Establishment of time and temperature limits to control the levels of Escherichia coli (Migula) Castellani and Chalmers in smoked salmon during cold meal preparation

The population kinetics of Escherichia coli in smoked salmon was determined at various temperatures typically used in cold meal preparation. Growth parameters such as the maximum specific growth rate and lag time were estimated using the Baronyi model. The time-temperature for E. coli, total coliforms and total aerobic plate counts at each stage of the process was profiled and used as the basis for simulating the process conditions. Time and temperature limits for the present process producing products with E. coli levels conforming to the specifications adopted by an inflight catering company in the Philippines were used as bases for establishing the time and temperature limits for the new process called "Non-Stop". The prevalence of E. coli in smoked salmon was mainly found at the surface of the meat. Growth was observed at 5, 10, 25 and 35 deg C. The Baronyi model was used to estimate growth parameters because of its sensitivity particularly at the transition from the lag and log-phases of bacterial growth. Highest maximum specific growth rate (1.16/hr) for E. coli in smoked salmon was observed at 35 deg C. Growth rates decreased as temperatures decreased while log times increased with low temperatures. A slight decline of cell populations was observed under freezing conditions. An exponential model was used to describe the reduction rate of the cell numbers. An average reduction rate of 0.001/hr was observed for E. coli during freezing. Using a speadsheet model, Monte Carlo simulation was performed using parameters obtained from the profiles of the smoked salmon and the cold meal preparation process. Accuracy and bias factors for the model were computed to be 0.967 and 1.03 for E. coli, 0.89 and 1.22 for total coliforms, and 0.502 and 1.99 for aerobic plate counts, respectively. All, except the aerobic plate counts (APC) model, had accuracy and bias factors within limits defined for determining the applicability of a model. Contamination rate during the processing was considered to be the factor to address the gap for the APC model. A simulation of the "non-stop" process was performed using time and temperature limits similar to the "conveyor belt" process (freezer storage at less than or equal to -12 deg C; thawing at less than or equal to 10 deg C for not more than 48 hours, preparation/tray set-up at less than or equal to 15 deg C for not more than 90 minutes; and, final holding at less than or equal to 5 deg C for not more than 24 hours). Results show that the finished product will have an E. coli level in the range of -1.457 to 2.628 log cfu/g, with a mean of 0.370 plus or minus 0.02 log cfu/g. The E. coli concentration in the finished product was found to consistently conform to the standard of 10 cfu/g