探討加入糖尿病共照網之病患與未加入者之糖化血色素控制與健康行為改變是否有異，並檢視兩組病患之糖尿病相關健康行為改變，分別與何種健康信念之改變有關及其影響之程度。方法：採類實驗研究法，收案對象為第二型糖尿病病患，分成「共照組」與「非共照組」，依性別、年齡配對，以1:2比例分別收案208與421人，並於首次收案時及6個月後，進行糖化血色素與問卷資料收集。結果：自迴歸多變項分析顯示，共照組之後測糖化血色素較低(S=-0.18)，後測健康行為得分亦較高(S=0.13)。共照組後測健康行為與自覺利益(6=0.29)、行動線索(6=0.15)之提升具顯著關聯，非共照組後測健康行為則與自覺罹患性(6=0.13)、自覺利益 (6=0.19)之提升，以及自覺障礙(6=-0.16)、行動線索(6=-0.12)之下降具顯著關聯。結論：共照組病患之糖化血色素控制與健康行為改變均較非共照組佳，但兩組之健康行為改變係由不同健康信念所驅動，故未來糖尿病健康照護介入，可參考本研究之發現，於兩組分別針對不同之健康信念設計合適之介入策略。（台灣衛誌2013 ； 32(6) ： 537-550)

Objectives: The aims of this study were to determine if patients in a diabetes shared care group had better glycated hemoglobin (HbA1c) control and changes in health behaviors than did their regular care counterparts, and to explore which changes in health beliefs contributed to changes in health behaviors in these two groups. Methods: A quasi-experiment, this study recruited 208 and 421 patients with type II diabetes in the shared care and regular care groups, respectively, at a 1:2 ratio, matched by gender and age. HbA1c levels were measured and survey data about diabetes-related health behaviors and health beliefs were collected at both pre-test and 6-month post-test. Results: Multivariate linear autoregression analysis indicated that patients in the shared care group had significantly lower HbA1c levels (p=-0.18) than did those in the regular care group at post-test. Similarly, patients in the shared care group had significantly higher health behavior scores (p=0.13) than did their regular care counterparts at post-test. In the shared care group, increased health behavior scores at post-test were directly related to elevated levels of perceived benefits (p=0.29) and cues to action (p=0.15). By contrast, in the regular care group, increased health behavior scores at post-test were positively associated with increased levels of perceived susceptibility (p=0.13) and perceived benefits (p=0.19), as well as decreased levels of perceived barriers (p=-0.16) and cues to action (p=-0.12). Conclusions: Compared with diabetic patients in the regular care group, those in the shared care group had better HbA1c control and changes in health behaviors. The diabetes-related health behavior changes could be attributed to different changes in health beliefs in the two groups. These findings could inform future interventions to improve diabetes care. (Taiwan J Public Health. 2013;32(6):537-550)