Exploring the mediating effect of process of care on the volume-outcome relationship for diabetes care: An example in Pay-For-Performance

• Main Authors: Chia-Chi Wei, 魏嘉琦
• Other Authors: Kuo-Piao Chung
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/51681074637891056733

碩士 === 國立臺灣大學 === 健康政策與管理研究所 === 99 === Background: Over the past few years, the relationship of volume and outcome has been widely discussed in many medical conditions, especially in surgical procedures. A lot of studies support that higher volume is associated with better outcome in a number of medical conditions. However, researches has not clearly identified whether a similar association exists for chronic medical condition, such as diabetes care, whether volume also has an important implication is unclear. And past studies have also considered that the process of care may affect the relationship of volume and outcome. But it still cannot explain the meaning and relationship between volume, quality of process and outcome. Objective: First, this study sought to identify the relationship of physician volume with quality of process and outcome in diabetes care, for Pay-for-Performance Payment Program’s patients cared for by physicians. Second, by using an mediation analysis, this study investigated the volume, process of care and outcome which respectively represent the three dimensions of quality of care─structure, process and outcome, to identify the mechanism underlying the association between volume, quality of process and outcome. Finding out whether quality of process may mediate the relationship of volume and outcome, attempting to clarify the reason why volume is associated with outcome. Methods: In this cross-sectional study, select the diabetes care in Pay-for-Performance Payment Program as an example and used secondary database from Taiwan’s National Health Insurance Research Database and VPN in 2008. Physicians’ patient volume was calculated based on the number of diabetes patients in Pay-for-Performance Payment Program. Physicians’ outcome performance was the rate of patients whose HbA1C average value was under control among all of the patients. Physicians’ process of care performance was the rate of patients who achieve the performance goal (check HbA1C value twice per year) among all of the patients. Using multiple regression analysis to verify the relationship of physician volume with process of care and outcome. Furthermore, using mediation analysis to verify whether process of care is an mediating variable in the relationship of volume and outcome. Results: In multiple regression analysis, after adjustment for physician characteristics and patient characteristics, physician volume was positively related to outcome and process of care, and process of care would mediates the relationship of volume and outcome, and it’s a complete mediation effect. In fact, the impact of volume on outcome was through quality of process care, that is, the impact of volume on outcome was indirectly, caused by the process of care. And this mediation effect was moderated by patients’ complications. Conclusions: This study suggests that, for diabetes care in Pay-for-Performance Payment Program, physicians’ patient volume would affect outcome, but this association is because of process of care performance. Based on this verification, cared for by physicians or hospital with greater numbers of diabetes patients in Pay-for-Performance Payment Program may obtain better process of care, then obtain better outcome. By mediation analysis, explained how the impact of volume on outcome, and highlighted the important of process of care. However, emphasizing volume was insufficient, because the implication of volume is too wide. Thus, emphasizing process of care was a more important concept, and perhaps such a relationship and phenomenon for some medical conditions which more emphasized on the process of care, was the most important point.