Salmonella in Peanut Products: A Manageable Risk

1
Salmonella in Peanut ProductsA Manageable Risk

• Don L. Zink, Ph.D.
• U.S. Food Drug Administration
• Center for Food Safety and Applied Nutrition

2
Incidence of Laboratory-confirmed Bacterial and
Parasitic Infections
Data from CDC, http//www.cdc.gov/mmwr/preview/mmw
rhtml/mm5813a2.htm?s_cidmm5813a2_e
3
Salmonella sp. Outbreaks and Illnesses Linked to
FDA-Regulated Foods, By Year
4
Salmonella sp. Outbreaks and Illnesses Linked to
FDA-Regulated Foods, By Year(excludes eggs)
5
Salmonella sp. Outbreaks and Illnesses Linked to
FDA-Regulated Foods, 1996-2009,Processed foods
only
6
How Salmonella Becomes a Problem
Outside Environment
Growth Niche
Wet Dry Cycle
Multiplication And Spread
Food Food Contact Surfaces
7
Why Salmonella is a Problem

• It survives long periods of dehydration
• In can be all over a food plant that looks
  clean
• It can live and grow in a crack in the floor
• Cleaning and sanitizing alone wont get rid of it
• Its nearly impossible to find by product testing
  alone
• Control requires a continual effort
• It is a low-dose pathogen

8
Design Salmonella out of the Plant
Equipment mounted to be Elevated above the floor
Floors that are impervious to water and dirt
Equipment that is clean and Well-maintained
9
Design Salmonella out of the Plant
Prevent the movement of water from floor-to-floor
10
Design Salmonella out of the Plant
Keep an eye on your maintenance staff and
contractors
11
The Truth About Finished Product Testing for
Salmonella

• Finished product testing has a purpose
• Verification of other control measures.
• Creating high hygiene areas where product is
  exposed
• Using validated processes to kill Salmonella
• Limit traffic between raw and processed areas
• Qualifying suppliers
• Proper environmental testing
• A single positive means something is wrong

12
The Truth About Finished Product Testing for
Salmonella

• Finished product testing will NOT detect low
  rates of contamination
• Salmonella is typically present at LOW rates of
  contamination
• 60 sub-samples from a lot cannot reliably detect
  a 1 defect rate (you need 300 sub-samples to
  have a 95 chance of finding a lot with a 1
  defect rate). 95 is not good enough for food
  safety
• Finished product testing IS NOT a control program

13
The Truth About Finished Product Testing for
Salmonella

• Retesting a lot that initially tests positive is
  a fools game
• A single positive on a lot sample makes the lot
  positive, regardless of how many samples later
  test negative
• The odds favor that a retest sample will be
  negative
• False positives almost never occur if the test is
  carried to completion with molecular or serotype
  ID

14
Observations from Recent Outbreaks Involving
Processed Foods

• A failure to appreciate the importance of
  Salmonella control in certain high risk foods
• Too much reliance on finished product testing
• A failure to implement environmental monitoring
  or ineffective implementation of monitoring
• Inadequate response to positive Salmonella
  findings in finished products and the environment
• Poor implementation of GMPs
• Inadequate qualification of critical ingredient
  suppliers
• Inadequate cleaning, sanitation and pest control