Assigning these cases to the Rachs-1 categories, there were 64 level I, 188 level II, 90 level III, 12 level IV cases, with 32 "others". The predicted mortality for the 354 categorized cases calculates to be 17.4 (Rachs-1) and 10.2 (STS).
The ROCmax method categorized cases into the same category as ISSAIS 59% of the time, and into the same or an adjacent category 92% of the time.
The GEMmax method categorized cases into the same category as ISSAIS 69% of the time, and into the same or an adjacent category 94% of the time.
The GEMmin method categorized cases into the same category as ISSAIS 72% of the time, and into the same or an adjacent category 95% of the time.
We categorized cases using the Brighton collaboration clinical case definition and calculated the national incidence rate.
To validate the cases, we categorized cases as likely if they had more than 1 diagnosis code for a fracture-healing complication recorded in the database.
On the basis of the findings at autopsy, we further categorized cases as an infectious cause of death (ICOD) (ncod = 99) and infection incidental to death (ninc = 26).
When only lobular carcinoma in situ (LCIS) or atypical lobular hyperplasia (ALH) was present (n=2), we categorized cases as non-proliferative.
We have applied the multinomial regression analysis method to a GWA study of T2D from the main WTCCC experiment [ The Wellcome Trust Case Control Consortium, 2007], where cases were categorized cases according to obesity.
Old lymph node stage measures categorized cases according to whether they had none, one to three, or four or more involved nodes, and recently according to more detailed subdivisions [ 10, 11] (see Additional files 2, 3, 4, 5 and 7).
