Skip to main content
. Author manuscript; available in PMC: 2009 Aug 19.
Published in final edited form as: N Engl J Med. 2009 Feb 19;360(8):753–764. doi: 10.1056/NEJMoa0809329

Table 4.

Ability of the Pharmacogenetic and Clinical Algorithms to Correctly Identify Patients Requiring Low or High Doses of Warfarin.

Actual or Predicted Dose Group ≤21 mg/wk ≥49 mg/wk
Pharmacogenetic Algorithm Clinical Algorithm Pharmacogenetic Algorithm Clinical Algorithm
Validation cohort
No. of patients predicted to require extreme dose 232 162 48 15
No. of patients actually requiring extreme dose 324 324 125 125
No. of patients correctly predicted to require extreme dose 160* 108 31* 9
Percent of patients correctly predicted to require extreme dose 49.4 33.3 24.8 7.2
Derivation-plus-validation cohort
No. of patients predicted to require extreme dose 1250 829 243 98
No. of patients actually requiring extreme dose 1711 1711 625 625
No. of patients correctly predicted to require extreme dose 929* 571 165 57
Percent of patients correctly predicted to require extreme dose 54.3 33.4 26.4 9.1
*

P<0.001 for the comparison of the pharmacogenetic algorithm with the clinical algorithm, with the use of a two-tailed sign test.

This value represents the number of patients correctly predicted to require the extreme dose divided by the number of patients who actually required the extreme dose.