Type of Document	Thesis
Author	Gu, Ning Yan
Author's Email Address	nyg03@fsu.edu
URN	etd-08192004-140344
Title	Medical Care Variations in Florida
Degree	Master of Science
Department	Economics, Department of
Advisory Committee	Advisor Name Title Gary M. Fournier Committee Chair Mary A. Burke Committee Co-Chair Thomas W. Zuehlke Committee Member
Keywords	Rates Hip Fractures Hip Replacement Vaginal Delivery Cesarean Section Corotid Endarterectomy Districts Stent Implant Coronary Artery Disease Stent Angioplasty Bypass Medical Care Variations Utilization Rates Acute Medical Procedures Physicians Panel Data Hmo Descriptive Data Analysis Florida Patient Records
Date of Defense	2004-08-09
Availability	unrestricted
Abstract Background: Literature on regional variation in health care has covered many settings at different periods of time and various locations. Substantial evidence on variations in the use of medical procedures has presented a powerful case for its existence, not only in small regions, but also worldwide. Based on empirical data analysis from previous studies on regional variation in acute medical procedures, there is limited evidence that these conditions exist throughout the State of Florida. Purpose: The purpose of this thesis is to catalog recent medical practice patterns throughout the State of Florida by constructing expanded empirical data sets that cover all 11 districts over a 12-year time span, from 1992 to 2003. Our research, for the first time, provides detailed descriptive analysis on medical care variations in Florida. Data set: The data set used in this report comes from Florida hospital discharge data gathered between 1992 and 2003. Patient records in this data set are from 264 hospitals. There are, roughly, between 1.8 and 2.4 million observations annually which are divided into four quarterly reports. We then regroup the data semi-annually to conduct our research on various illnesses. Methods: Specific medical procedures and some relative diagnostic procedures are carefully selected for conducting this research. The selected medical procedures are: coronary artery bypass grafting (CABG or bypass), percutaneous transluminal coronary angioplasty (PTCA or angioplasty), coronary stent implant (stent), cesarean section births (C-section), carotid endarterectomy and hip replacement, respectively. Aided by SAS computer software, relevant patient records are extracted from the original data set. Depending on the nature of the illness under consideration, the total observations in the sample data sets range from tens of thousands to hundreds of thousands. These large samples enable us to conduct the research with minimized probability of biasness. Based on these sample data sets, specific information on medical procedure patterns and patientsˇ¦ demographic background are retrieved for all 11 districts and over a 12-year time span. Numerous rate comparisons based on various factors like year, district and patientsˇ¦ age are then computed and graphed using the sample data sets. Results: As a consequence of our massive data processing on various illnesses, we find that utilization rates on different acute medical procedures exhibit not only unique patterns but also persistent variations across 11 districts in the 12-year period. Depending on the nature of the illness, some may demonstrate higher variations than others. For instance, invasive treatments for heart problem patients like bypass and angioplasty have much higher regional variations than the invasive treatment like hip replacement. Further, we find that the utilization rates of different medical procedures show either divergent or convergent patterns over time. For example, in our sample data, coronary stent implant sample data set is extracted from the patient records on angioplasty. That is, for those patients who receive coronary stent implant, they must undergo angioplasty but for those patients who receive angioplasty, they may or may not undergo coronary stent implant. We find that utilization rates of coronary stent implant (introduced in 1995) show convergent pattern in 2003 while there is no immediate convergence expected in the near future for either bypass or angioplasty rates (see Chapter 3). Most importantly, although no sign of convergence, patients in a district with younger population have higher probability of receiving bypass and angioplasty than patients in a district with relatively older population. This pattern of variations shows a 180„a change on coronary stent implant patients, that is, prior to convergence, patients in a district with younger population have lower probability of receiving coronary stent implant than patients in a district with relatively older population. Moreover, results obtained from studies on other health conditions like carotid endarterectomy, childbirth and hip replacement all demonstrate, although with different magnitude, persistent medical care variations in Florida.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access    Thesis.pdf 2.07 Mb 00:09:34 00:04:55 00:04:18 00:02:09 00:00:11