| The growing momentum towards hospital quality measurement and public reporting of hospital performance has presented hospitals with a dramatically changing environment. An immediate consequence of these changes has been the rise in importance of coding accuracy in hospital administrative data. Existing literature, however, indicates considerable variation in hospital coding accuracy across the nation. Studies investigating causes of coding inaccuracy have revealed poor physician documentation in the medical record to be the primary explanatory factor. While there has been limited empirical research into reasons for poor documentation, federal audits into the issue have suggested that the problem might be owing to a lack of knowledge exchange among hospital administrators, physicians, and coders, in regard to how documentation and coding can impact organizational outcomes such as comparative quality rankings.;This study explores the importance of organizational knowledge exchange as a factor impacting hospital-coding performance, and, to this effect, develops a theoretical framework for knowledge sharing network effectiveness in professional complex systems. The essence of this framework is that knowledge exchange in professional complex systems may be most effective when there is integration through centralization, hierarchy, and brokerage (with leaders connecting different professional subgroups to each other and to the external environment), rather than through density (where everyone is directly connected to everyone else). This framework was used to design an exploratory and comparative field study of knowledge exchange among hospital administrators, physicians, and coders in a changing environment of hospital quality measurement and reporting. The study involved conducting individual interviews with the quality administrator, physician administrator, and coding administrator in four hospitals, two showing good coding performance and two showing poor coding performance in the context of hospital quality measurement. Additionally, an online survey was conducted with staff and managers in the quality, coding, and medical staff subgroups in the same four facilities.;The study found that good coding performance was systematically associated with broad and conceptual knowledge exchange content, wherein leaders played a proactive role in providing ongoing education regarding the changing environment and the role of documentation/coding in hospital quality measurement. From a network structure perspective, good performance was associated with networks rich in centralization, hierarchy, and brokerage (with little or no density). In both "good coding" facilities, a core group of senior administrators played a key role in coordinating knowledge exchange across subgroups and between subgroups and the external environment. By contrast, "poor coding" facilities showed minimal interaction and integration. Additionally, the study identified various leadership/management and organizational characteristics that were systematically associated with coding performance. The study concludes with a series of recommendations for public managers and policymakers. |
