Meta-analysis is a cornerstone of evidence-based research, offering a systematic approach to combine and synthesize data from multiple studies. However, as research questions become more nuanced and datasets more diverse, the complexity of meta-analyses increases significantly. This is where the Data Extraction for Complex Meta-Analysis (DECiMAL) guide comes into play, providing a structured framework to navigate these complexities.[1]
Extracting data for meta-analysis can be a complex task, especially when dealing with diverse study designs, outcomes, and data formats. Traditional data extraction methods may not be sufficient to handle the complexities of modern research, leading to potential biases and inconsistencies in the analysis. The DECiMAL guide addresses these challenges by offering a detailed methodology for data extraction, ensuring all relevant information is captured and analyzed consistently. This promotes standardization, reduces bias, and enhances transparency in the meta-analysis process, ultimately leading to more reliable and informative results. DECiMAL covers a wide range of data types, including continuous, binary, and time-to-event outcomes, as well as more complex data structures, such as multiple treatment arms and correlated outcomes. By addressing these complexities, DECiMAL helps researchers conduct rigorous and reproducible meta-analyses.[1-3]
The DECiMAL guide comprises several core components designed to tackle the complexities of meta-analyses. First, it stresses the importance of a clearly defined research question, utilizing the Population, Intervention, Comparison, and Outcome (PICO) criteria to guide the data extraction process. A comprehensive literature search is essential, and DECiMAL advocates for a systematic approach across various databases to capture all relevant studies while minimizing publication bias, with meticulous documentation of the search strategy. The guide’s detailed data extraction template captures a broad range of data points, ensuring consistency and completeness. Addressing heterogeneity is another key aspect, with DECiMAL offering guidance on statistical methods like subgroup analyses and meta-regression to understand variability between studies. For data synthesis and analysis, DECiMAL provides best practices, including the use of fixed-effect and random-effects models and a multivariate approach for diagnostic accuracy studies. It also emphasizes the assessment of bias using standardized tools and advocates for transparent reporting according to guidelines, such as PRISMA, which supports replication and enhances research credibility.[4]
The DECiMAL guide provides a detailed methodology for extracting various types of data for meta-analysis. For time-to-event data, such as cancer recurrence, hazard ratios and their uncertainties should be collected, and it should be noted if Kaplan-Meier plots or life tables are reported. For rate data, like migraine episode frequency, the total number of person-years at risk should be collected. If this information is not available, the average length of follow-up and the total number of patients at study end can be used to approximate person-years. Binary and categorical variables should use numerical coding for responses and additional coding for other responses, while both numbers of patients randomized and those completing the trial should be extracted. Continuous and ordinal variables should be consistently reported in chosen units, with both final values and changes from the baseline being combined if baselines are equal. The guide ensures comprehensive data collection and helps identify and address potential issues early on, enhancing the consistency and accuracy of complex meta-analyses.[4]
While DECiMAL provides a comprehensive framework for data extraction, it has certain limitations. For instance, it does not delve into specific statistical techniques for handling missing data or converting summary statistics. Additionally, while DECiMAL is primarily designed for aggregate data meta-analyses, it may not be directly applicable to individual patient data meta-analyses. The guide primarily addresses considerations related to data extraction for subsequent meta-analyses but provides limited information on the practical and technical aspects of data extraction itself. Furthermore, DECiMAL is designed specifically for data extraction in aggregate data meta-analyses, and its methods do not apply to individual patient data meta-analyses.[4]
The DECiMAL guide marks significant progress in meta-analysis, especially for handling complex datasets. By standardizing data extraction, addressing heterogeneity, and enhancing transparency, DECiMAL ensures that meta-analytical results are robust and reliable. With the increasing volume and complexity of research data, adopting comprehensive tools like DECiMAL will be essential for preserving the integrity and effectiveness of meta-analyses. For researchers undertaking complex meta-analyses, DECiMAL provides a structured approach to navigating the challenges of data extraction and analysis. Following its guidelines can improve the quality and impact of research findings, offering valuable contributions to the scientific community.
Become A Certified HEOR Professional – Enrol yourself here!
References:
- Brown SA, Upchurch SL, Acton GJ. A framework for developing a coding scheme for meta-analysis. West J of Nurs Res. 2003;25:205–22.
- Centre for Evidence-Based Medicine. Data Extraction Tips: Meta-Analysis [Internet]. Oxford: University of Oxford. 2023; Available from: https://www.cebm.ox.ac.uk/resources/data-extraction-tips-meta-analysis.
- Effective Practice and Organisation of Care (EPOC). Data collection form. EPOC Resources for review authors. Norwegian Knowledge Centre for the Health Services. 2013;Available from: http://epoc.cochrane.org/epoc-specific-resources-review-authors.
- Pedder H, Sarri G, Keeney E, Nunes V, Dias S. Data extraction for complex meta-analysis (DECiMAL) guide. Syst Rev. 2016 Dec 13;5(1):212. 5. Afifi M, Stryhn H, Sanchez J. Data extraction and comparison for complex systematic reviews: a step-by-step guideline and an implementation example using open-source software. Syst Rev. 2023 Dec 1;12(1):226.