. 2021 Aug 17;20(1):91–104. doi: 10.1007/s40258-021-00675-z

Table 3.

Base-case distributions for the probabilities of minor, moderate or severe harm for a harmful, undetected error for all error types and with both current and user-tested guidelines

Error type	Dirichlet distributions (minor harm, moderate harm, severe harm)
Error type	Current guidelines	User-tested guidelines
I1	Dirichlet (0.1, 0.1, 0.1)^a	Dirichlet (0.1, 0.1, 0.1)^a
I2	Dirichlet (0.1, 0.1, 0.1)^a	Dirichlet (0.1, 0.1, 0.1)^a
I3	Dirichlet (5.1, 5.6, 0.1)	Dirichlet (2.1, 2.1, 0.1)
I4	Dirichlet (0.1, 0.1, 0.1)^a	Dirichlet (0.1, 1.1, 0.1)^a
I5	Dirichlet (0.1, 6.1, 0.1)	Dirichlet (0.1, 0.1, 0.1)^a
I6	Dirichlet (0.1, 27.1, 0.1)	Dirichlet (3.1, 45.1, 0.1)
I7	Dirichlet (0.1, 0.1, 0.1)^a	Dirichlet (0.1, 0.1, 0.1)^a
I8	Dirichlet (0.1, 3.1, 0.1)	Dirichlet (0.1, 6.1, 0.1)
I9	Dirichlet (4.1, 29.1, 0.1)	Dirichlet (0.1, 11.1, 0.1)
I10	Dirichlet (0.1, 0.1, 0.1)	Dirichlet (0.1, 0.1, 0.1)^a
I11	Dirichlet (19.1, 3.6, 5.1)	Dirichlet (0.1, 2.1, 1.1)

These distributions were applied to node 4 of Fig. 1

^aDistribution based entirely on a priori assumption of 0.1 for each error type and severity (see Sect. 2.6.1)