Skip to main content

Table 6 Probabilistic multiplication of subfactors

From: Dispelling urban myths about default uncertainty factors in chemical risk assessment – sufficient protection against mixture effects?

Reference

Database used

Distribution parameters

Results

Sheehan et al. 1990 [70]

Interspecies variation

Ratios of tumour incidences (TD50s) for 190 chemicals in mice and rats

Median = 2.6

 

Intraspecies variation

Ratios of acute lethality (LD50s) for adult and newborn mammals for 238 chemicals

Median = 2.4

 

Overall assessment

  

Values exceeding 100: 11.8% predicted, 10% observed

Baird et al. 1996 [107]

Interspecies variation

69 pesticides tested in different animal species, allometrically adjusted for body surface (Dourson et al. 1992)

Median = AS1

 
  

GSD = 5

 

Intraspecies variation

Probit dose–response slopes from 490 acute lethality experiments using rats [75], assuming two different levels of protection;

Basic approach:

 
  

Median = 2.7

 
  

GSD = 2.3

 
 

Basic approach: 1/100,000

Alternative approach:

 
 

Alternative approach: 1/1,000

Median = 5.3

 
  

GSD = 1.4

 

Overall assessment

RfDs or RfCs for 126 compounds with NOAELs from chronic bioassays in IRIS database

Basic approach:

Fraction of RfDs within the lower 5% of distribution of potential threshold values2 ;

  

Median = AS x 3

All: 56%

  

P95 = AS x 50

Mice: 23%

  

P99 = AS x 220

Rats: 39%

  

Alternative approach:

Dogs: 98%

  

Median = AS x 5

 
  

P95 = AS x 63

 
  

P99 = AS x 194

 

Vermeire et al. 1999 ; Vermeire et al. 2001 [5, 49]

Interspecies variation

184 substances tested in mice, rats and dogs

GM = AS

Factor 12 (4 for allometric scaling x 3 for remaining uncertainty) coincides with 73rd percentile.

  

GSD = 4.5

 
  

P95 = AS x 19

 
  

P99 = AS x 65

 

Intraspecies variation

Theoretical, to be consistent with default factor 10, P99 = 10 [108]

Median = 1 + 3

 
  

GSD = 1.6

 

Overall assessment

 

GM = AS x 4

Percentile of the default factor 100: 79% (NOAEL in mouse), 88% (NOAEL in rat)3

  

GSD = 4.7

 
  

P95 = AS x 53

 

Gaylor and Kodell 2000 [109]

Interspecies variation

Binary aquatic interspecies comparisons from dozens to over 500 agents [72]

Median = 1

 
  

GSD = 1.66

 

Intraspecies variation

Probit dose–response slopes from 490 acute lethality experiments using rats [75] adapted by Dourson and Stara [31]

Median = 1

Default value of 10 corresponds to the 92nd percentile

  

GSD = 1.64

 

Overall assessment

 

Median = 1

 
  

GSD = 2.33

 
  

P95 = 46

 
  

P99 = 230

 

Schneider et al. 2005 [106]

Interspecies variation

63 antineoplastic agents in humans and five different animal species [68]

GM = AS x 0.97

 
  

GSD = 3.45

 
  

P95 = AS x 6.7

 
  

P99 = AS x 15

 

Intraspecies variation

Human database for predominantly healthy adults developed by Hattis et al. [79]

GM = 3.8

 
  

GSD = 4.3

 
  

P95 = 44

 
  

P99 = 117

 

Overall assessment

Our own calculation

GM = AS x 3.7

Proportion of substances for which the default factor 100 would not be exceeded:

  

GSD = 5.4

AS based on caloric demand; 76% (mouse), 85% (rat)

  

P95 = AS x 82

AS based on surface area; 64% (mouse), 79% (rat)

  

P99 = AS x 295

 

Hasegawa et al. 2010 [110]

Interspecies variation

63 antineoplastic agents in humans and five different animal species adapted from [68]

GM = AS

 
  

GSD = 3.23

 
  

P95 = 48.2 (mice)

 
  

P95 = 27.5 (rats)

 

Intraspecies variation

Rat young/newborn NOAEL ratios for 18 industrial chemicals [98]

GM = 3

 
  

GSD = 1.38

 
  

P95 = 5.09

 

Overall assessment

 

P95 = 155 (mice)

 
  

P95 = 88.7 (rats)

 
  1. References
  2. 1 Allometric scaling factor according to body surface area (mouse = 14, rat = 6).
  3. 2 For 231 RfDs including some derived from LOAELs and/or sub-chronic bioassays.
  4. 3 Allometric scaling factors based on caloric demand were used.