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

Wen Wang; Mei Liu; Qiao He; Mingqi Wang; Jiayue Xu; Ling Li; Guowei Li; Lin He; Kang Zou; Xin Sun

doi:10.1186/s12874-023-01922-8

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

BMC Med Res Methodol. 2023 Apr 20;23(1):95. doi: 10.1186/s12874-023-01922-8.

Authors

Wen Wang^{1

2

3}, Mei Liu^{1

2

3}, Qiao He^{1

2

3}, Mingqi Wang^{1

2

3}, Jiayue Xu^{1

2

3}, Ling Li^{1

2

3}, Guowei Li^{4

5

6}, Lin He⁷, Kang Zou¹, Xin Sun^{8

9

10}

Affiliations

¹ Chinese Evidence-based Medicine Center and Cochrane China Center, West China Hospital, Sichuan University, 37 Guo Xue Xiang, Chengdu, 610041, Sichuan, China.
² NMPA Key Laboratory for Real World Data Research and Evaluation in Hainan, Chengdu, China.
³ Sichuan Center of Technology Innovation for Real World Data, Chengdu, China.
⁴ Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, ON, L8S 4L8, Canada.
⁵ Center for Clinical Epidemiology and Methodology, Guangdong Second Provincial General Hospital, Guangzhou, 510317, Guangdong, China.
⁶ Biostatistics Unit, Research Institute at St. Joseph's Healthcare Hamilton, Hamilton, ON, L8N 4A6, Canada.
⁷ Intelligence Library Center, West China Hospital, Sichuan University, Chengdu, 610041, Sichuan, China.
⁸ Chinese Evidence-based Medicine Center and Cochrane China Center, West China Hospital, Sichuan University, 37 Guo Xue Xiang, Chengdu, 610041, Sichuan, China. sunxin@wchscu.cn.
⁹ NMPA Key Laboratory for Real World Data Research and Evaluation in Hainan, Chengdu, China. sunxin@wchscu.cn.
¹⁰ Sichuan Center of Technology Innovation for Real World Data, Chengdu, China. sunxin@wchscu.cn.

Abstract

Background: Routinely 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.

Methods: Observational 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.

Results: A 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).

Conclusions: Some 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.

Keywords: Drug treatment effect; Reporting characteristics; Routinely collected health data.

Publication types

Systematic Review
Research Support, Non-U.S. Gov't

MeSH terms

Humans
Information Sources*
Journal Impact Factor
Observational Studies as Topic
PubMed
Research Design
Routinely Collected Health Data*