U.S. Department of Agriculture
Food Safety and Inspection Service
Washington, DC 20250



 

HACCP Implementation: First Year Salmonella Test Results

January 26, 1998 to January 25, 1999

 

I. Background

The Food Safety and Inspection Service is the Agency within the U.S. Department of Agriculture responsible for ensuring the safety, wholesomeness, and accurate labeling of meat, poultry, and egg products. On July 25, 1996, FSIS issued its landmark rule, Pathogen Reduction; Hazard Analysis and Critical Control Point (HACCP) Systems.

The Pathogen Reduction and HACCP rule: (1) requires all meat and poultry plants to develop and implement a system of preventive controls, known as HACCP, to improve the safety of their products, (2) sets pathogen reduction performance standards for Salmonella that slaughter plants and plants producing raw ground products must meet, (3) requires all meat and poultry plants to develop and implement written standard operating procedures for sanitation, and (4) requires meat and poultry slaughter plants to conduct microbial testing for generic E. coli to verify the adequacy of their process controls for the prevention of fecal contamination. Implementation of the rule began on January 27, 1997. Large plants, which are those federally-inspected establishments employing 500 or more employees, became subject to the Salmonella testing requirements on January 26, 1998.

II. Sample Collection

In accordance with the Pathogen Reduction and HACCP rule, each federally-inspected establishment must meet product-specific Salmonella performance standards once they become subject to the Salmonella testing requirements in the rule. Salmonella samples are collected by federal inspectors and analyzed in one of three Agency laboratories.

To determine whether federally-inspected plants meet the Salmonella performance standards for products covered under the Pathogen Reduction and HACCP rule, the Agency has codified the maximum number of samples that can be positive for Salmonella in a statistically-based set of samples. The number of samples in a sample set varies by product, and represents the number of samples that FSIS collects and analyzes for Salmonella before determining whether the plant-specific product performance standard is met. These Salmonella performance standards are found at 9 CFR 310.25(b) and 381.94(b). Performance standards were calculated to provide an 80% probability of passing when the establishment is operating at the national baseline prevalence of positive Salmonella results.

Between January 26, 1998, and January 25, 1999, federal inspectors collected samples for Salmonella analysis from 204 large plants that produced products in one of seven categories: broilers, swine, steers and heifers, cows and bulls, ground beef, ground turkey, and ground chicken. The Agency conducted 10,023 Salmonella analyses on samples collected from large plants during the first year of HACCP pathogen reduction testing. These 10,023 total analyses are comprised of samples from products in the initial sample set as well as samples from additional (follow-up) sample sets. Of these 10,023 total analyses, 9,464 analyses were from these seven product categories up to and including one complete sample set. Results from 9,161 analyses in four of the seven product categories are the focus of this report.

III. Results Overview

Results for the first year of the large plant Salmonella testing program from January 26, 1998, through January 25, 1999, are available for four product classes: broilers, swine, ground beef, and ground turkey. During the first year of HACCP implementation, Salmonella samples were collected from 199 large federally-inspected establishments producing one of these four products. These four product classes are the only ones to date with complete sample sets from ten or more establishments.

Although results are considered preliminary in that they do not represent a random sample of all domestic meat and poultry production, Salmonella prevalence in large plants for each of these four products was lower after the first year of HACCP implementation than in baseline studies conducted before HACCP implementation (Table 1). Although it is unlikely that all of these reductions are solely attributable to the implementation of HACCP, the Agency nevertheless finds these results encouraging.

Results indicate that Salmonella, found on 20 percent of broiler carcasses in pre-HACCP baseline studies, was found on 10.9% of broiler carcasses after the first year of HACCP implementation. For swine, Salmonella was found on 8.7% of carcasses in pre-HACCP baseline studies, and 6.5% of carcasses after HACCP implementation. In ground beef, Salmonella was found in 7.5% of samples in pre-HACCP baseline studies, and 4.8% of samples after HACCP implementation. In ground turkey, Salmonella was found in 49.9% of samples in pre-HACCP baseline studies, and 36.4% of samples after HACCP implementation. For these four product classes combined, 100 (88%) of the 114 plants with complete data sets met their respective Salmonella performance standards (Table 2).

IV. Product-Specific Results

Broilers

Large federally-inspected establishments that slaughter broilers must meet the 20% broiler Salmonella performance standard found at 9 CFR 381.94(b). For broilers, this equates to a maximum of 12 Salmonella positive samples in a 51-sample set. Plants that exceed 12 positive Salmonella samples in the 51-sample set must initiate corrective action to meet the performance standard and undergo targeted Salmonella testing in accordance with 9 CFR 381.94(b)(3).

Between January 26, 1998 and January 25, 1999, the Agency collected and analyzed 5,697 carcass rinse samples for Salmonella from 124 large broiler plants. The Salmonella prevalence in the 5,697 broiler carcass rinse samples was 10.9% compared to 20% in pre-HACCP baseline studies (Table 1). Although comparisons between Salmonella prevalence during the first year of compliance testing and previous baseline studies must be done cautiously, the Agency finds these results encouraging.

In broilers, 51 samples comprise a complete sample set. Seventy-six (76) broiler plants had complete sample sets during this period (Table 2). Of the 48 broiler plants without complete sample sets, 26 had 45 or more samples collected by the end of the first year of HACCP implementation and are therefore within six working days of having complete sample sets. Nine additional broiler plants began having Salmonella samples collected in December 1998 or January 1999 and have not undergone sample collection for a sufficient time to collect a full sample set. These nine plants were identified during small plant identification activities in late 1998. Therefore, of the 124 broiler plants that had samples collected for Salmonella in the first year of HACCP implementation, 111 either have complete sample sets, are within 6 working days of completing their sample sets, or have only recently begun sample collection. Of the remaining 13 broiler plants without complete data sets, 11 had between 40 and 44 samples collected and analyzed by January 25, 1999 and are therefore within eleven working days of completing a sample set. The final two broiler plants had 33 samples collected and analyzed by January 25, 1999.

In the 76 broiler plants with complete sample sets, the plant-specific prevalence of Salmonella ranged from 0% to 47.1% (Table 3). Of these 76 plants, 69 (91%) met the broiler Salmonella prevalence performance standard (Table 2). Of these 69 plants, 47 (68%) had a Salmonella prevalence of less than 10% in complete sample sets, while 27 (39%) had a Salmonella prevalence of less than 4% in complete sample sets. Of the 7 broiler plants that did not meet the Salmonella performance standard, 5 had a Salmonella prevalence of less than 30%. Follow-up Salmonella testing in the seven broiler slaughter plants that did not meet the performance standard has begun.

Swine

Large federally-inspected establishments that slaughter swine must meet the 8.7% swine Salmonella performance standard found at 9 CFR 310.25(b). For swine, this equates to a maximum of 6 positive Salmonella samples in a 55-sample set. Plants that exceed 6 positive Salmonella samples in the 55-sample set must initiate corrective action to meet the performance standard and undergo targeted Salmonella testing in accordance with 9 CFR 310.25(b)(3).

Between January 26, 1998 and January 25, 1999, the Agency collected and analyzed 1,532 carcass sponge samples for Salmonella from 33 large swine plants. The Salmonella prevalence in the 1,532 swine carcass swabs was 6.5% compared to 8.7% in pre-HACCP baseline studies (Table 1). Although comparisons between Salmonella prevalence during the first year of compliance testing and previous baseline studies must be done cautiously, the Agency finds these results encouraging.

In swine, 55 samples comprise a complete sample set. Seventeen (17) swine plants had complete sample sets during this period (Table 2). Of the 16 swine plants without complete sample sets, 7 had 50 or more samples collected by the end of the first year of HACCP implementation and are therefore within five working days of having complete sample sets. Three additional swine plants began having Salmonella samples collected in January 1999 and have not had sample collection for a sufficient time to collect a full sample set. These three plants were identified during small plant identification activities in late 1998. Therefore, of the 33 swine plants that had samples collected for Salmonella in the first year of HACCP implementation, 27 either have complete sample sets, are within 5 working days of completing their sample sets, or have only recently begun sample collection. The remaining six swine plants without complete data sets range between 26 and 45 (n=2) samples collected and analyzed by January 25, 1999.

In the 17 swine plants with complete sample sets, the plant-specific prevalence of Salmonella ranged from 0% to 45.5% (Table 4). Of these 17 plants, 12 (71%) met the 8.7% swine Salmonella prevalence performance standard (Table 2). Each of these 12 plants had a Salmonella prevalence of less than 5% in complete sample sets. Of the 5 swine plants not meeting the Salmonella performance standard, 3 had a Salmonella prevalence between 11 and 15%. Follow-up Salmonella testing in two of the five swine slaughter plants that did not meet the performance standard has begun. The three swine slaughter plants that did not meet the performance standard but that have not begun follow-up testing have only recently completed (and failed) the first set. These three plants are being scheduled for follow-up testing.

Ground Beef

Large federally-inspected establishments that produce ground beef must meet the 7.5% ground beef Salmonella performance standard found at 9 CFR 310.25(b). For ground beef, this equates to a maximum of 5 positive Salmonella samples in a 53-sample set. Plants that exceed 5 positive Salmonella samples in the 53-sample set must initiate corrective action to meet the performance standard and undergo targeted Salmonella testing in accordance with 9 CFR 310.25(b)(3).

Between January 26, 1998 and January 25, 1999, the Agency collected and analyzed 1,184 samples for Salmonella from 25 ground beef-producing plants. The Salmonella prevalence in the 1,184 ground beef samples was 4.8% compared to 7.5% in pre-HACCP baseline studies (Table 1). Although comparisons between Salmonella prevalence during the first year of compliance testing and previous baseline studies must be done cautiously, the Agency finds these results encouraging.

In ground beef, 53 samples comprise a complete sample set. Ten (10) ground beef-producing plants had complete sample sets during this period (Table 2). Of the 15 ground beef-producing plants without complete sample sets, 9 had 47 or more samples collected by the end of the first year of HACCP implementation and are therefore within six working days of having complete sample sets. Therefore, of the 25 ground beef-producing plants that had samples collected for Salmonella in the first year of HACCP implementation, 19 either have complete sample sets or are within 6 working days of completing their sample sets. The remaining six ground beef-producing plants without complete data sets range between 18 and 45 samples collected and analyzed by January 25, 1999.

In the 10 ground beef-producing plants with complete sample sets, the plant-specific prevalence of Salmonella ranged from 0% to 11.3% (Table 5). Of these 10 ground beef-producing plants, 9 (90%) met the 7.5% performance standard (Table 2). Of these 9 plants, 7 (78%) had a Salmonella prevalence of less than 5% in complete sample sets. Follow-up Salmonella testing in the one ground beef-producing plant that did not meet the performance standard has begun.

Ground Turkey

Large federally-inspected establishments that produce ground turkey must meet the 49.9% ground turkey Salmonella performance standard found at 9 CFR 381.94(b). For ground turkey, this equates to a maximum of 29 positive Salmonella samples in a 53-sample set. Plants that exceed 29 Salmonella positive samples in the 53-sample set must initiate corrective action to meet the performance standard and undergo targeted Salmonella testing in accordance with 9 CFR 381.94(b)(3).

Between January 26, 1998 and January 25, 1999, the Agency collected and analyzed 748 samples for Salmonella from 17 ground turkey-producing plants. The Salmonella prevalence in the 748 ground turkey samples was 36.4% compared to 49.9% in pre-HACCP baseline studies (Table 1). Although comparisons between Salmonella prevalence during the first year of compliance testing and previous baseline studies must be done cautiously, the Agency finds these results encouraging.

In ground turkey, 53 samples comprise a complete sample set. Eleven (11) ground turkey-producing plants had complete sample sets during this period (Table 2). Of the 6 ground turkey-producing plants without complete sample sets, 2 had 48 or more samples collected by the end of the first year of HACCP implementation and are therefore within five working days of having complete sample sets. Three additional ground turkey-producing plants began having Salmonella samples collected in either December 1998 or January 1999 and have not had sample collection for a sufficient time to collect a full sample set. These three plants were identified during small plant identification activities in late 1998. Therefore, of the 17 ground turkey-producing plants that had samples collected for Salmonella in the first year of HACCP implementation, 16 either have complete sample sets, are within 5 working days of completing their sample sets, or have only recently begun sample collection. The remaining ground turkey-producing plant without a complete data set had 30 samples collected and analyzed by January 25, 1999.

In the 11 ground turkey-producing plants with complete sample sets, the plant-specific prevalence of Salmonella ranged from 18.9% to 67.9% (Table 6). Of these 11 ground turkey-producing plants, 10 (91%) met the 49.9% performance standard (Table 2). Of these 10 plants, 6 (60%) had a Salmonella prevalence of less than 40% in complete sample sets. Follow-up Salmonella testing in the one ground turkey-producing plant that did not meet the performance standard has begun.

Other Products

Large plants producing products in three other product categories had 303 samples collected and analyzed for Salmonella during the first year of HACCP implementation. Salmonella samples were collected from three steer/heifer plants, one cow/bull plant, and one plant producing ground chicken. Due to the limited number of plants in these three product categories during the first year of Salmonella compliance testing, results from these three product categories will be reported at a later date.


Table 1. Prevalence of Salmonella in meat and poultry products: Post-HACCP implementation results from large plants-January 26, 1998 through January 25, 1999.*

Class of Product Salmonella
Performance
Standard (%)**
Post-HACCP implementation
Salmonella Prevalence (%, n=no. samples)
Broilers20.0%10.9% (n=5697)
Swine 8.7%6.5% (n=1532)
Ground Beef7.5%4.8% (n=1184)
Ground Turkey49.9%36.4% (n=748)
*Reflects testing results from products with 10 or more complete sample sets

**9 CFR 310.25(b) and 381.94(b)



Table 2. Large plant HACCP Salmonella testing performance - January 26, 1998 through January 25, 1999.

Class of Product* Number of Samples in Set Maximum Number of Positives in Set Plants With Complete Sets Percent (Number) Meeting Performance Standard**
Broilers51127691% (69)
Swine5561771% (12)
Ground Beef5351090% (9)
Ground Turkey53291191% (10)
Total11488% (100)
*Reflects testing results from products with 10 or more complete sample sets

**9 CFR 310.25(b) and 381.94(b)



Table 3. Distribution of Salmonella prevalence from 76 large federally-inspected broiler establishments with one complete sample set. January 26, 1998 to January 25, 1999.
Percent of 51 broiler
carcass rinse samples
with Salmonella*
Number (%) of large
broiler plants**
0 – 5%27 (36%)
5.1 – 1020 (26%)
10.1 – 1510 (13%)
15.1 – 208 (11%)
20.1 – 23.64 (5%)
23.7 – 305 (7%)
35 – 401 (1%)
45 – 501 (1%)
Total76 (100%)
* Less than 23.6% equates with the maximum number of positives at 9 CFR 381.95(b)

** Total percentage may not equal 100 due to rounding



Table 4. Distribution of Salmonella prevalence from 17 large federally-inspected swine establishments with one complete sample set. January 26, 1998 to January 25, 1999.
Percent of 55 swine
carcass swab samples
with Salmonella*
Number (%) of large
swine plants**
0.0 – 5.0%12 (71%)
5.1 – 8.70 (0%)
8.8 – 110 (0%)
11.1 – 153 (18%)
15.1 – 201 (6%)
45.0 – 501 (6%)
Total17 (100%)
* Less than 11% equates with the maximum number of positives at 9 CFR 310.25(b)

** Total percentage may not equal 100 due to rounding



Table 5. Distribution of Salmonella prevalence from 10 large federally-inspected ground beef-producing establishments with one complete sample set. January 26, 1998 to January 25, 1999.
Percent of 53 ground
beef samples with
Salmonella*
Number (%) of large
ground beef plants**
0 – 5.0%7 (70%)
5.1 – 7.51 (10%)
7.6 – 9.51 (10%)
9.6 – 151 (10%)
Total10 (100%)
* Less than 9.5% equates with the maximum number of positives at 9 CFR ;310.25(b)

** Total percentage may not equal 100 due to rounding



Table 6. Distribution of Salmonella prevalence from 11 large federally-inspected ground turkey-producing establishments with one complete sample set. January 26, 1998 to January 25, 1999.
Percent of 53 ground
turkey samples with
Salmonella*
Number (%) of large
ground turkey plants**
15 – 20%2 (18%)
30 – 351 (9%)
35.1 – 403 (27%)
40.1 – 451 (9%)
45.1 – 501 (9%)
50.1 – 552 (18%)
65 – 701 (9%)
Total11 (100%)
* Less than 55% equates with the maximum number of positives at 9 CFR 381.95(b)

** Total percentage may not equal 100 due to rounding




Click here for further information about this publication.

Click here to learn more about the Office of Public Health and Science.



URL: http://www.fsis.usda.gov/ophs/salmdata.htm