Inactivation of Escherichia coli O157:H7 in beef roasts cooked under selected cooking conditions was evaluated. Eye of round roasts were each inoculated at five sites in the central plane with a five-strain cocktail of E. coli O157:H7 at ca. 6.3 log CFU per site and cooked to center temperatures of 56 to 71°C in a convection oven set at 120, 140, 180, or 200°C, in a conventional oven set at 120 or 210°C, and in a slow cooker set on high or low. Prime rib roasts were each inoculated at 10 sites throughout the roast with the same E. coli O157:H7 cocktail at ca. 6.6 log CFU per site and cooked in the conventional oven set at 140 or 180°C to center temperatures of 58 to 71°C. The number of sites yielding E. coli O157:H7 after cooking decreased with increasing roast center temperature for the eye of round roasts cooked in the convection oven or in the slow cooker at a given setting, but this trend was not apparent for roasts of either type cooked in the conventional oven. Reductions of E. coli O157 in both types of roasts were generally less at the center than at other locations, particularly locations closer to the surface of the meat. When eye of round roasts were cooked to the same center temperature in the convection oven, the reduction of E. coli O157:H7 increased with increasing oven temperature up to 180°C and decreased after that. The reduction of E. coli O157:H7 in replicate roasts cooked under conditions in which the organism was not eliminated during cooking mostly differed by >1 log CFU per site. However, E. coli O157:H7 was not recovered from any of the inoculation sites when eye of round roasts were cooked to 65, 60, 60, or 63°C in the convection oven set at 120, 140, 180, and 200°C, respectively; cooked to 63 or 71°C in the conventional oven set at 120 and 210°C, respectively; or cooked to 63°C in the slow cooker set at high or low. For prime rib roasts, E. coli O157:H7 was not recovered from any of the inoculation sites in roasts cooked to 71 or 58°C in the conventional oven set at 140 and 180°C, respectively.

Consumption of undercooked intact beef cuts with the surfaces sufficiently cooked poses no microbiological risks because such tissue from healthy animals is generally sterile (14). However, beef roasts on retail sale in North America have often been mechanically tenderized by incision with banks of thin blades (23). Such mechanical tenderization can carry bacteria from the surface into the previously sterile deep tissues (20). The internalized bacteria can include enteric pathogens such as Salmonella and Escherichia coli O157:H7, and consumers can be infected if these bacteria are not sufficiently inactivated during cooking (10, 12). North American regulatory authorities have determined that mechanically tenderized beef should be identified and labeled as such, and product-specific instructions for cooking the meat should be provided (18, 30, 31). The U.S. Department of Agriculture, Food Safety Inspection Service (USDA-FSIS) (30) criterion for safe cooking of mechanically tenderized beef is that the temperature history at any point in an item of meat during cooking must result in ≥5-log reductions of E. coli O157:H7. During cooking, the coldest point in a roast usually will be the geometric center of the cut (25). Therefore, the USDA-FSIS (30), based on data from the thermal inactivation of Salmonella in ground beef (17), has mandated that labels on mechanically tenderized beef carry recommendations to cook the meat to temperatures as measured at the center of 160°F (71.1°C) or of 145°F (62.8°C) with a holding time of 3 min before consumption. Whether consumer adherence to these recommendations during cooking of beef would be sufficient to achieve a ≥5-log reduction of E. coli O157:H7 at all points in roasts is uncertain. Available reports on inactivation of E. coli O157:H7 or Salmonella during cooking of mechanically tenderized or brine injected beef roasts are few, and most studies have addressed cooking of large beef roasts under commercial conditions (27, 29, 32, 33). In commercial practice, beef roasts are commonly cooked in ovens operated at low temperatures, i.e., ≤250°F (121.1°C), and the roasts then may be held for prolonged periods after cooking in warming ovens set at 60°C (27, 29, 32). Recent simulations of commercial cooking of beef prime rib roasts to internal temperatures up to 71°C did not reliably eliminate ≥5 log CFU of E. coli O157:H7 (27).

Culinary instructions for cooking beef roasts under domestic circumstances differ widely, and most are for cooking beef roasts in a conventional oven. The specified temperature settings are 140 to 220°C (5, 6). Domestic ovens fitted with a convection function are widely available, and instructions for cooking beef roasts in a convection oven are sometimes appended to instructions for conventional oven cooking. The general instructions for convection oven cooking are that either the specified cooking time (i.e., minutes per kilogram or pound) be reduced by 25 to 30% while retaining the same cooking temperature or the specified cooking time is retained with the set oven temperature reduced by at least 15°C (4, 11). Slow cookers are mostly used to cook complete dishes, which can include beef roasts, for periods of several hours (2, 3). Beef roasts cooked in this manner will be well done. There are, however, instructions for cooking beef roasts to a rare condition in a slow cooker operated on a high or low setting for less than 2 h (1, 34). Information appears to be lacking on how oven temperature or the use of conventional or convection ovens or slow cookers for cooking affects the survival of E. coli O157:H7 in beef roasts cooked to the same target temperature measured at the center of the meat.

Studies were carried out (i) to determine the effect of oven temperature on the survival of E. coli O157:H7 in beef roast by cooking eye of round roasts in a convection oven operated at different temperatures and (ii) to determine the effect of cooking appliances by comparing the survival of the organism in eye of round roasts cooked at selected temperatures in convection and conventional ovens and in a slow cooker. Roasts used for cooking had been inoculated with E. coli O157:H7 at various points in the deep tissues to simulate the internalization of the organism resulting from mechanical tenderization. The effect of fat and bone on the survival of E. coli O157:H7 also was determined using prime rib roasts inoculated with this organism.

Preparation of inocula.

Five strains of nontoxigenic E. coli O157:H7 from the culture collection of the Department of Food Science at the University of Manitoba (Winnipeg, Manitoba, Canada) were used to prepare suspensions of stationary-phase cells in half-strength brain heart infusion (Difco, BD, Sparks, MD) as previously described (15). Each cocktail suspension contained E. coli O157:H7 at 8.54 ± 0.07 log CFU/ml and the food dye FD&C blue #1 (Calico Food Ingredients, Kingston, Ontario, Canada) at about 200 ppm (16, 19).

Inoculation of roasts.

Vacuum-packaged eye of round (semitendinosus) primal cuts and prime rib roasts were obtained from a beef packing plant and a local retail store, respectively. The prime rib roasts (2.10 ± 0.14 kg) and cylindrical eye of round roasts (7 cm, 0.61 ± 0.06 kg) prepared from the primal cuts were stored at 1 ± 1°C and used for cooking within 2 weeks. Roasts were inoculated in groups of four. Each eye of round roast was inoculated at five points: the geometric center and points 1 and 3 cm from each edge of the roast along the center line. The two points 3 cm from each edge were along the center line approximately 0.5 cm to either side the geometric center. Each prime rib roast was inoculated at 10 points: the geometric center of the rib eye muscle and in the central plane 2 cm from the edge of the rib eye muscle, four points in the approximately 1-cm-thick layer of fat covering the rib eye muscle surface, and four points in the muscle tissue within 2 mm of bone. At each point to be inoculated, the tissue was disrupted as previously described (15) but with a sterile needle (18 gauge by 1.5 in. [3.8 cm]; BD, Mississauga, Ontario, Canada). The disrupted tissue was inoculated using a pipette fitted with a sterile needle with 10 μl of the five-strain cocktail suspension of E. coli O157:H7. The inoculated roasts were held at 4°C overnight until used for cooking.

Cooking of roasts.

The benchtop conventional oven (model KCO 274 SSO, KitchenAid, St. Joseph, MI) operated without convection (bake mode setting) and the benchtop convection oven (model 241B toaster oven, Bravetti, Mississauga, Ontario, Canada) used in this study were of similar sizes (width by depth by height approximately 330 by 300 by 200 mm and 350 by 310 by 230 mm, respectively). The doors of both ovens were hinged at the bottom. The temperatures of roasts and ovens were monitored using stainless steel probes that carried type K thermocouple sensors (Thermoworks, Lindon, UT). Each probe was connected through a thermocouple thermometer (Dualog R, Cole-Parmer Instruments, Vernon Hills, IL) to a computer with which temperatures were recorded at 10-s intervals. For roast cooking in the ovens, a probe was placed within the oven with the probe held in place by closing the door on the armored cord of the probe. The oven was turned on, and the temperature setting was adjusted as required to obtain the appropriate temperature as measured by the probe. A second probe was inserted through one end of a roast with the tip at the approximate geometric center of the roast. For roast cooking in a slow cooker (4-liter capacity; model SCVP500B-CN, Crock Pot, Boca Raton, FL), the handle in the center of the cooker lid was removed. The roast was placed in the vessel with the surface penetrated for inoculation facing up, and water was added until the vessel was filled to the 2-liter mark, at which point all but the top surface of the roast was immersed in water. The lid was put in place, and the probe was threaded through the 8-mm-diameter hole in the lid and inserted into the roast.

The effect of oven temperature on the survival of E. coli O157:H7 during cooking was determined by cooking eye of round roasts in the convection oven operated at 120, 140, 180, and 200°C. Although the maximum temperature on the temperature dial of the convection oven is 450°F (232°C), the maximum temperature attainable was only 200°C. To compare the effect of cooking appliances on the survival of E. coli O157:H7 during cooking, eye of round roasts were also cooked in a conventional oven operated at 120 and 210°C and in a slow cooker operated on high and low settings. The target center temperature was initially set at 60, 56, 56, or 58°C for oven temperatures of 120, 140, 180, and 200 or 210°C, respectively, and at 60°C for both slow cooker settings. The target center temperature for each appliance temperature setting was subsequently raised until no E. coli O157:H7 was recovered from any of the inoculation sites or to 60°C if elimination of E. coli O157:H7 was attained at center temperatures of <60°C. Prime rib roasts were cooked in the conventional oven operated at 140 and 180°C to an initial target center temperature of 58°C for determination of the effect of fat and bone on the survival of E. coli O157:H7 during cooking. As with eye of round roasts, the target center temperature for prime rib roasts cooked at either temperature was subsequently raised until no E. coli O157:H7 was recovered from any of the inoculation sites or to 60°C if elimination of E. coli O157:H7 was attained at center temperatures of <60°C. Thus, in this study, eye of round roasts were cooked in the convection oven at 120, 140, 180, and 200°C to center temperatures of 60, 63, and 65°C; 56, 58, and 60°C; 56, 58, and 60°C; and 58, 60, and 63°C, respectively. Eye of round roasts were cooked in the conventional oven at 120 and 210°C to center temperatures of 58, 60, and 63°C; and 58, 60, 63, and 71°C, respectively. Eye of round roasts were cooked in the slow cooker at both high and low settings to center temperatures of 60 and 63°C. Prime rib roasts were cooked in the conventional oven at 140 and 180°C to center temperatures of 58, 60, 63, and 71°C; and 58 and 60°C, respectively.

Each combination of target roast center temperature and oven or slow cooker setting was regarded as one cooking condition. Thus, in total, eye of round roasts and prime rib roasts were subjected to 23 and 6 cooking conditions, respectively. For each cooking condition, three trials with one inoculated roast each were conducted. The fourth inoculated roast was not cooked and was used as a control for determination of the initial level of E. coli O157:H7 at inoculation sites. A total of 92 eye of round roasts and 24 prime rib roasts were used in this study.

Excision of inoculated tissue and enumeration of E. coli O157:H7.

For excision of inoculated tissue from each raw or cooked roast, the surface penetrated for inoculation of the roast was seared to a depth of about 3 mm with a heated stainless steel plate. A block of muscle or fat tissue centered on each point of inoculation was cut from the roast aseptically. About 1 g of muscle or fat tissue, which included all the dyed tissue, was then excised from each block. Each portion of excised tissue was placed in a stomacher bag fitted with a filter sleeve and pummeled with 10 ml of 0.1% peptone water (Difco, BD) in a stomacher for 2 min. The stomacher fluid was withdrawn from the filter sleeve, and 1 ml was used to prepare 10-fold dilutions in 0.1% peptone water. Each dilution and all of the undiluted stomacher fluid was filtered through a hydrophobic grid membrane filter (HGMF; Oxoid, Nepean, Ontario, Canada). Each filter was placed on a plate of lactose monensin glucuronate agar (Oxoid), which is formulated for the recovery of injured cells (13). Plates were incubated at 35°C for 18 to 24 h. The grid squares containing blue colonies were counted for each suitable filter, with the assumption that all such colonies were E. coli O157:H7. A most probable number (MPN) was calculated from each count of squares using the formula MPN =N ×ln(N/NX), where N is the total number of squares on a filter and X is the count of squares containing blue colonies. The detection limit of the HGMF method is 1 CFU per inoculation site.

Analysis of microbiological data and cooking time.

Counts of E. coli O157:H7 recovered from inoculated sites were transformed to log values (log CFU per inoculated site). The time required to reach the target oven temperature after a roast was placed in the oven was the come-up time. The time required for cooking a roast to a target center temperature was the cooking time. Sets of cooking times were from groups of roasts that were cooked to the same final temperature at the geometric center under the same cooking conditions, and sets of come-up times were from groups of roasts that were cooked in the same oven operated at the same temperature. A Shapiro-Wilk test was used to test each set of cooking times or come-up times for a normal distribution. When data sets to be compared were normally distributed or mostly normally distributed, mean values for the sets were separated using a Tukey test. All tests were performed using Minitab, version 16 (Minitab Inc., State College, PA).

Initial E. coli O157:H7 populations.

For uncooked inoculated eye of round roasts, E. coli O157:H7 was recovered from the geometric center (n = 23) and from points in the central plane 1 cm (n = 46) and 3 cm (n = 46) from the edge of the roast at means (±SD) of 6.31 ± 0.30, 6.33 ± 0.23, and 6.28 ± 0.26 log CFU per site, respectively. For uncooked inoculated prime rib roasts, the E. coli O157:H7 was recovered from the center of the rib eye muscle (n = 6), points in the central plane 2 cm from the outer edge of that muscle (n =6), muscle tissue within 2 mm of the bone (n = 24), and the fat tissue covering the muscle surface (n = 24) at 6.58 ± 0.18, 6.57 ± 0.21, 6.57 ± 0.20, and 6.64 ± 0.37 log CFU per site, respectively.

Cooking in a convection oven.

All sets of come-up times for the convection oven were normally distributed (P > 0.05), with means (±SD) of 3.67 ± 0.69, 4.15 ± 1.29, 4.44 ± 0.41, and 11.33 ± 2.79 min for oven temperatures of 120, 140, 180, and 200°C, respectively. The come-up times ranged from 2.83 to 5.66 min and 5.83 to 15.83 min for oven temperatures of ≤180 and 200°C, respectively. The mean come-up time for 200°C was significantly longer (P < 0.05) than that for the lower temperatures. During cooking, the convection oven temperatures were 120.6 ± 1.1, 140.4 ± 1.6, 179.9 ± 1.3, and 199.7 ± 2.0°C at set temperatures of 120, 140, 180, and 200°C, respectively.

At each oven temperature, the mean cooking time generally increased with increasing target center temperature (Table 1), but the difference was mostly <5 min and was not significant (P > 0.05). When roasts were cooked to 60°C (the temperature to which roasts were cooked at all four oven temperatures), the mean cooking time was significantly longer (P < 0.05) at 120°C than at 140, 180, and 200°C, and there was no significant difference (P > 0.05) between cooking times for roasts cooked at 140 and 180°C or at 180 and 200°C (Table 1). At each oven temperature, the number of inoculated sites from which E. coli O157:H7 was recovered after cooking decreased with increasing target temperature at roast centers: 10, 3, and 0 sites; 9, 7, and 0 sites; 2, 1, and 0 sites; and 3, 2 and 0 sites for roasts cooked at 120°C to 60, 63, and 65°C; at 140°C to 56, 58, and 60°C; at 180°C to 56, 58, and 60°C; and at 200°C to 58, 60, and 63°C, respectively. E. coli O157:H7 was recovered from all three types of inoculation sites (geometric center and 1 and 3 cm from edges of the roast) when roasts were cooked at 120°C to 60 and 63°C and at 140°C to 56 and 58°C. For cooking conditions under which E. coli O157:H7 was not eliminated, reductions of the organism was always greater at points close to the roast surface (points 1 cm from each edge) than points close to or at the roast center (centers and points 3 cm from each edge). The reductions of E. coli O157:H7, in relation to roast center temperatures and oven temperatures, decreased in the following order: 56°C at center, 180°C > 140°C; 58°C at center, 180°C > 200°C > 140°C; 60°C at center, 180°C = 140°C > 200°C >120°C; and 63°C at center, 200°C > 120°C. The extent to which E. coli O157:H7 was inactivated at the same type of inoculation site in roasts that were cooked under the same condition and yielded more than one positive sample for that particular type of inoculation site mostly differed by >1 log CFU per site. Such variation between roasts that were cooked at 120°C to an internal temperature of 60°C, at 140°C to internal temperatures of 56 or 58°C, or at 200°C to an internal temperature of 58°C was 3 to 5 log CFU per site. Even so, E. coli O157:H7 was reduced by >5 log CFU per site throughout the meat when roasts were cooked at 180°C to an internal temperature of 58°C, and no E. coli O157:H7 was recovered from any of the inoculation sites when roasts were cooked at 120, 140, 180, and 200°C to internal temperatures of 65, 60, 60, and 63°C, respectively.

TABLE 1.

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a convection oven

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a convection oven
Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a convection oven

Cooking in a conventional oven.

When eye of round roasts were cooked in the conventional oven operated at 120 and 210°C, the come-up times for the 120°C setting were normally distributed (P > 0.05), but those for 210°C were not normally distributed (P < 0.05). The mean come-up times were 8.74 ± 2.08 and 7.54 ± 2.11 min for oven temperatures of 120 and 210°C, respectively. During cooking, the mean oven temperatures at set temperatures of 120 and 210°C were 120.1 ± 0.7 and 210.1 ± 1.3°C, respectively. When eye of round roasts were cooked in the conventional oven to a center temperature of 58, 60, and 63°C, the mean cooking time was significantly shorter (P < 0.05) at 210°C than at 120°C (Table 2). At either oven temperature, the mean cooking time increased with increasing target temperature at roast centers; however, those differences were not significant (P > 0.05). The mean times for cooking roasts in the conventional oven at 120°C to center temperatures of 60 and 63°C were 90.3 and 105.0 min, respectively. These times were 2.3 min (3%) and 13.3 min (13%) longer, respectively, than those for cooking roasts in the convection oven operated 120°C to the same center temperatures. When eye of round roasts were cooked at 120 and 210°C, the number of samples yielding E. coli O157:H7 after cooking did not always decrease with increasing target roast center temperature, as did such roasts cooked in the convection oven. Reduction of E. coli O157:H7 increased with increasing final roast center temperature when roasts were cooked at 120°C to 58 to 63°C, but the reduction in roasts cooked at 210°C to those center temperatures was variable. For center temperatures of 58, 60, and 63°C, reduction of E. coli O157:H7 was greater in roasts cooked at 120°C than in those cooked at 210°C. As for eye of round roasts cooked in the convection oven, greater reduction of E. coli O157:H7 was always attained at points close to the surface than at the center of roast, and reduction of the organism between roasts cooked under the same condition differed by >1 log CFU per site. E. coli O157:H7 at all inoculated points was eliminated when roasts were cooked to 63 or 71°C at 120 and 210°C, respectively.

TABLE 2.

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a conventional oven

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a conventional oven
Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in a conventional oven

Cooking in a slow cooker.

When eye of round roasts were cooked in the slow cooker to a center temperature of 60 and 63°C, the mean cooking time was significantly shorter (P < 0.05) for the high setting than for the low setting (Table 3). However, at either cooker setting, the mean times for cooking roasts to 60 and 63°C were not significantly different (P > 0.05). E. coli O157:H7 was recovered from only roast centers at >2.8 log CFU per site when roasts were cooked to 60°C, but no E. coli O157:H7 was recovered from any of the inoculated sites when roasts were cooked to 63°C, regardless of the temperature setting of the cooker.

TABLE 3.

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in water in a slow cooker

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in water in a slow cooker
Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three eye of round roasts cooked in water in a slow cooker

Effect of fat and bone on the survival of E. coli O157:H7.

When prime rib roasts were cooked in the conventional oven at 140 and 180°C, the come-up times were normally distributed, with mean values of 14.83 ± 8.06 and 11.47 ± 4.29 min, respectively. At both oven temperatures, the mean cooking time did not always increase with increasing target temperature at roast center (Table 4). When prime rib roasts were cooked at 140°C to center temperatures of 58, 60, and 63°C, E. coli O157:H7 was recovered from 6, 2, and 4 of the 30 inoculation sites, respectively, and the reduction of E. coli O157:H7 at roast centers increased with increasing target center temperature, with 2 and 1 site from roasts cooked to 58 and 60°C yielding E. coli O157:H7 at >5 log CFU per site and all sites from centers of roasts cooked to 63°C yielding ≤2.39 log CFU per site. E. coli O157:H7 was recovered at 0.60 log CFU per site from one site 2 cm from the edge of the ribeye muscle in roasts cooked at 63°C, but no E. coli O157:H7 was recovered from such sites in roasts cooked to 58 and 60°C. The organism was recovered from muscle tissue adjacent to the bone from roasts cooked to 58°C only. E. coli O157:H7 were reduced by 5.52 to 6.60 log CFU per site in the fat tissue covering the muscle surface of prime rib roasts cooked under the conditions examined. No E. coli O157:H7 was recovered from any of the inoculated sites when prime rib roasts were cooked at 140°C to 71°C or at 180°C to 58 and 60°C.

TABLE 4.

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three prime rib roasts cooked in a conventional oven

Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three prime rib roasts cooked in a conventional oven
Cooking times and reductions in Escherichia coli O157:H7 at inoculation sites in groups of three prime rib roasts cooked in a conventional oven

The temperature setting dials and electronic displays of modern domestic ovens suggest that the temperature set for an oven will be consistently achieved and maintained. However, actual oven temperatures can vary by as much as 32°C from set temperatures (22, 26). It is then not surprising that the maximum temperature on the dial of the convection oven used in this study was 30°C higher than the attainable maximum temperature. Even so, after the set temperatures had been attained, the mean oven temperatures during cooking varied little, with standard deviations of ≤3.3°C. This finding of the stability of oven temperature during cooking is in agreement with previous reports with respect to oven temperature profiles (8, 27). Ovens used in this laboratory research were by necessity much smaller than regular household ovens, which limited the size of the roasts that could be evaluated. Some further work can be conducted to verify whether findings from those ovens are applicable to regular household ovens and larger roasts.

Direct comparison with published work on the microbiological effects of cooking small roasts under domestic conditions is difficult because such information is largely lacking. However, some studies have been conducted on thermal inactivation of Salmonella (7, 9, 29, 32) or E. coli O157:H7 (27) in large beef roasts cooked under commercial conditions. Snyder et al. (29) reported that 6 h of cooking in a controlled vapor oven set at 54.4°C with an oven relative humidity of 30 to 90% resulted in temperatures of 48.9 to 53.3°C at the geometric centers of three beef roasts (5.9 to 6.4 kg). Under such conditions, Salmonella which was introduced through inserted dialysis tubing, at the geometric center of roasts was reduced by 4.56 to 7.65 log CFU/g. Salmonella in beef roasts (4 to 6 kg) that had been mechanically tenderized was reduced by 3.58 and 3.87 log CFU/g when roasts were cooked in a convection oven at 120°C to an internal temperature of 48.9°C and in a conventional oven at 190°C to internal temperatures of 43.3 or 48.9°C, respectively (9, 32). Blankenship (7) studied pathogen inactivation in 21 beef roasts (4.9 to 5.7 kg) that had been contaminated by needle injection of 2 × 107 Salmonella cells along the longitudinal geometric center line. The roasts were cooked in a pilot oven at 110°C to 13 internal temperatures of 58.3 to 64.1°C. Salmonella was recovered from the centers of 2 of the 21 roasts (one cooked to 58.3°C and one cooked to 60.8°C), but all 21 cooked roasts were positive for Salmonella either in the drip or the net in which the roast was held during cooking. When mechanically tenderized roasts (8 to 9 kg) were cooked in a commercial oven at 121°C to internal temperatures of 37.8, 48.9, 60, or 71.1°C, 2.7- to 4.0-log reductions of E. coli O157:H7 were found, and no differences in reduction of the organism were observed among the roasts cooked to 48.9, 60, or 71.1°C (27).

Despite the differences in types of ovens and oven temperatures used in those studies, >5-log reductions of either Salmonella or E. coli O157:H7 were not consistently achieved by cooking roasts under any of those conditions alone. In the present study, the large variation in reductions of E. coli O157:H7 in roasts that had been cooked under the same conditions were in agreement with the findings in these previous reports. The rate at which any point within a roast heats and the homogeneity of heating will be affected by both the conditions to which the roast is exposed during cooking and factors intrinsic to the roast. Inconsistent reductions of internalized bacteria in large roasts have been attributed to nonhomogeneous heating within the meat (27). The discrepancy in reductions also may be caused by other factors, such as variations in the distribution of bacterial cells in the meat resulting from different inoculation methods. Cooking large roasts under commercial conditions often includes a searing step before cooking and a holding step in a warming oven after cooking, and the combined lethality from searing, cooking, and holding could result in >5-log reductions of Salmonella or E. coli O157:H7 (27, 29, 32). In the present study, when roasts were cooked in convection or conventional ovens operated at 180°C to an internal temperature of 58°C or in a slow cooker on either setting to an internal temperature of 63°C, >5-log reductions of E. coli O157:H7 were consistently attained. Breslin et al. (8) reported that in beef roasts (sizes similar to those of the eye of round roasts used in the present study) that had been inoculated with Salmonella by vacuum tumbling and cooked in a commercial moist-air convection oven at 91.1°C to a final temperature of 71.1°C, Salmonella was consistently reduced by >5.6 log CFU. The findings of the present study and those of Breslin et al. indicate that cooking of small roasts to instantaneous internal endpoint temperatures of ≤71°C can consistently result in >5-log reductions of internalized E. coli O157:H7 or Salmonella.

The variation in the thermal conductivity of muscle, bone, and fat (21) will inevitably affect the survival of bacteria in or adjacent to these tissues during heating of the meat. The survival of E. coli O157:H7 in the fat tissue and muscle tissue adjacent to the bone was no greater than that at the geometric center of prime rib roasts cooked under the conditions of this study. This finding suggests that conditions allowing effective reduction of E. coli O157:H7 at the roast center should be at least equally effective for reducing the organism associated with the fat and bone of prime rib roasts.

In this study, the reduction of E. coli O157:H7 in eye of round roasts cooked to the same center temperature increased with increasing oven temperatures of 120 to 180°C. A possible reason for this increasing reduction could be that the velocity of air in the oven increased with increasing oven temperature (28), which resulted in a more efficient replacement of the cold air surrounding the meat by the warm oven air and, in turn, a more homogeneous heating of the meat. This hypothesis was supported by the finding that cooking prime roasts in the conventional oven at 180°C was more effective for reducing the numbers of internalized E. coli O157:H7 than cooking at 140°C when the roasts were cooked to an internal temperature of 58 or 60°C. However, the benefit of a more homogeneous heating of the meat resulting from a better replacement of cold air at higher temperatures will be offset by the increased partial drying or browning of the meat surface at those temperatures, because thermal conductivity decreases with decreasing water content on the meat surface (24).

Cooking eye of round roasts to 60°C in the convection oven operated at 140 and 180°C eliminated E. coli O157:H7 at all inoculation sites, but reductions resulting from cooking such roasts in the slow cooker to 60°C were <1.30 to 6.30 log CFU per site, which indicates that when roasts are cooked to the same final center temperature cooking in a convection oven is more effective than cooking in a slow cooker for reducing E. coli O157:H7. When eye of round roasts were cooked at 120°C to internal temperatures of 60 and 63°C, greater reductions of E. coli O157:H7 were attained with the conventional oven than with the convection oven. The reason for this is not clear because one would expect the convection function to lead to more efficient heating of the meat. Even though the cooking times for the conventional oven were 2.3 min (3%) and 13.3 min (13%) longer than those for the convection oven when eye of round roasts were cooked at 120°C to 60 or 63°C, respectively, the difference in cooking times between the two ovens was no more than that between roasts cooked under the same conditions. These findings indicate that for convection oven cooking, reducing the cooking time by 25 to 30% or the oven temperature by at least 15°C from the times and temperatures listed in the culinary instructions for conventional oven cooking may result in undercooked roasts (4, 11).

Taken together, the findings of this study indicate that the survival of E. coli O157:H7 in roasts cooked to the same center temperature differs with types and settings of the cooking appliance. Even so, a 5-log reduction of E. coli O157:H7 throughout the roast is attainable when eye of round roasts are cooked at any setting in a slow cooker to 63°C and when eye of round roasts and prime rib roasts are cooked in convection and conventional ovens, respectively, at 180°C to and internal temperature of 58°C. Thus, adherence to the USDA-FSIS (30) recommendation of cooking roasts to a center temperature of 71.1 or 63°C with a holding time of 3 min at oven or slow cooker temperature settings appropriate to roast sizes should ensure the microbiological safety of such products.

This study was carried out as part of the E. coli O157:H7 Research and Education Strategy of the Canadian Cattlemen's Association (CCA), with funding provided by the Strategies Initiatives Program of the Alberta Livestock and Meat Agency. The contribution of Mr. M. Klassen (CCA) to development and management of the study is greatly appreciated.

1.
Andrews
,
A. J.
2014
.
How to cook a medium rare roast beef in a crock-pot
. .
2.
Anonymous
.
1999
.
Basic beef pot roast
. .
3.
Anonymous
.
2015
.
All American roast beef
.
Available at: http://allrecipes.com/recipe/all-american-roast-beef/. Accessed 5 December 2014
.
4.
Anonymous
.
2015
.
Cooking with convection oven
. .
5.
Anonymous
.
2015
.
Oven roasting
.
Available at: http://www.beefinfo.org/?ID=17&ArticleID=37. Accessed 10 October 2014
.
6.
Anonymous
.
n.d.
How to roast beef
. .
7.
Blankenship
,
L.
1978
.
Survival of a Salmonella typhimurium experimental contaminant during cooking of beef roasts
.
Appl. Environ. Microbiol
.
35
:
1160
1165
.
8.
Breslin
,
T. J.
,
M. I.
Tenorio-Bernal
,
B. P.
Marks
,
A. M.
Booren
,
E. T.
Ryser
, and
N. O.
Hall
.
2014
.
Evaluation of Salmonella thermal inactivation model validity for slow cooking of whole-muscle meat roasts in a pilot-scale oven
.
J. Food Prot
.
77
:
1897
1903
.
9.
Calle-Madrid
,
M. A.
2012
.
Validation of thermal processing to control Salmonella spp. and Clostridium perfringens during prime rib preparation from intact and non-intact meat
.
M.S. thesis. University of Nebraska, Lincoln
.
10.
Catford
,
A.
,
M.-C.
Lavoie
,
B.
Smith
,
E.
Buenaventura
,
H.
Couture
,
A.
Fazi
, and
J. M.
Farber
.
2013
.
Findings of the health risk assessment of Escherichia coli O157 in mechanically tenderized beef products in Canada
.
Int. Food Risk Anal. J
.
3
(
3
):
1
12
.
doi:10.5772/56713
.
11.
Devries
,
D.
2014
.
How long to cook a roast in a convection oven
. .
12.
Echeverry
,
A.
,
J. C.
Brooks
,
M. F.
Miller
,
J. A.
Collins
,
G. H.
Loneragan
, and
M. M.
Brashears
.
2009
.
Validation of intervention strategies to control Escherichia coli O157:H7 and Salmonella Typhimurium DT 104 in mechanically tenderized and brine-enhanced beef
.
J. Food Prot
.
72
:
1616
1623
.
13.
Entis
,
P.
, and
P.
Boleszczuk
.
1990
.
Direct enumeration of coliforms and Escherichia coli by hydrophobic grid membrane filter in 24 hours using MUG
.
J. Food Prot
.
53
:
948
952
.
14.
Gill
,
C. O.
1979
.
Intrinsic bacteria in meat
.
J. Appl. Bacteriol
.
47
:
367
378
.
15.
Gill
,
C. O.
,
J.
Devos
,
M. K.
Youssef
, and
X.
Yang
.
2014
.
Effects of selected cooking procedures on the survival of Escherichia coli O157:H7 in inoculated steaks cooked on a hot plate or gas barbecue grill
.
J. Food Prot
.
77
:
919
926
.
16.
Gill
,
C. O.
,
L. F.
Moza
, and
S.
Barbut
.
2009
.
Survival of bacteria in less than thorough cooked, brine-injected steaks
.
Food Control
20
:
501
507
.
17.
Goodfellow
,
S. J.
, and
W. L.
Brown
.
1978
.
Fate of Salmonella inoculated into beef for cooking
.
J. Food Prot
.
41
:
598
605
.
18.
Health Canada
.
2014
.
Regulations amending the food and drug regulations (mechanically tenderized beef)
. .
19.
Hinkens
,
J. C.
,
N. G.
Faith
,
T. D.
Lorang
,
P.
Bailey
,
D.
Buege
,
C. W.
Kaspar
, and
J. B.
Luchansky
.
1996
.
Validation of pepperoni processes for control of Escherichia coli O157:H7
.
J. Food Prot
.
59
:
1260
1266
.
20.
Johns
,
D. F.
,
C. L.
Bratcher
,
C. R.
Kerth
, and
T.
McCaskey
.
2011
.
Translocation of surface-inoculated Escherichia coli into whole muscle nonintact beef striploins following blade tenderization
.
J. Food Prot
.
74
:
1334
1337
.
21.
Lawrie
,
R. A.
, and
D. A.
Ledward
.
2006
.
The storage and preservation of meat. I. Temperature control
,
p
.
189
234
.
In
R. A.
Lawrie
and
D. A.
Ledward
(
ed.
),
Lawrie's meat science
.
Woodhead Publishing
,
Boca Raton, FL
.
22.
Liddon
,
A.
2012
.
What your oven isn't telling you
. .
23.
Maddock
,
R.
2008
.
Mechanical tenderization of beef
. .
24.
Marcotte
,
M.
,
A. R.
Taherian
, and
Y.
Karimi
.
2008
.
Thermophysical properties of processed meat and poultry products
.
J. Food Eng
.
88
:
315
322
.
25.
Obuz
,
E.
,
T. H.
Powell
, and
M. E.
Dikeman
.
2002
.
Simulation of cooking cylindrical beef roasts
.
LWT Food Sci. Technol
.
35
:
637
644
.
27.
Porto-Fett
,
A. C. S.
,
B. A.
Shoyer
,
H.
Thippareddi
, and
J. B.
Luchansky
.
2013
.
Fate of Escherichia coli O157:H7 in mechanically tenderized beef prime rib following searing, cooking, and holding under commercial conditions
.
J. Food Prot
.
76
:
405
412
.
28.
Sato
,
H.
,
T.
Matsumura
, and
S.
Shibukawa
.
1987
.
Apparent heat transfer in a forced convection oven and properties of baked food
.
J. Food Sci
.
52
:
185
188
.
29.
Snyder
,
P. O. J.
,
J. E.
Kennedy
,
G. F.
Dagan
,
W. L.
Shelton
, and
M. J.
Gibson
.
2010
.
Controlled vapor oven cooking and holding procedures used for the reduction of Salmonella and prevention of growth of Clostridium perfringens in boneless beef rib roasts
.
Food Prot. Trends
30
:
223
229
.
30.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2013
.
Compliance guideline for validating cooking instructions for mechanically tenderized beef products
.
Available at: http://www.fsis.usda.gov/shared/PDF/Comp_Guide_MTB.pdf. Accessed 10 August 2014
.
31.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2013
.
Descriptive designation for needle or blade tenderized (mechanically tenderized) beef products
. .
32.
Wendelburg
,
J. W.
,
D. L.
Lambert
,
H.
Thippareddi
,
C. L.
Kastner
,
R. K.
Phebus
, and
J. L.
Marsden
.
2002
.
Salmonella spp. risk assessment for cooking of blade tenderized prime rib
,
p
.
52
53
.
In
Cattlemen's Day, Kansas State University
.
Agricultural Experiment Station and Cooperative Extension Service, Kansas State University
,
Manhattan
.
33.
Wiegand
,
K. M.
,
S. C.
Ingham
, and
B. H.
Ingham
.
2012
.
Evaluating lethality of beef roast cooking treatments against Escherichia coli O157:H7
.
J. Food Prot
.
75
:
48
61
.
34.
Younkin
,
M.
2013
.
Perfectly simple and sliceable crockpot roast beef
. .