Design And Implementation Of Supervisor System Of Municipal Hospital’s Shared Data Quality

In recent years, cross-hospital medical info sharing draws much attention from medical reform and digitalization. Medical information sharing throws significant impact on governmental administration on medical sector, integration of clinical and prophylactic, clinic research, plus patient’s self-care. Shanghai Medical Interlink Project (SMIP), dedicated to municipal hospital’s data sharing, is established to fulfill above-mentioned objectives. This thesis will focus on SMIP, design and implementation of supervisor system of municipal hospital’s shared data quality, so as to ensure SMIP’s smooth running as a platform of medical info sharing.It is unrealistic to manually QC medical data which flows into SMIP at the level of GB scale on daily basis. That’s why we need to establish a systematic mechanism to analyze, evaluate and feedback for the quality of medical data sharing, set up correct medical archive patients’diagnosis and treatment, minimize the possibility of missed and wrong reporting, track the quality of uploaded data in real-time mode, and feedback the system performing status to hospital management, info division and HIS vendor. This system features with the following:a relatively complete and reasonable data analysis and evaluation KPIs, well-defined all indexes regarding description, specification and calculation methodology; high precision to field level of database for all data; automatically evaluation of data upload ratio based on the internal logics between table and data; providing visualized presentation of all potential issues which are shared via internet, forwarding key data quality issue to decision makers via Email and SMS. This system contains3modules:definition of data quality evaluation, QC of data quality at both front-end and server-end, feedback platform of data quality.As validated by application system which operates every day, system designed based on this thesis is capable of quality control (identifying, locating and reporting potential flaw of media-care sharing data), and replying to all planned management requirements.