• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Animal & Veterinary

  • Print
  • Share
  • E-mail

Use of Risk Model to Calculate Resistance Thresholds

SLIDE 1:

Use of Risk Model to Calculate Resistance Thresholds

Mary Bartholomew
Expert Mathematical Statistician
Center for Veterinary Medicine
January 24, 2001

 

SLIDE 2: Risk-based Approach

Human Cases Caused by Resistant Bacteria(attribute)k-res
<-----
Quality of Meat with Resistant Bacteria(exposure)

 

SLIDE 3: A Measurable Human Health Impact, H(x)

H(x) = current measurable level of human health impact resulting from the use of an antimicrobial drug in food-producing animals causing resistance to an antimicrobial drug used in human medicine

 

SLIDE 4: For Measurable Values Q and H(x) it Follows That

k-res = H(x)/Q

 

SLIDE 5: Since T(x) is a Particular Value of H(x), There is a Particular Value of Q for Which the Same Relationship Holds,

k-res = H(x)/Qt(x)

 

SLIDE 6: How Q is estimated

Q =( total pounds of product consumed) x (proportion of sampled pounds containing bacteria) x (proportion of samples from which resistant bacteria are isolated (h))

 

SLIDE 7: Defining Q and Qt(x)

Proportion of samples from which resistant bacteria are isolated is

  • called h in defining Q (as just shown)
  • called t(x) in defining Qt(x)

 

SLIDE 8:

t(x) - the maximum allowable prevalence of resistant bacteria isolated from animal-derived food that does not cause an unacceptable impact on illness x in humans

 

SLIDE 9: During a period of stable

  • Consumption
  • Rates of carriage of bacteria on product
  • Rates of human prescriptions

 

SLIDE 10:

H(x)/Q = T(x)/Qt(x)

and further,

H(x)/h = T(x)/t(x)

 

SLIDE 11:

This allows a solution for t(x)

t(x) = [T(x) *h]/H(x)

 

SLIDE 12: Maximum Impact Method

  • Without current information, can use Max
  • Assume all isolates from food commodity are resistant
  • Assume all human cases attributable to the food commodity are caused by resistant bacteria
  • Then h = 1, Hmax = all attributable cases

 

SLIDE 13:

A graph titled, "Human health impact of antimicrobial drug resistance attributed to use of an antimicrobial drug in food-producing animals.

X axis - Prevalence (impact) of resistance among food product exposure unit (h)

Y axis - Human Health Impact H(x)

 

SLIDE 14: Thresholds for Multiple Species

  • Σai[ti(x) k-resi] T(x)
  • uses weights, ai, to allocate t(x) for each species so that the total is less than T(x)