This dissertation describes how to apply pharmacokinetic simulations and modeling in a clinical setting to monitor and adjust drug dosing in special patient populations. Pharmacokinetic simulations were used to investigate efficacy and risk of drug toxicity of a new dosing regimen for aminoglycoside antibiotics when administered to renal failure patients. The current method of administering aminoglycosides to renal failure patients is to dose the drug during the last half hour of dialysis sessions. The new proposed method suggests dosing the drug during the first half hour of the dialysis session. Using one-compartment model infusion equations, both methods were simulated to predict drug peaks, troughs and area under the curves. These parameters were used to compare both dosing regimens to find out if the proposed dosing regimen can be suggested in a clinical setting to obtain the same efficacy and lower risk of drug toxicity. The dissertation then describes a prospective clinical study in chronic renal failure patients who received the same tobramycin dose using current and proposed dosing regimens. Results from the clinical study confirm pharmacokinetic simulations and modeling outcomes. Results suggest that both regimens have the same efficacy, but the new proposed method is expected to have lower risk of drug toxicity. The dissertation also describes a retrospective study for vancomycin dosing in renal failure patients. The objective was to confirm that pharmacokinetic modeling could be used to predict and adjust vancomycin dosing for this special population. Vancomycin trough concentrations obtained from patient medical records were compared to predictions obtained using a pharmacokinetic model. It was concluded that there was no statistically significant difference between actual and predicted vancomycin trough concentrations. These results suggest that the pharmacokinetic model can be used to predict and adjust vancomycin dosing to chronic renal failure population. The last part of this dissertation describes evaluation of insulin glargine effect on glycemic control and weight change in a diabetic population. Glycemic control and weight of patients before and after initiation of insulin glargine were evaluated retrospectively. Results showed that initiation of insulin glargine improved glycemic control while weight remained relatively stable. / Graduation date: 2005

This descriptive, correlational designed study was concerned with noncompliance with therapeutic regimens, a pervasive clinical problem which is confounded with the lack of a strong link among theory, research and practice. The focus was on one of the constructs included the Reciprocal Interaction Model of Compliance Behaviors, which was derived using a modified grounded theory methodology and following various theory building prescriptions. The overall purpose was to begin to evaluate the goodness-of-fit of this empirically, qualitatively and retroductively generated explanation of compliance behaviors. The specific aims were to develop, refine and test a 6-point response, 64-item Likert-type instrument, Olivas' Health Motivation Scale - OHMS, that adequately measures the construct, Health Motivation: the force within the patient which is developed as he/she gains experience with his/her illness as a function of time. It has two major dimensions: expectations and values. Health Motivation as indexed by an expectations/values interaction was predicted to impact compliance as measured by dietary and medication measures, both objective and subjective estimates. Using trait and nomological construct perspectives, the OHMS was systematically evaluated by internal and external association criteria and therefore validity and reliability estimates, with a purposive sample of 84 heterogeneous hemodialysis patients who represented two cultures (Anglo and Hispanic), varying in gender, age and length in hemodialysis. Internal consistency reliability and trait construct validity were derived through Cronbach's alpha and principal components factor analysis. Refined OHMS Scales had alphas and thetas ranging from .58 to .89. Explained scale variance ranged from .54 to .84. Epistemic coefficients, the validity links between concept and operational measures, ranged from .76 to .94. Internal validity of the design, estimated through multiple regression, was concluded to be satisfactory. External association assessment via multiple regression produced mixed findings. Select expectations, in linear combination with select values, explained varying degrees of the variance, in select compliance measures, R² = .11 to .44. Through empirical modeling via path analysis, select subject characteristics (ethnicity, length on dialysis, age) were found to have direct or indirect relationships with compliance. Theory, research, and practice based limitations and recommendations were made from the results of the study. (Abstract shortened with permission of author.)

The purpose of this study was to identify determinants of compliance behavior. Since compliance among dialysis patients increases survival rate, it is important for social workers to identify patients who are at risk of noncompliance.