published_or_final_version / Public Administration / Master / Master of Public Administration

published_or_final_version / Public Administration / Master / Master of Public Administration

This study examined if physician assistants (PAs) are cost-beneficial to employers. In an era of cost accountability, questions arise about whether a visit to a PA for an episode of care differs from a visit to a physician, and if PAs erode their cost-effectiveness by the manner in which they manage patients. Four common acute medical conditions seen by PAs and physicians within a large health maintenance organization were identified to study. An episode approach was undertaken to identify all laboratory, imaging, medication and provider costs for these diagnoses. Over 12,700 medical office visits were analyzed and assigned to each type of provider and medical department. Patient variables included age, gender, and health status. A multivariate analysis identified significant cost differences in each cohort of patients. In every condition managed by PAs, the total cost of the visit was less than that of a physician in the same department. This was significant for episodes of shoulder tendinitis, otitis media, and urinary tract infections. In no instance were PAs statistically different from physicians in use of laboratory and imaging costs. In each instance the total cost of the episode was less when treated by a PA. Sometimes PAs ordered fewer laboratory tests than physicians. There were no differences in the rate of return visits for a diagnosis between physicians and PAs. Patient differences were held constant for age, gender, and health status. This study affirms that PAs are not only cost-effective from a labor standpoint but are also cost-beneficial to those who employ them. In most cases, they order resources for diagnosis and treatment in a manner similar to physicians for an episode of care, but the cost of an episode of an illness is more economical overall when the P A delivers the care. This study validates the federal policy of support for primary care P A education and suggests that PA employment should be expanded in many sectors of the health care system. These findings and the results of this cost-benefit model are evidence of its validity in predicting health care costs.

Abstract not available

There has been an almost 60 percent increase in health care expenditures in the US in the past seven years. Employer-sponsored health coverage premiums have increased significantly (87 percent) in this same period. Besides the cost of care for chronic conditions such as migraine, arthritis and diabetes, absenteeism linked to these diseases also adds financial strain. Current health financial models focus on past spending instead of modeling based on current health burdens and future trends. This approach leads to suboptimal health maintenance and cost management. Identifying the diseases which affect the most employees and are also the most costly (in terms of productivity, work-loss-days, treatment etc) is necessary, since this allows the employer to identify which combination of policies may best address the health burdens. The current predictive health model limits the amount of diseases it models since it ignores incomplete data sets. This research investigated if by using Bayesian methodology it will be possible to create a comprehensive predictive model of the health burdens being faced by corporations, allowing for health decision makers to have comprehensive information when choosing policies. The first specific aim was to identify which diseases were the most costly to employers both directly and indirectly, and the pathogenesis of these diseases. Co-morbidity of diseases was also taken into account as in many cases these diseases are not treated independently. This information was taken into account when designing the models as the inference was disease specific. One of the contributions of this thesis is coherent incorporation of prior information into the proposed expert model. The Bayesian models were able to estimate the predicted disease burdens for corporations, including predicting the percentage of individuals with multiple diseases. The model was also comparable to, or better than current estimators on the market with limited input. The outputs of the model were also able to give further insight into the disease interactions which creates an avenue for further research in disease management.

This paper considers the adoption of Health Information Technology (HIT) by physician clinics with ten or fewer physicians. The paper considers the theoretical economics literature on technology adoption for a new technology and has a place in the empirical tests of these models. The two major hypotheses tested in the paper are that the probability of adopting HIT increases with the number of physicians working at the clinic and if the clinic is part of a chain of clinics, and that it also increases with increased competition at the market level measured by the number of clinics per 10,000 residents in a county. To test these hypotheses, the paper first estimates a baseline logit model followed by three hazard rate models. In each case, clinic size is found to have positive though not significant effect on the probability of adoption (in the logit model) or to decrease the predicted time to adoption for the clinic (in the hazard rate models), being in a chain of clinics is found to have a strong positive and significant on the probability of adoption, and increased competition is found to have a positive though not significant effect on the probability of adoption.

Thesis (Ph.D.)--Brown University, 2008. / Vita. Includes bibliographical references.

Source: Masters Abstracts International, Volume: 46-01, page: 0311. Adviser: Carl Hacker. Includes bibliographical references.

Introduction: New treatments for stage IV (adv) NSCLC have emerged this past decade. Recent pharmacoeconomic research has focused on cost of treatment, comparative costs of therapies, and cost/cost effectiveness of adding a biologic to traditional therapy. Drug cost is thought to be a primary driver of cost change in NSCLC, yet to our knowledge, characterization of the direct cost of NSCLC has not been published since the new treatments have emerged in the guidelines. Our primary objective was to characterize the direct and segmented cost of adv NSCLC from 2000-9. We also want to determine cost impact of new therapies, and cost trend from 2000-9. Methods: This PharMetrics claims database study includes diagnosed NSCLC patients [greater than or equal to] 20 yo. Small cell lung cancer was excluded. Claims were divided into disease segments and time periods representative of changes in therapy ("pre" (2000-2), "transition" (2003-5), and "current" (2006-9) periods). Descriptive statistics (median, interquartile range (IQR)), chi-square test (nominal data), and Wilcoxan rank sum tests were performed on the data. To adjust for baseline confounders, multivariate least squares regression models were created. Results: Costs are reported as medians in terms of per patient per month (pppm). Overall monthly cost (n=969) was $10,281 pppm. Diagnosis cost $6,601 pppm, active treatment cost $9,287 pppm, and end-of life cost $12,215 pppm. There was no difference in cost between the “transition” (n=439) and “current” (n=503) periods overall or for any segment of disease. Comorbidities had no effect on cost. For patients receiving at least 5 months of active treatment medication (n=316) total median cost was $144,147 per patient ($9,371 pppm). Discussion: There was no difference in cost between the transition and current periods, in regards to either overall cost or segmented cost. The most expensive segment was end-of-life, with a median cost exceeding $12,000 pppm. Surprisingly, comorbidities had no effect on cost. Newer agents (biologics, TKIs, and pemetrexed) represent only a modest portion of cost, with a majority of cost for stage IV NSCLC comprised of non-drug costs. / text

published_or_final_version / Public Administration / Master / Master of Public Administration