Feasibility of post-marketing active safety surveillance for pediatric medication based on big medical data: a survey on information systems in 17 children′s hospitals in China

Abstract

Objective To explore the feasibility of post-marketing active safety surveillance for pediatric medication based on big medical data. Methods A questionnaire on hospital information systems was conducted using cross-sectional study method in 17 children′s hospital of Beijing Children′s Hospital Group during May to July, 2016. Variables for post-marketing active safety surveillance for pediatric medication was designed in the questionnaire which included 146 questions of six aspects, database features, degree of computerization (including 21 variables), degree of structuring (involving 39 variables), data type, data encoding, degree of linkage on information systems, and data sharing willingness. The data were analyzed using Excel 2007 software. Results The recovery rate of the questionnaire was 100%. There were 17 children′s hospitals, of them, 14 hospitals were level 3 A and 3 hospitals were level 3 B, distributed in 17 provinces or regions in China. Data starts from 1999 to 2014 in the different hospitals′ information systems with a total of 60 million children′s medical records. The 21 variables used to evaluate the degree of computerization computerized in all departments in 10 hospitals and in half of the departments in 2 hospitals. Of the 39 variables used to evaluate the degree of structuring, 10 variables structured in the 17 hospitals, 23 variables structured in the 10-16 hospitals, and 6 variables structured in the 2-6 hospitals. However, there was a big difference in the data types and data coding among the information systems. The data encoding were mostly based on the hospital standard. Four modules of variable information (baseline information, drug exposure, diagnosis information, laboratory test) could be linked by a uniform number in 10 hospitals, partially linked in 7 hospitals. Meanwhile, 12 hospitals have the willing to share the desensitization data to support post-marketing drug safety studies. Conclusions Most children′s hospitals in this study have willing and conditions to integrate data of different modules of information systems to form the data chain. It is feasible to monitor actively the post-marketing safety of children′s drugs based on the big medical data. However, the medical data need to be further standardized firstly because of the big difference in the data structure and types among the medical systems. Key words: Hospital information systems; Child; Data collection; Safety medication