Classification results of equation.

This study aimed to assess the discriminant factors and determine the cutoff value which can predict or classify the group of prescribed and non-prescribed biosimilar among Thai physicians. The online surveys were distributed to physicians via major three medical associations in Thailand. Five psychological variables were obtained from the surveys (familiarity, attitude toward biosimilar medication, attitude toward biosimilar practice scenarios, attitude toward naming biosimilars in prescriptions, and attitude toward pricing of biosimilar). The point-based system was used to score all variables and transformed to percentage. The assumptions were tested before using discriminant function analysis (DFA). Total 82 respondents were analyzed. Data of all variables met the assumptions of DFA. Familiarity was the most influential factor to differentiate the group of physicians, followed by attitude toward biosimilar medication. The cutoff value for group differences was −0.600. The accuracy rate of discriminant function equations was 82.9% overall for the stepwise method. The study concluded that the psychological factors such as familiarity with biosimilars and attitude toward biosimilars would play a significant role to classify between prescribed and non-prescribed biosimilar groups.