The present invention concerns methods and systems for improving the accuracy of predicting resistance of a disease to a therapy. In one embodiment of the invention, mean and standard deviation (SD) values of fold change in normalized sensitivity, relative to a laboratory wild type standard pathogen or malignant cell are calculated to demonstrate that the patient samples display inherently different degrees of variation in susceptibility to each therapy. In another embodiment, the 2.times.SD value for each therapy is used as the cut-off between sensitive (within normal susceptible range) and resistant (above normal susceptible range).