Article
Data extraction in systematic reviews: a survey on current practice, methods and research priorities
Search Medline for
Authors
-
Roland Brian Büchter
- Witten/Herdecke University, Institute for Research in Operative Medicine (IFOM), Deutschland
-
Tanja Rombey
- Technische Universität Berlin, Department of Health Care Management, Berlin, Deutschland
-
Tim Mathes
- University Medical Center Göttingen, Institute for Medical Statistics, Göttingen, Deutschland
-
Hanan Khalil
- La Trobe University, School of Psychology and Public Health, Department of Public Health, Australien
-
Carole Lunny
- St. Michael’s Hospital, Unity Health Toronto, Li Ka Shing Knowledge Institute, Toronto, Kanada
-
Danielle Pollock
- University of Adelaide, Health Evidence Synthesis, Recommendations and Impact (HESRI), School of Public Health, Faculty of Health and Medical Sciences, Adelaide, Australien
-
Livia Puljak
- Catholic University of Croatia, Center for Evidence-Based Medicine and Healthcare, Kroatien
-
Andrea C. Tricco
- St. Michael’s Hospital, Unity Health Toronto, Li Ka Shing Knowledge Institute, Toronto, Kanada; University of Toronto, Epidemiology Division and Institute of Health Policy, Management, and Evaluation, Dalla Lana School of Public Health, Toronto, Kanada
-
Dawid Pieper
- Brandenburg Medical School (Theodor Fontane), Institute for Health Services and Health System Research, Deutschland; Brandenburg Medical School (Theodor Fontane), Center for Health Services Research, Deutschland
| Published: | March 12, 2024 |
|---|
Outline
Text
Background/research question: Data extraction in systematic reviews (SRs) involves taking information from primary studies for further refinement, analysis and synthesis, typically using dedicated data extraction forms. Carefully designed and piloted extraction forms reduce errors, ensure relevance and transparency, and provide a reliable historical record to support future updates. Previous research shows that data extraction errors are common and take many forms [1]. However, much of the guidance on conducting SRs provides incomplete advice on how to approach data extraction – and little is known about current practice [2]. Comparative research on methods is also limited [3]. We aimed to conduct a survey of systematic review authors to understand current data extraction practice and identify further research needs.
Methods: We developed and piloted an international online survey and distributed it through relevant organisations, social media and personal networks in 2022. The survey consisted of 27 closed questions and 2 open-ended questions. Closed questions were analysed using descriptive statistics and open questions were analysed using content analysis.
Results: 142 respondents completed the survey. The use of adapted (65%) or newly developed extraction forms (62%) was common. Generic forms were rarely used (14%). Spreadsheet software was the most popular extraction tool (83%). Methodologists, content experts and statisticians were considered most important in developing and piloting extraction forms, and statisticians were not often involved (28%). Pilot testing the forms with a few studies prior to full data extraction was common (74%). Independent and duplicate extraction was considered the most trustworthy approach to minimise errors (64%). About half of the respondents agreed that blank forms and/or raw data should be publicly accessible. The impact of different methods on error rates (60%) and the use of tools to support data extraction (46%) were identified as research gaps.
Conclusion: Systematic reviewers used varying approaches to data extraction. Methods to reduce errors and the use of support tools such as (semi-)automation tools are key research gaps. The survey provides useful insights that can be used to develop and advance future guidance for data extraction in SRs.
Competing interests: We have no conflicts of interest to disclose.
References
- 1.
- Mathes T, Klaßen P, Pieper D. Frequency of data extraction errors and methods to increase data extraction quality: a methodological review. BMC Med Res Methodol 2017;17:152.
- 2.
- Büchter RB, Weise A, Pieper D. Reporting of methods to prepare, pilot and perform data extraction in systematic reviews: analysis of a sample of 152 Cochrane and non-Cochrane reviews. BMC Med Res Methodol. 2021;21:240.
- 3.
- Robson RC, Pham B, Hwee J, et al. Few studies exist examining methods for selecting studies, abstracting data, and appraising quality in a systematic review. J Clin Epidemiol. 2019;106:121-135.
