Defense Date


Document Type


Degree Name

Doctor of Philosophy


Health Administration

First Advisor

Thomas T.H. Wan


This study was undertaken to address the need of professionals responsible for assuring the quality of hospital care for a framework for understanding and evaluating quality assurance mechanisms and their impact on hospital quality of care. Primary data were collected from 70 Virginia short term acute care general hospitals on the design and resources of their quality assurance programs in 1986. Adverse outcome data for 1986 were collected from the Medical Society of Virginia Review Organization. Hospital structural data were obtained from the American Hospital Association computer data base and the Federal Register. The intermediate outcome variables are: rate of unexpected return to the operating room, rate of treatment/medication problems, rate of in-hospital trauma, rate of medical instability at discharge, and rate of unexpected deaths.

Exploratory analyses of hospital size and specialization demonstrate that size positively affects the numbers of RNs in quality assurance, the number of quality assurance professionals with academic degrees above the associate level, and negatively affect the ratio of quality assurance personnel full-time equivalents (FTEs) — both total and professional — to total hospital FTEs. Hospital specialization negatively affects the ratio of quality assurance personnel FTEs — both total and professional — to total hospital FTEs.

Structural equation models, causally relating the adequacy of quality assurance design and resources to adverse outcomes of hospitalization, were used to test the causal relationships. The model supports the work of Donabedian and of Deming. The model demonstrates the effects of quality assurance constructs on perceived organizational commitment to quality assurance and commitments effect on process-related outcomes. Process-related outcomes are strongly and positively related to the terminal measure of unexpected deaths.

When size and specialization are controlled, some changes are noted in the model. The R2 increases, the Chi-square/df ratio increases and the adjusted goodness of fit ratio decreases. This change was not unexpected due to the statistical significance of the percent of board certified physicians (BRDCERT) on the outcome variable unexpected death (DEDPROBR).


Scanned, with permission from the author, from the original print version, which resides in University Archives.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission