Defense Date


Document Type


Degree Name

Doctor of Philosophy



First Advisor

Dr. Cynthia K. Kirkwood

Second Advisor

Dr. Patricia W. Slattum


Prescription drug abuse is a continuing problem in the United States. Educating physicians on issues related to prescription drug abuse is a key factor in preventing and treating this problem. High variability has been found in substance abuse terminology in the literature, textbooks, and FDA-approved product labeling. This dissertation describes a survey study designed to address how the variability in substance abuse terminology, specifically package inserts, affects the prescribing decisions made by physicians.A random sample of 1008 physicians currently licensed and residing in the Commonwealth of Virginia received a letter of explanation, a self-administered questionnaire, and a follow-up reminder and thank you. To increase response rate, a second questionnaire was sent to non-responders. Prescribing decisions made by physicians were measured as three variables: comfort level with a prior physician's choice, likelihood of refilling the prescription, and likelihood of prescribing a drug or drug class as the first physician seeing a particular patient. Physicians were presented with four case scenarios which included package insert information and selected patient characteristics. Other factors affecting physicians' decisions in prescribing controlled substances include ideas about addiction, and characteristics of the physician, patient, disease state, and drug. The patient case scenarios and other items on the questionnaire addressed these covariates. Based on the number of deliverable questionnaires returned and included in analysis, the response rate was 32.3%. More physicians associated abuse, craving, drug-seeking behavior, psychological dependence, and withdrawal with addiction than with drug dependence, while more physicians felt that physical dependence and tolerance were necessary for drug dependence. The most frequently used sources for drug information were the Physicians' Desk Reference (PDR), package inserts, and pharmacists. Four linear regression models were created for physician prescribing decisions. Physician, patient and package insert characteristics were all significant (p


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

June 2008