Date of Award

Summer 1999

Document Type


Degree Name

Doctor of Philosophy (PhD)


Urban Services--Health Services

Committee Director

Gail Grisetti

Committee Member

Stacey Plichta

Committee Member

Clare Houseman


The purpose of this study was to explore relationships among population characteristics, patient post-hospital resource use and patient health outcomes in terms of the Andersen health behavioral model. An additional purpose of this study was to examine a discharge planning screening instrument used in a large urban hospital in Taiwan to determine if it predicts resource use and patient outcomes. A longitudinal research design was adopted for this study. Data were collected from the general medicine departments of an 800-bed university teaching hospital. Out of 109 patients who were interviewed, 78 participants qualified (participation rate = 71.6%).

The majority of participants were aged 65 to 84 (97%) with an average age of 74.9 (SD = 5.92) and female (61.5%). Ninety percent of the participants lived with a spouse or family members. Nearly all participants (95%) had family support while they were sick. Over 60% of the participants had a low individual income. Over half of the participants had visited an ER or a hospital in the past year. Functional status scores for the participants revealed that most of them had problems with bathing, toileting, riding the bus, walking, and shopping. Three quarters of the participants were fully alert, half of the participants had some kind of cognitive problems. Approximately 20% had a hearing/visual deficit, and approximately 60% had more than one medical problem.

The results of the multivariate analyses indicated that 14 out of 25 independent variables were able to predict hospitalization outcomes, postdischarge resource use, or patient health outcomes. Provider-related enabling factors were strong predictors of patient outcomes. However, 11 variables were not predictors of patient outcomes or resource use. The findings did not fully support the Andersen model. The variables identified in this study might not fully explain the constructs of the Andersen model. Recommendations for a discharge planning screening instrument, the Taiwanese government, health care professionals, and future studies are discussed.