在醫療服務的研究領域中，醫療資源的使用與其相關因素的探討一直是許多人關心的焦點。以美國為例，每年有數百萬元的政府預算耗用在與醫療有關的服務上。儘管美國政府千方百計的想要降低醫療服務的花費，並且推出各種替代方案來控制醫療花費，成效並不顯著。有鑑於每年有龐大的經費被用在醫療服務的領域，瞭解醫療資源使用的相關因素就顯得十分的重要。 在研究醫療資源使用的領域中，Andersen的行為模式（behavioral model）是廣為學者們引用的理論模式。該模式認為，一個人是否使用醫療服務會受下列三個主要因素所影響：傾向因素（predisposing factors），此因素包括個人的性別、年齡、種族、婚姻狀況、教育程度、宗教信仰、與對醫療專業的觀感等；使能因素（enabling factors），此因素包括收入、財產總額、有無健康保險、保額高低等；需求因素（need factors），包括個人自評身體健康程度、疾病診斷、肢體失能程度、個人自評心理健康程度等。 此項研究的樣本是取自美國1994年「健康與退休調查」（Health and Retirement Study, HRS）。該調查在1993年以面談的方式，由訪問員藉由問卷調查親訪1萬2,652位年齡介於51歲到61歲的民眾。在問卷當中，民眾被問到有關他們的健康問題、退休狀況與經濟收入情形。經過剔除不合研究條件的樣本後，共有9,005筆記錄被使用到本研究中。 研究結果顯示，「需求因素」在預測醫療服務的使用上最具有顯著性。自評生、心理健康為「很差」的人比起其他答項的人更有可能使用醫療服務。在「使能因素」方面，有健康保險的人使用醫療服務的可能性是沒有健康保險者的1.4倍。有職業者使用醫療服務的可能性只有無業者的72%。在「傾向因素」方面，研究發現與過去的文獻相去不遠，亦即年齡的增加及較高的教育水準均和較多的醫療服務使用有正相關。與過去文獻不同的是，本研究發現「宗教」在預測醫療服務的使用上也有其重要性。基督教徒與天主教徒在使用醫療服務的可能性上分為是無宗教信仰者的1.3倍與1.4倍。本研究也發現了一些與過去文獻不同的結果：美國黑人使用醫療服務的可能性是白人的1.27倍（odds ratio=1.27, p<0.001）。由本研究的發現我們認為在設計相關的醫療政策時應該儘量以「需求」為導向，而非以年齡為標準，產生齊頭式的平等〔如美國的「醫療照護」方案（Medicare）〕。另外，少數族裔較可能使用醫療服務的發現是否與他們長期受到社會不平等所導致的健康不良有關，值得再進一步的研究。

Issues regarding health service utilization are one of the major concerns in the area of health service research. Each year, the U.S. government spends millions of dollars in financing health-related services. In spite of the efforts from the government to control the expenses, the escalating trend of health service expenditures does not seem to come to an end. Since so much money has been spent in this area, it is essential to understand how people use health services and to explore possible characteristics which associate with service use. Andersen’s Behavioral Model (ABM) has been applied widely in studying service utilization behaviors. The model specifies that a person’s decision to use health service is based on the combination of the following three factors: 1. the predisposing factors, which include demographic factors (age, gender, race, marital status, etc.,) social structural variables (education, religion, etc.,) and health beliefs, 2. the enabling factors which include household income, insurance coverage, etc., and 3. the need factors which include subjective health status and objective health status. Data for this study were drawn from the 1994 Health and Retirement Study (HRS), first wave. The survey interviewed 12,652 respondents in 1993 who were born between 1931 and 1941. Respondents were asked face-to-face questions about their health, retirement, and economic status. Nine thousand and five cases (9,005) were selected for this study. Results from this study show that the need factors were the most significant factors in predicting the occurrence of health service utilization behaviors. People who perceived their physical health and/or emotional health as “poor” were more likely to use services than those in other categories. In the enabling factors, those who were covered by health insurance were 1.4 times more likely to use health services than those who were not. Those who were employed had only 72% chance to use health services compared to those who were not employed. In the predisposing factors, similar to what previous studies have found, advancing age and higher educational levels were positively related to health service utilization. Religion played a role in predicting service use, a finding which was not studied before. Protestants and Catholics were 1.3 and 1.4 times more likely to use health services, respectively, compared to those who had no religious preferences. This study also found an interesting result which is different from previous studies: African Americans were more likely to use health services than Whites (odds ratio = 1.27, p<0.001.) The implication of this study calls for a need-based health policy rather than an age-based policy like Medicare. The disadvantageous health status of minorities also deserves more attention.