本研究將利用國民健康局的老人健康與生活狀況調查之 1989、1993、
1996、1999 及2001 年五期追蹤調查資料，探討老人之醫療服務利用行為。
主要研究方法是將系統化的非正式社會支持體系(戶基護航網絡Household-base
Convoy)引進到Anderson 的醫療服務利用模型中，以分析老人的醫療服務利用行為。其
目的有：
1、政策意涵：
本研究將利用 Anderson 的醫療服務利用模型分析三種老人醫療服務利用行為，包
括用藥規則性、健檢之利用及就醫次數。將非正式社會支持體系加入到Anderson 模型，
可以達到控制其他變數來突顯或淨化非正式社會支持系統成員對上述三種老人醫療服
務利用行為之影響。衛生單位可以透過最有影響力的非正式社會支持成員來促進老人之
醫療服務利用行為。
2、研究方法探討：
本研究將“戶基護航網絡”用來代表非正式社會支持體系，並加入到Anderson 模型中
可檢討模型解釋力是否會增加。
3、理論發展：
理論上，戶基護航網絡是一個動態觀念，網絡成員會隨生命過程(Life course)而改
變。利用國健局的老人健康與生活狀況貫時性調查資料，我們可以檢驗老人之戶基護航
網絡是否隨時間而改變。這種改變會不會影響到老人之醫療服務利用行為。
Abstract
In this study, the longitudinal survey data on” The health and living status of the elderly
in Taiwan” collected by Bureau of Health Promotion in 1989,1993, 1996, 1999, and 2003 will
be used to investigate the health utilization behaviors of the elderly in Taiwan. The
investigation relies on introducing a concept “household-based convoy” into the Andersen’s
health utilization model to achieve the following three purposes:
1. Policy implications
In this study, Andersen model will be used to analyze three types of health
utilization behaviors including regularity of medicine taking, the use of overall health
examination, and numbers of clinic visit. The concept of household-based convoy is added to
Andersen model in order to reflect an informal support system. We suppose that the inclusion
of the convoy concept will enable us to point out which member of an informal support
system is most influential in increasing the elderly’ health utilization behaviors. Health
department may use the information about informal social support to promote the health
status of the elderly.
2. Methodological implications
Andersen model has been criticized for having a relatively low explaining ability.
In this study, we would like to know whether the household-based convoy has the ability to
increase the model’s explanation power.
3. Theoretical implications
Theoretically, the household-based convoy is assumed to be a dynamic concept. Its
members are supposed to be different in different life-course stages. Using the data mentioned
above, we are enabled to examine whether the household-based convoy of the elderly will
change in life-course stages. Moreover, it is also interesting to examine whether the change
may affect the health utilization behaviors of the elderly.
