The relatively low number of positive postoperative
bleeding (P-OB) responses in relation to the total
number of patients limits the complexity of the models
that can be tested. Thus we have not estimated interactions,
and the start model was therefore P-OB ~
gender + diagnosis + age, where ‘~’ indicates that the
postoperative response is modeled with ‘gender’,
‘diagnosis’ and ‘age’ as explanatory variables.