Date of Award
Fall 2000
Document Type
Dissertation - Restricted
Degree Name
Doctor of Philosophy (PhD)
Department
Mechanical Engineering
First Advisor
Marklin, Richard
Second Advisor
Cariapa, Vikram
Third Advisor
Merrill, Stephen
Abstract
Current physiologic monitoring alarm systems are weak because they exhibit a high false positive rate and true positive alarms that are often clinically irrelevant. In addition, they only measure what is convenient, safe, cost effective, and physically possible rather than what would be most useful. While an extensive body of literature exists on proposed improvements to these alarms, solutions presented are either not likely to be implemented commercially or are simplistic and not well grounded in physiology. Without a commercially viable solution, the solution might as well not exist. The objective of this research is to identify a process that tailors the current alarm system of a physiologic monitor in such a way as to more accurately identify ominous problems with a patient. The modification of the alarm system will be limited to defining a more appropriate set of alarm thresholds for existing physiologic variables. The key to the solution is to use a tool to peer inside the body to the variables that one would like to measure if it were practical. Mathematical models of physiologic processes will be used to gain access to these elusive variables. The relationship will be explored between these primary "wellness" variables and the familiar measurable variables from a physiologic monitor. My research will establish the optimum alarm limits for measurable variables by relating them to wellness variables.