Data source profile reporting by studies that use routinely collected health data to explore the effects of drug treatment

BMC Medical Research Methodology

卷:23

Research

Kang Zou¹  Mei Liu²  Ling Li²  Xin Sun²  Mingqi Wang²  Qiao He²  Jiayue Xu²  Wen Wang²  Guowei Li³  Lin He⁴ 

[1] Chinese Evidence-based Medicine Center and Cochrane China Center, West China Hospital, Sichuan University, 37 Guo Xue Xiang, 610041, Chengdu, Sichuan, China;

[2] Chinese Evidence-based Medicine Center and Cochrane China Center, West China Hospital, Sichuan University, 37 Guo Xue Xiang, 610041, Chengdu, Sichuan, China;NMPA Key Laboratory for Real World Data Research and Evaluation in Hainan, Chengdu, China;Sichuan Center of Technology Innovation for Real World Data, Chengdu, China;

[3] Department of Health Research Methods, Evidence and Impact, McMaster University, L8S 4L8, Hamilton, ON, Canada;Center for Clinical Epidemiology and Methodology, Guangdong Second Provincial General Hospital, 510317, Guangzhou, Guangdong, China;Biostatistics Unit, Research Institute at St. Joseph’s Healthcare Hamilton, L8N 4A6, Hamilton, ON, Canada;

[4] Intelligence Library Center, West China Hospital, Sichuan University, 610041, Chengdu, Sichuan, China;

关键词: Routinely collected health data;    Drug treatment effect;    Reporting characteristics;   

DOI  :  10.1186/s12874-023-01922-8

 received in 2022-04-21, accepted in 2023-04-13,  发布年份 2023

来源: Springer

【 摘 要 】

BackgroundRoutinely collected health data (RCD) are important resource for exploring drug treatment effects. Adequate reporting of data source profiles may increase the credibility of evidence generated from these data. This study conducted a systematic literature review to evaluate the reporting characteristics of databases used by RCD studies to explore the effects of drug treatment.MethodsObservational studies published in 2018 that used RCD to explore the effects of drug treatment were identified by searching PubMed. We categorized eligible reports into two groups by journal impact factor (IF), including the top 5 general medical journals (NEJM, Lancet, JAMA, BMJ and JAMA Internal Medicine) and the other journals. The reporting characteristics of the databases used were described and compared between the two groups and between studies citing and not citing database references.ResultsA total of 222 studies were included, of which 53 (23.9%) reported that they applied data linkage, 202 (91.0%) reported the type of database, and 211 (95.0%) reported the coverage of the data source. Only 81 (36.5%) studies reported the timeframe of the database. Studies in high-impact journals were more likely to report that they applied data linkage (65.1% vs. 20.2%) and used electronic medical records (EMR) (73.7% vs. 30.0%) and national data sources (77.8% vs. 51.3%) than those published in other medical journals. There were 137/222 (61.7%) cited database references. Studies with database-specific citations had better reporting of the data sources and were more likely to publish in high-impact journals than those without (mean IF, 6.08 vs. 4.09).ConclusionsSome deficits were found in the reporting quality of databases in studies that used RCD to explore the effects of drug treatment. Studies citing database-specific references may provide detailed information regarding data source characteristics. The adoption of reporting guidelines and education on their use is urgently needed to promote transparency by research groups.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

40507_2023_167_Article_IEq533.gif

40507_2023_167_Article_IEq531.gif

Fig. 6

40507_2023_167_Article_IEq524.gif

13638_2023_2232_Article_IEq92.gif