTP) as a rapid, semi-quantitative approach for estimating Salmonella contamination levels in ground beef. Growth rates of 14 Salmonella strains (inoculated at log 1 to -2 CFU/g) were characterized in lean ground beef mTSB enrichments and time-to-detection was determined using culture and molecular detection methods. Enrichments were sampled at five timepoints and results were used to construct a prediction model of estimated contamination level by TTP (superscript indicates time in hours) defined as TTP4: ≥5 CFU/g; TTP6: ≤5, ≥1 CFU/g; TTP8: ≤1, ≥0.01 CFU/g; with samples negative at 8 h estimated ≤0.01 CFU/g. Model performance measures showed high sensitivity (100%) and specificity (83% and 93% for two detection methods) for samples with a TTP4, with false negative rates of 0%.