In March of 1999, FDA initiated a 1000 sample survey focused on high volume imported fresh produce. Broccoli, cantaloupe, celery, cilantro, culantro, loose-leaf lettuce, parsley, scallions (green onions), strawberries and tomatoes were collected and analyzed for Salmonella, and E. coli O157:H7. All commodities except for cilantro, culantro, lettuce and strawberries were analyzed for Shigella. Twenty-one countries were represented in the collection and sampling of fresh produce.
Of 1003 samples that were collected and analyzed, 96% were not contaminated with Shigella, Salmonella, and/or E. coli O157:H7. Forty-four samples (4% of the total number sampled) were contaminated with either Shigella or Salmonella while 0% of the produce items were contaminated with E. coli O157:H7. Of the 44 contaminated samples, 35 (80%) were contaminated with Salmonella and 9 (20%) were contaminated with Shigella.
Due to the presence of violative samples, 21 firms were placed on detention without physical examination (DWPE). Seven firms were placed on DWPE due to the presence of Shigella in one composite while 14 firms were placed on DWPE due to the presence of Salmonella in two composites. One firm was placed on DWPE for the presence of both Shigella and Salmonella.
Removal of a product, shipper, grower or importer from DWPE is a decision made by FDA based on evidence establishing that the condition(s) that gave rise to the appearance of a violation have been resolved. To date, 11 of the 21 firms have been removed from DWPE.
Although the incidence of foodborne illnesses linked to fresh produce is low, over the last several years the proportion of foodborne illnesses associated with domestic and imported fresh fruits and vegetables has increased. In January of 1997, President Clinton announced a Food Safety Initiative designed to improve the safety of the nation's food supply. The Department of Health and Human Services (DHHS), the Department of Agriculture (USDA), and the Environmental Protection Agency (EPA) sent a report to the President, in May of 1997, that identified fresh produce as an area of concern. In October of 1997, President Clinton announced a plan entitled Produce & Imported Foods Safety Initiative to provide further assurance that fruits and vegetables consumed by the American public meet the highest health and safety standards.
The problems in this area are self-evident. Most fresh fruits and vegetables are grown in non-sterile environments. Growers have less control over conditions in the field compared to an enclosed production facility. The surfaces of produce have natural microflora composed of microorganisms that are generally not significant to human health, however, low-level contamination of produce with pathogenic microorganisms may sporadically occur. Harvesting, washing, cutting, slicing, packaging and transporting offer opportunities for produce contamination. Most fresh produce is likely to be consumed raw without undergoing processes, such as cooking, that inactivate harmful microorganisms.
In October of 1998, the Food and Drug Administration (FDA) released a guidance document entitled "Guidance for Industry -- Guide to Minimize Microbial Food Safety Hazards for Fresh Fruits and Vegetables." This document outlines good agricultural and good manufacturing practices (GAPs and GMPs) that reduce the risk of microbial contamination of fresh produce. This voluntary guidance does not impose unnecessary or unequal restrictions or barriers on either domestic or foreign products. Areas covered by the guide include water quality, manure management, worker training, field and facility sanitation, and transportation. Guidelines are set to control each potential source of contamination throughout the production process, from the farm to point-of-sale. In the absence of use of GAPs and GMPs, the risk of microbial contamination increases along with the likely extent of the contamination. Implementing GAPs and GMPs increases the assurance that microbial safety hazards will be minimized.
To assist in the development of policy for the Produce & Imported Foods Safety Initiative, the Center for Food Safety and Applied Nutrition (CFSAN) needed data on the incidence and extent of pathogen contamination on selected imported produce. In March of 1999, FDA initiated a 1000 sample survey focused on high-volume imported fresh produce. Eight produce items were selected for the survey: broccoli, cantaloupe, celery, cilantro, loose-leaf lettuce, parsley, scallions (green onions) and strawberries. Loose-leaf lettuce products included radicchio, escarole, endive, chicory and others. These commodities were selected based on the criteria described below. Produce was analyzed for Salmonella, Shigella, Escherichia coli O157:H7, aerobic plate count (APC), and coliforms. Collection of 125 samples was planned for each of the eight produce items.
The objectives of the imported produce survey were to:
Our purpose was not to attempt to detect every incidence of low-level, sporadic contamination but to detect those levels of contamination that might result from a failure to follow adequate GAPs and GMPs as specified by FDA guidance.
Criteria for Selection of Commodities in Survey
There were five risk factors considered in ranking commodities to determine which ones would be selected for the produce survey. Each produce item was assigned a score between 1 and 15 based on the sum of the values given for each of the five risk factors. Each factor was given 1 (low risk), 2 (medium risk) or 3 (high risk) points based on the associated risk. The following risk factors were used:
Sample Collection
Collections were to reflect as many country/producer combinations as possible. All samples were collected aseptically and were shipped under refrigeration to the appropriate District's servicing laboratory. Each sample consisted of ten (10) sub-samples. For cilantro, culantro, loose-leaf lettuce, parsley and scallions (green onions), enough product was collected for each sub-sample to equal 454 g (16 oz.). For broccoli, cantaloupe, celery and tomatoes, each sub-sample consisted of one head/bunch of produce. For bulk shipments of strawberries, each sub-sample consisted of 16 oz. of product. If the product was shipped in retail sized containers, enough containers were collected to equal 16 oz. of product.
Microbial Analysis
Produce was prepared in a manner that closely simulated minimal consumer preparations (e.g., visible dirt removed, stems trimmed, outer leaves removed) but did not include a thorough wash step. Each produce sample consisted of 10 sub-samples, five (5) of which were used to derive two (2) composite samples for Salmonella and the remaining five (5) were used to derive two (2) composite samples for Shigella. The 10 sub-samples were also used to test for the presence of E. coli O157:H7. Laboratories were required to analyze for:
For this assignment, FDA did not approve or recommend reconditioning of detained products. All detained products were destroyed.
If a product/grower combination was found to be violative (i.e., the produce tested positive for Salmonella, Shigella or E. coli O157:H7), subsequent shipments were either detained without physical examination (DWPE)(1) or sampled as "suspect" products under the appropriate compliance program.
Based on the following guidelines, a single confirmed positive does not necessarily indicate that a firm was placed on DWPE. For example, either one (1) composite positive for Salmonella or one (1) sub-sample positive for E. coli O157:H7 in the absence of a second confirmed positive (within the same sample) would not result in DWPE for an implicated firm, however, the entry from which the sample was taken was considered "violative" and refused admission into the U.S.
Confirmed presence of Shigella on produce samples is considered to be a correctable human failure to follow GAPs and GMPs (i.e., human fecal contamination) and represents a potentially significant health hazard. Therefore, it was recommended that the current shipment and subsequent shipments would be detained and the grower/shipper be placed on DWPE after one (1) positive composite was found.
Confirmed presence of Salmonella or E. coli O157:H7 on produce samples may be due to a chance event (e.g., wildlife in field) and would not necessarily indicate poor agricultural practices. However, more than a single sporadic positive could be an indicator of poor practices. Therefore, for samples found with one (1) composite positive for Salmonella or one (1) sub-sample positive for E. coli O157:H7, the shipment yielding the positive sample would be refused admission into the U.S. and the next 10 shipments of that product from that grower/shipper were sampled. If a second positive was detected, the shipper/grower was recommended for DWPE. If a second positive was not detected among the next ten (10) shipments, intensified sampling was discontinued. If two (2) positive samples for Salmonella or E. coli O157:H7 were found in the same entry, the shipper/grower would be placed on DWPE.
Under this assignment, if the grower of the contaminated produce was identifiable (e.g., through records or labeling), a farm inspection by FDA could be initiated with the concurrence of the Office of Regulatory Affairs (ORA) and CFSAN. If a farm inspection was refused or if the findings of the inspection were indicative of unsanitary practices that would lead to microbial contamination, then the grower was placed on DWPE. It was not FDA's intent to conduct farm inspections in all cases where DWPE was imposed, but to do so in some cases for information gathering purposes.
Removal from DWPE
FDA made the decision to remove firms from DWPE based on information packages submitted by the firms that documented the corrective actions taken to remove the conditions producing the violations. The following criteria were considered in the evaluation for the removal of a firm from DWPE:
Modifications to the Survey
Due to certain constraints (e.g., time, availability of produce), modifications to both data collection and analysis components of the original assignment occurred throughout the assignment.
Products were substituted for others when it appeared that not enough entries of an assigned product were available. For example, tomatoes were added to the original assignment and were used to complete the 1000 sample quota. Culantro, a seasoning herb, was added to the assignment when it was found to have a high incidence of contamination. Celery, lettuce, and parsley analyses were discontinued before 125 samples were collected and analyzed due to the unavailability of the products. Collection and analysis of broccoli was discontinued after 36 negative samples under the assumption that broccoli is frequently cooked, and the cooking process would be sufficient to destroy pathogens.
Shigella analysis was performed on all commodities except cilantro, culantro, lettuce and strawberries as specified by the field assignment. While loose-leaf products were not required to be tested for Shigella, one sample of radicchio (loose-leaf lettuce) was analyzed for Shigella upon request of the collecting district. Two composite positives were found in the sample and reported.
To reduce the time needed to perform the microbial analysis, APC and coliform analysis were discontinued early in the assignment.
Countries Represented by Samples
Countries that supply a greater amount of the specific produce items to be sampled per year to the United States (e.g., cantaloupes from Mexico) were sampled more often than countries that supply a limited amount of the specified produce item (e.g., cantaloupes from Canada).
Data Analysis
A total of 1003 samples were collected and analyzed from 21 countries. Produce imported from Mexico, Canada, Costa Rica, Guatemala, the Netherlands, Honduras, Belgium, Italy, Israel, Chile, Peru, Columbia, Trinidad & Tobago, New Zealand, Nicaragua, the Dominican Republic, France, Argentina, Ecuador, Haiti and Korea was sampled. Six countries provided 25 or more samples for analysis: Mexico, Canada, Costa Rica, Guatemala, the Netherlands and Honduras.
Of the 1003 samples, 96% tested negative for pathogen contamination while 4% were positive for pathogen contamination. No shipments were contaminated with E. coli O157:H7. Table 1 shows the number of samples collected and analyzed and the number of confirmed positives for each produce item. Although Table 1 lists the countries from which specific produce items were sampled, the identity of the countries from which contaminated produce was detected is not included.
The three produce items with the greatest incidence of pathogen contamination were cilantro, cantaloupe and culantro, accounting for 1.6, 1.1, and 0.6%, respectively, of the overall contamination (4.4%). The remaining produce items each contributed 0.3% or less to the overall contamination. Broccoli and tomatoes were not found to be contaminated and therefore did not contribute to the overall contamination level of 4.4%. These percent values were calculated based on the number of positives for a particular produce item divided by the total number of samples (1003), multiplied by 100. The sum of these percent values totals 4.4%, which is the percentage of 1003 imported produce samples that were collected, analyzed and identified as contaminated.
| Produce item | # sampled | # positive | Positives-% of # sampled | Positives- % of total | Countries sampledab | ||||||||||||||||||||||||
| broccoli | 36 | 0 | 0.0 | 0.0 | 3, 16 | ||||||||||||||||||||||||
| cantaloupe | 151 | 11 | 7.3 | 1.1 | 3, 6, 7, 8, 10, 11, 12, 16, 19 | ||||||||||||||||||||||||
| celery | 84 | 3 | 3.6 | 0.3 | 3, 16 | ||||||||||||||||||||||||
| cilantro | 177 | 16 | 9.0 | 1.6 | 3, 5, 6, 16, 20, 21 | ||||||||||||||||||||||||
| culantro | 12 | 6 | 50.0 | 0.6 | 6, 21 | ||||||||||||||||||||||||
| lettuce | 116 | 2 | 1.7 | 0.2 | 2, 3, 4, 9, 10, 13, 14, 15, 16, 17, 20 | ||||||||||||||||||||||||
| parsley | 84 | 2 | 2.4 | 0.2 | 3, 5, 6, 13, 14, 16, 20 | ||||||||||||||||||||||||
| scallions | 180 | 3 | 1.7 | 0.3 | 3, 10, 13, 14, 16 | ||||||||||||||||||||||||
| strawberries | 143 | 1 | 0.7 | 0.1 | 1, 2, 3, 16, 18 | ||||||||||||||||||||||||
| tomatoes | 20 | 0 | 0.0 | 0.0 | 2, 16, 17 | ||||||||||||||||||||||||
| Totals | 1003 | 44 | ------ | 4.4 | ------ | ||||||||||||||||||||||||
|
a Indicates that produce from that country was collected and analyzed but not necessarily contaminated.
| |||||||||||||||||||||||||||||
Structural characteristics of the produce items may have contributed to the incidence of contamination detected. For example, webbing on the outer surface of a cantaloupe might provide a suitable environment for bacteria to attach and survive. On the other hand, parsley, a botanical related to the structurally similar cilantro and culantro, might be expected to have similar incidences of contamination. However, parsley had a much lower incidence of contamination than either cilantro or culantro. Therefore, while structural characteristics of the produce are important, there are other factors to consider such as growing conditions and the agricultural practices used by the growers.
Figure 1 shows the incidence of the particular pathogens on produce samples confirmed positive for contamination. No samples were positive for E. coli O157:H7. Of the 44 confirmed positive produce samples, 35, or 80% were positive for Salmonella and 9, or 20%, were positive for Shigella. These findings, along with further collection and analysis led to some firms being placed on Detention Without Physical Examination (DWPE), and, in some cases, farm inspections.
The incidence of Salmonella on the total number of produce items sampled was 3.5% (35/1003) and the incidence of Shigella was approximately 1.0% (9/1003). The incidence of Salmonella on produce may have been associated with contamination from human contact, such as unsanitary food handlers, or from environmental sources. Salmonella is widespread in poultry and swine and is often found in water, soil, animal feces, and on food contact surfaces. This list is not inclusive, however, and the contamination might have been derived from another source. Shigella is transmitted via the fecal-oral route and would likely be transmitted by unsanitary handling by infected food handlers. This microorganism is also found in water polluted with human feces. A farm that does not follow GAPs and GMPs increases the risk for contaminating produce with Salmonella, Shigella, or E. coli O157:H7.
a Broccoli and tomatoes were collected and analyzed but are not included in the figure because none were positive for pathogen contamination.
b None of the samples collected and analyzed were positive for E. coli O157:H7.
Twenty-one firms were placed on DWPE and four (4) FDA-initiated farm inspections were conducted as follow-up. Examination of facilities and observation of standard practices helped to identify sources of contamination and possible corrective actions. Likewise, review of packages submitted from firms requesting removal from DWPE provided useful information about potential sources of contamination and successful corrective actions.
Since the completion of the assignment, 11 of the 21 firms, or 52.4% of the implicated firms that were placed on Import Alert # 99-23, "Detention Without Physical Examination of Raw Fresh Fruits and Vegetables from Listed Manufacturers, Shippers, and/or Growers due to the Presence of Pathogenic Contamination" have been removed from DWPE.
Of the 21 implicated firms, seven firms were placed on DWPE due to the presence of Shigella in one (1) composite while 14 firms were placed on DWPE due to the presence of Salmonella in two (2) composite samples. Tables 2 and 3 show the number of firms placed on DWPE for each produce item, the number of firms removed from DWPE and identity of the country of origin.
Although parsley and strawberry samples were contaminated with Salmonella, neither item resulted in DWPE for the respective shipper/growers, based on the guidelines for a firm to be placed on DWPE. Similarly, other contaminated produce samples did not result in DWPE for the shipper/growers. These samples had a composite sample positive for Salmonella but a second positive was not found, so the firms were not placed on DWPE. Also, one firm found to be violative for Shigella on parsley and scallions, identified and implemented corrective actions before the firm could be placed on DWPE. One shipper/grower of celery was placed on DWPE after both Salmonella and Shigella contamination were detected on the fresh product.
| Produce item | # of firms placed on DWPE | # of firms removed from DWPE | Country |
| cantaloupe | 1 | 0 | Costa Rica |
| cantaloupe | 1 | 0 | Guatemala |
| cantaloupe | 1 | 0 | Mexico |
| celery | 2 | 1 | Mexico |
| lettuce | 1 | 1 | Chile |
| scallions | 1 | 1 | Canada |
| Produce item | # of firms placed on DWPE | # of firms removed from DWPE | Country |
| cantaloupe | 4 | 4 | Mexico |
| celery | 1 | 1 | Mexico |
| cilantro | 6 | 3 | Mexico |
| culantro | 2 | 1 | Costa Rica |
| culantro | 1 | 0 | Trinidad & Tobago |
Based on farm investigations and other information, a failure to follow GAPs and GMPs was often associated with the findings of pathogen contamination. In particular, inadequate manure management and lack of appropriate field and transport sanitation practices was most frequently associated with overall contamination. Specific problems included fields that were open to domestic animals and were fertilized by untreated animal manure, equipment and tools not being sanitized, unsanitary harvesting and/or packing equipment (e.g., woven plastic bags to collect culantro after harvest) and packing practices, and unsanitary methods of transportation (e.g., trucks washed with non-chlorinated water and/or cleaned infrequently). And in at least one instance, a firm placed on DWPE could not provide documentation to certify the cleanliness of the water used for irrigation and fertilization.
Of the 21 firms placed on DWPE, 14 firms were based in Mexico, one (1) firm in Canada, three (3) firms in Costa Rica, one (1) firm in Guatemala, one (1) firm in Chile and one (1) firm in Trinidad & Tobago. Firms from Mexico made a substantial effort to be removed from DWPE. At the completion of this assignment, nine (9) of the 14 firms based in Mexico, or 64.3%, successfully identified and corrected all identifiable sources of potential contamination and were subsequently removed from DWPE.
All firms removed from DWPE identified, implemented and verified all identifiable corrective actions. Corrective measures included fencing areas to eliminate entry by animals, cessation of use of untreated animal manure as fertilizer, food safety training for employees involved in the harvest and/or the maintenance of harvest equipment used in the field, replacement of woven plastic harvesting bags with easily cleanable plastic crates, and increased frequency of truck cleaning with a sanitizer (e.g., chlorine) added to the cleaning water. Proactive measures focused on implementing GAPs and GMPs to ensure a safer food supply for export.
As a consequence of this assignment, food safety training was implemented at firms placed on DWPE for harvest workers to teach proper practice for harvesting and maintaining the harvest equipment in a sanitary manner. Successful implementation of food safety training was a factor in the removal of some firms from DWPE. For at least two (2) firms removed from DWPE, voluntarily implementing changes that followed GAPs outlined in the guide (e.g., training on worker health and hygiene, sanitation of fields and packing facilities, and manure management) helped to reduce the levels of contamination that had been detected. Voluntarily submitted test results documented at least five (5) consecutive clean produce samples as analyzed by an independent third party for at least two (2) of the firms removed from DWPE.
The widespread distribution and consumption of fresh produce demands a better understanding of the risks associated with fresh produce. As a complement to the Imported Produce Sampling Assignment, CFSAN initiated a similar study (the Domestic Produce Sampling Survey) focusing on domestic produce. Eight (8) produce items will be collected and analyzed: cantaloupe, celery, cilantro, loose-leaf lettuce, parsley, scallions (green onions), strawberries and tomatoes. Similar to the Import Assignment, one thousand (1000) produce samples from different regions across the United States will be collected and analyzed for E. coli O157:H7, Shigella, and Salmonella. The domestic survey is to be completed by May 2001. In addition, CFSAN is in the process of issuing a follow-up imported produce assignment targeted at commodities that had a high rate of contamination in the initial imported produce assignment.
1 DWPE means that an import shipment is refused entry into U.S. commerce unless the importer presents evidence, e.g., test results, to FDA showing that the item meets U.S. safety requirements. DWPE can be imposed when violative findings for a grower/shipper are of a nature that suggest that future shipments from that grower/shipper may also be violative. DWPE is imposed to protect consumers from potentially contaminated subsequent shipments from that grower/shipper until the firm implements appropriate corrective measures.
FDA Survey of Imported Fresh Produce, November 1, 1999 (REVISED Imported Produce Assignment - FY 99 DOEP # 99-4)