Patients under methadone flexi dispensing (MFlex) program are required to do methadone dosages trends for down-up case since no parameters were employed to identify the patient who has potential rate of recovery. So that, the existing system need further enhancement towards classification and optimization. The objective is to determine the Mahalanobis distance (MD), evaluate the degree of contribution, and diagnose the unknown data in the MFlex program using Mahalanobis-Taguchi system (MTS). The data is obtained at the Bandar Pekan clinic that includes 16 parameters. Two types of MTS methods are used for classification and optimization using RT-Method and T-Method respectively. As a result, the RT-Method is able to classify between healthy and unhealthy samples, while the T-Method is able to evaluate the significant parameters in terms of degree of contribution. 4 unknown samples have been diagnosed with different number of positive and negative degree of contribution to achieve lower MD. The best-proposed solution is type 3 of 4 modifications because it shows the lowest MD value than others. In conclusion, a Bandar Pekan clinic pharmacist confirmed that MTS can solve a classification and optimization problems in the MFlex program.
