I know these groups have had good luck using the assays described in the attachments. Factors that might cause variations in results could include the thermocycler used and the quality of the polymerase so there might be some optimization on your end if you do not use similar equipment or reagents.
Do you want to characterize pure cultures or detect these serotypes in clinical samples?If you want to characterize pure cultures and you have the capacity for sequencing gnd-typing might be more straight forward to type STEC strains as you can also identify a wider variety of serotypes.