17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Received: 23 April 2021 Revised: 3 February 2022 Accepted: 26 February 2022 DOI: 10.1002/jrsm.1557 SOFTWARE FOCUS Graphical Representation of Overlap for OVErviews: GROOVE tool Javier Pérez-Bracchiglione1,2,3 s Meza1 | Shrikant I. Bangdiwala4 | | Nicola Ena Niño de Guzma n2 | Gerard Urrútia2,5 | Xavier Bonfill2,5 | Eva Madrid1 1 Interdisciplinary Centre for Health Studies (CIESAL), Universidad de Abstract Valparaíso, Viña del Mar, Chile Overlap of primary studies among systematic reviews (SRs) is one of the main meth- 2 Iberoamerican Cochrane Centre, Sant odological challenges when conducting overviews. If not assessed properly, over- Pau Biomedical Research Institute lapped primary studies may mislead findings, since they may have a major influence (IIB-Sant Pau), Barcelona, Spain 3 Departament of Pediatrics, Obstetrics and either in qualitative analyses or in statistical weight. Moreover, overlapping SRs may Gynecology, Preventive Medicine and represent the existence of duplicated efforts. Matrices of evidence and the calculation Public Health, Universitat Autònoma de of the overall corrected covered area (CCA) are appropriate methods to address this Barcelona, Barcelona, Spain 4 issue, but they seem to be not comprehensive enough. In this article we present Department of Health Research Methods, Evidence, and Impact, McMaster Graphical Representation of Overlap for OVErviews (GROOVE), an easy-to-use tool University, Hamilton, Ontario, Canada for overview authors. Starting from a matrix of evidence, GROOVE provides the 5 Clinical Epidemiology Service, Hospital number of included primary studies and SRs included in the matrix; the absolute de Sant Pau, CIBER Epidemiología y Salud Pública (CIBERESP), Barcelona, number of overlapped and non-overlapped primary studies; and an overall CCA Spain assessment. The tool also provides a detailed CCA assessment for each possible pair of SRs (or “nodes”), with a graphical and easy-to-read representation of these results. Correspondence Javier Pérez-Bracchiglione, Angamos 655, Additionally, it includes an advanced optional usage, incorporating structural mis- Viña del Mar, Chile. singness in the matrix. In this article, we show the details about how to use Email: javier.perez.b@gmail.com GROOVE, what results it achieves and how the tool obtains these results. GROOVE Funding information is intended to improve the overlap assessment by making it easier, faster, and more JP-B was funded by a FONDECYT Grant friendly for both authors and readers. The tool is freely available at http://doi.org/10. 1212037 from the Chilean National 17605/OSF.IO/U2MS4 and https://es.cochrane.org/es/groovetool Agency of Research and Development (ANID). ENDG was funded by a Rio KEYWORDS Hortega contract from the Instituto de Salud Carlos III [CM17/00199]. GU corrected covered area, overlap, overviews of systematic reviews, systematic reviews as contribution to this paper was supported topic by a CONICYT grant (MEC80170060), government of Chile. Highlights • The development of overviews of reviews has brought new methodological challenges, such as the assessment of overlap of primary studies among sys- tematic reviews. • Current methods and guidance for addressing overlap still have important limitations, especially when assessing a large body of evidence. • This new and freely-available tool named GROOVE provides a simpler and more comprehensive way for assessing the overlap of primary studies among systematic reviews included in an overview. This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2022 The Authors. Research Synthesis Methods published by John Wiley & Sons Ltd. Res Syn Meth. 2022;13:381–388. wileyonlinelibrary.com/journal/jrsm 381 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 382  PEREZ-BRACCHIGLIONE ET AL. 1 | INTRODUCTION 2 | M E T HO DS AND I SSUE S W H EN ASSESSING OVERLAP The increasing number of systematic reviews (SRs) publi- shed during the last years has led to the development of There are several ways to deal with overlap at different new methods for broader evidence synthesis, such as stages during the development of the overview.8–11 overviews.1–4 An overview can be defined as a SR of SRs Overview authors may decide about eligibility, data instead of primary studies, and constitutes a relatively extraction, risk of bias assessment and synthesis/ new approach for broad evidence synthesis.4,5 Neverthe- presentation of the results, based on the overlap of pri- less, some methodological challenges have arisen with mary studies among the included SRs. Indeed, a variety the development of this research design, with one of the of tools has been proposed for guiding these deci- most important being the assessment of overlap of pri- sions.5,12–14 Nevertheless, currently, most authors of mary studies among SRs. overviews do not report any method for handling over- The overlap of primary studies among SRs (also lap.15 Although some methods for assessing overlap named as overlapping SRs) refers to the double counting have been proposed, these are not widely used, and of the same primary study included in at least two there is considerable heterogeneity among guidelines (or more) SRs. Overviews including overlapping SRs are considering this issue.11,16,17 complex and, if not assessed properly, overlapped pri- Matrices of evidence and the calculation of the mary studies could have a major influence either in qual- corrected covered area (CCA) are probably among the itative analyses or in statistical weight,5 which could lead most exhaustive methods for measuring overlap.10,17,18 to misleading findings due to the undue influence of the Matrices of evidence are grids that link SRs with their results of the studies with greater overlap in the global included primary studies. This method is very useful assessment of the effect of the intervention of interest. when the analysed body of evidence is relatively small For example, if an overview does not correctly assess the since it displays a visual image that allows to capture overlap of primary studies, it could wrongly obtain overly (sensory impression) the overlap of studies between the precise estimates of the effect of certain interventions.6 different SRs that are being analysed within the overview. Moreover, neglecting the analysis of overlap among SRs On the other hand, CCA is a formula that quantitatively may disregard the opportunity of identifying duplicated assesses the overlap degree, gathering its variables from a efforts that may constitute a research waste phenome- matrix of evidence (Figure 1). According to the original non, with implications ranging from the existence of publication, a CCA of 0%–5% represents a slight overlap, redundant findings, to the misuse of funds and 6%–10% a moderate overlap, 11%–15% a high overlap, resources.7 and above 15% a very high overlap.18 F I G U R E 1 Matrix of evidence and calculation of corrected covered area (CCA) formula.18 Ticked cells represent a primary study included in the respective SR [Colour figure can be viewed at wileyonlinelibrary.com] 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License  PEREZ-BRACCHIGLIONE ET AL. 383 F I G U R E 2 Rationale of GROOVE. The tool calculates the corrected covered area (CCA) for the overall matrix, and it also calculates the CCA for each possible pair of reviews. In this example, it is possible to find a specific pair of reviews (1 and 4) that have no overlap (CCA = 0.0%), in a context of very high overall overlap (overall CCA = 22.2%). The tool provides the result of the CCA for each possible pair of reviews, but does not visually display the specific matrices [Colour figure can be viewed at wileyonlinelibrary.com] Despite being appropriate, both of the described CCA for a whole matrix of evidence, and, at the same time, methods have limitations. First, a matrix of evidence for each possible pair of SRs included in the overview. In becomes harder to interpret as the included body of evi- other words, GROOVE calculates the CCA that a matrix dence increases,11 since it depends on a visual impression would have if only two SRs were included into it, and that becomes more difficult to capture when the sheet of repeats this process for every possible pair of SRs (Figure 2). studies and reviews is very extensive. On the other hand, This approach allows detecting highly overlapped pairs of CCA provides an assessment for a whole matrix of evi- SRs in a context of low-overall overlap, as well as non-over- dence. It is possible for two or more specific SRs to be lapped pairs of SRs in a context of high-overall overlap, dis- highly overlapped among each other in a low overall CCA playing these results into an easy-to-read figure. scenario, as well as not being overlapped in a high overall GROOVE also provides an advanced optional new fea- CCA scenario. In this article, we present a new tool to ture: the consideration of structural missingness address these limitations. (or structural zeros) when calculating the CCA. Structural missingness is operationally defined as an intersection in the matrix of evidence that cannot acquire a value other 3 | INTRODU CI N G G ROOVE: than “0,” For example, a SR published in 2015 cannot GR APHIC AL REPRESENTATION O F include a primary study published in 2018 (i.e., the value OVERLAP FOR OVERVIEWS of this intersection cannot be other than “0”), therefore, this is a chronological structural missingness. Structural In order to assess in a simpler and more comprehensive missingness is a topic not yet described in the assessment way the overlap of primary studies among SRs included in of overlap of primary studies among SRs, but we think this an overview, we developed the GROOVE (Graphical Repre- new feature could provide a complementary approach to sentation of Overlap for OVErviews) tool. This tool is an overlap assessment. The GROOVE tool supports the inclu- Excel-based file which automatically calculates the overall sion of structural missingness at the moment of building 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 384  PEREZ-BRACCHIGLIONE ET AL. F I G U R E 3 Example of GROOVE's matrix of evidence filled for a previously published overview.19 Green cells marked with a “1” represent a primary study included in a SR, while the black cells marked with an “X” represent a structural zero. The image has been cropped [Colour figure can be viewed at wileyonlinelibrary.com] the matrix and when calculating the results, although this not have any influence on the final results. If the user is an optional feature intended to calculate the proportion wants to consider structural missingness, an “X” must over the real denominator. be annotated in all the intersections that are consid- ered structural zeros. Figure 3 shows an example of GROOVE's matrix of evidence fulfilled for a previously 3.1 | How to use GROOVE? published overview.19 • “2.1 GROOVE - colour version”: Once the user enters GROOVE is a downloadable tool developed in Excel. It all the information in the “1. Matrix” sheet, the tool has four sheets: will provide the results of the overlap analysis in this sheet. The results provided in this sheet are colour- • “0. Start here”: It explains briefly what GROOVE is, blind friendly.20 how to use the tool and the optional advanced usage • “2.2 GROOVE - greyscale version”: This tool is identi- considering structural missingness. cal to the “2.1 GROOVE - colour version”, except it • “1. Matrix”: This is the only sheet that the user directly provides the results in greyscale. manipulates when entering data. It provides an empty matrix of evidence, supporting up to 70 SRs and 1000 unique primary studies. The user must fill the SRs ID 3.2 | What results does GROOVE in the columns (light blue cells starting from C5), and achieve? the primary study ID in the rows (light red cells starting from A6). Afterwards, the user must mark All the results will be provided in the “2.1 GROOVE - col- with a “1” each intersection in the matrix representing our version” and the “2.2 GROOVE - greyscale version” a primary study included in its respective SR. There is sheets. The results are calculated automatically as the also space for annotating the topic of the matrix and matrix is filled, and are included in three sections within the reference of the primary studies, although this does these sheets: 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License  PEREZ-BRACCHIGLIONE ET AL. 385 F I G U R E 4 Overall results as displayed by the GROOVE tool. These results include the components of the CCA formula, the calculation of the CCA, the calculation of CCA adjusted by structural zeros, and the absolute number of overlapped and non- overlapped primary studies. This example was elaborated using the data from a previously published overview.19 The image has been cropped [Colour figure can be viewed at wileyonlinelibrary.com] • “Overall results”: This section provides the overall name to our tool. The tool provides a calculation of a overlap assessment for the whole matrix of evidence. CCA for each possible pair of SRs or “nodes”, dis- The tool summarises the number of reviews, index playing these results in a figure resembling a leaning publications and primary studies (including double triangle (Figure 5). Each node shows the CCA for two counting) included in the matrix. With this data, it cal- specific SRs. In other words, it shows what the result culates the covered area, the CCA, and provides the of the CCA would be if, instead of the whole matrix, interpretation of the overall overlap assessment, being we considered a matrix including only those two spe- slight if the CCA is <5%, moderate if it is ≥5% cific SRs. The figure also highlights with different col- and < 10%, high if it is ≥10% and < 15%, and very high ours each node according to their interpretation: In if CCA is ≥15%. If structural missingness was consid- the “2.1 GROOVE - colour version”, the nodes are ered when filling the matrix, the tool will also provide white if CCA is low, bluish green if it is moderate, yel- a CCA adjusted by structural zeros. If structural mis- low if it is high, and vermillion if it is very high. All singness was not considered when filling the matrix, these colours are colour-blind friendly.21 The “2.2 these results will be equal to the CCA not adjusted by GROOVE - greyscale version” displays the same structural missingness. The tool also indicates the results, with white being used for nodes with low number of unique primary studies (i.e., included in CCA, light grey for nodes with moderate CCA, dark only one SR), as well as the absolute number of pri- grey for nodes with high CCA, and black for nodes mary studies included in two or more SRs (Figure 4). with very high CCA. The thresholds for this inter- pretation are the same as the ones mentioned previ- ously. This figure also displays the total number of • “Graphical Representation of Overlap for OVErviews nodes that have low, moderate, high or very high (GROOVE)”: This graphical representation gives the overlap. 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 386  PEREZ-BRACCHIGLIONE ET AL. F I G U R E 5 Graphical representation of the results as presented by the GROOVE tool (colour version). The intersections or nodes in this figure represent every possible pair of SRs. For each node, the tool calculates the CCA for that specific pair of SRs. The results are highlighted in colours. The tool also provides the total number of nodes with slight, moderate, high and very high overlap. This example was elaborated using the data from a previously published overview19 [Colour figure can be viewed at wileyonlinelibrary.com] • “Graphical Representation of Overlap for OVErviews each row for each possible pair of SRs. If the result of (GROOVE) - Considering Structural Missingness”: If the sum is “2” for any pair of SRs, this means that the the user considered structural missingness when mak- primary study is overlapped. Then, the tool counts all ing the matrix (that is, if he/she annotated any “X” in the “2” in the respective columns. This count repre- the matrix), this figure will show the results of the sents the total number of overlapped primary studies assessments of CCA for each node, but with the CCA for a specific node. formula adjusted by structural missingness. This figure • Total number of rows (index publications, “r”): In this has the same visual interface as the one mentioned hidden leaning triangle, the tool calculates the number above. If no “X” were annotated in the matrix, the of rows which there would be in a matrix with two spe- results will be the same as those shown in the cific SRs. This is directly done by subtracting the total GROOVE not considering structural missingness. number of overlapped primary studies to the total number of studies (“N”). • Total number of structural missingness (“X”): In this 3.3 | How does GROOVE calculate these hidden leaning triangle, the tool calculates the number results? of structural zeros or “X” in the matrix for each node. To do this, the “1. Matrix” sheet counts, for each possi- The results are calculated based on numerous but simple ble pair of reviews, all the rows where one SR includes formulas in hidden columns and rows. In the “2.X the study (that is, an intersection with a “1”) and the GROOVE” sheets, there are four figures with the same other states a structural zero (that is, an intersection leaning triangle structure as the GROOVE. For each pos- with an “X”). The leaning triangle copies the count of sible node, these figures calculate the following: cases where this happens. • Total number of studies (including duplicates, “N”): With this information, GROOVE (and GROOVE The first of the hidden leaning triangle figures calcu- adjusted by structural missingness) has all the necessary lates the total number of studies for each node. To do data to calculate the detailed CCA. There is no need to this, the “1. Matrix” sheet sums all the “1” included in calculate the number of rows (“r”), as the assessment is each column. The result for each intersection represents presented for each pair of SRs, therefore, “r” is equal to 2. the sum of the results of the two corresponding columns. • Total number of overlapped primary studies: In this 4 | DISCUSSION hidden leaning triangle, the tool calculates the number of overlapped primary studies for each node. To do In this article, we present a new tool for assessing overlap this, the “1. Matrix” sheet sums all the “1” included in of primary studies among SRs. GROOVE is a simple and 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License  PEREZ-BRACCHIGLIONE ET AL. 387 easy-to-use tool which might be valuable for authors of matrix of evidence from which the CCA is calculated.13 overviews, allowing them to perform a comprehensive For example, authors could consider each row as a partic- analysis of the overlap. Besides the calculation of overall ular reference or as a publication thread (i.e., a group of CCA, GROOVE provides a new graphical representation references related to the same study or findings). This lack of the overlap among each pair of possible SRs. Other of clear guidance could lead to an inconsistent assessment graphical and visual approaches have been reviewed in of overlap among different authors. Second, the thresholds detail by Bougioukas et al.17 However, due to the com- for considering the CCA as slight, moderate, high or very plex and heterogeneous nature of possible findings when high are based on the original author recommendations, an overview is conducted, static figures, tables or graphs but the impact of this classification has not yet been would not be enough. In this context, GROOVE provides assessed.18 We are confident that GROOVE will contribute dynamic and customizable features to apply the tool to to address these future challenges, as well as to improve different scenarios, allowing data exploration and man- the development of overviews. agement. For instance, GROOVE enables the analysis of overlap for specific outcomes by creating matrices con- ACKNOWLEDGMENTS taining only the SRs and their primary studies focusing Javier Pérez-Bracchiglione is a PhD candidate at the Doc- on those outcomes,13 and it could certainly help for dis- torate Program on Biomedical Research and Public cussing the results. Indeed, adjusting a given matrix to a Health, Universitat Autònoma de Barcelona, Barcelona, specific clinical question within an overview by including Spain only primary studies and SRs meeting that specific criteria (which involves deleting certain rows and col- CONFLICT OF INTEREST umns in the matrix), could be useful to avoid underesti- The authors declare no conflict of interest. mation of overlap in certain scenarios. We also propose a modification to the CCA formula DA TA AVAI LA BI LI TY S T ATE ME NT by incorporating structural missingness. The consider- The tool presented in this paper is freely available at ation of structural missingness in the generation of a http://doi.org/10.17605/OSF.IO/U2MS4 and https://es. matrix of evidence may provide a complementary cochrane.org/es/groovetool21 approach to the overlap assessment. Structural missing data refers to data that is missing for a logical reason and ORCID that should not exist. For example, if authors incorporate Javier Pérez-Bracchiglione https://orcid.org/0000-0001- chronological structural missingness in their matrices 8738-2184 (i.e., considering as structural zeros all primary studies published after the conduction of a specific SR), they RE FER EN CES could logically and explicitly explain the reason why 1. Bastian H, Glasziou P, Chalmers I. Seventy-five trials and eleven some primary studies are not included in certain SRs. systematic reviews a day: how will we ever keep up? PLoS Med. Also, if the CCA for a certain pair of SRs is 100% when 2010;7(9):e1000326. doi:10.1371/journal.pmed.1000326 considering chronological structural zeros, but below 2. Thomson D, Russell K, Becker L, Klassen T, Hartling L. The 100% when not considering it, authors could choose to evolution of a new publication type: steps and challenges of analyse only the most recent SR for that node, since it producing overviews of reviews. Res Synth Methods. 2010;1 probably provides all the useful data by itself. Structural (3-4):198-211. doi:10.1002/jrsm.30 3. Smith V, Devane D, Begley CM, Clarke M. Methodology in missingness based on a chronological criteria is probably conducting a systematic review of systematic reviews of the most useful example of this feature, but overview healthcare interventions. BMC Med Res Methodol. 2011;11(1): authors may eventually consider other types of structural 15. doi:10.1186/1471-2288-11-15 missingness. For example, if a SR considers only studies 4. Hartling L, Chisholm A, Thomson D, Dryden DM. A descrip- published in English, and a primary study written in tive analysis of overviews of reviews published between 2000 Spanish is incorporated into the matrix, it could be con- and 2011. PLoS One. 2012;7(11):e49667. doi:10.1371/journal. sidered as a language structural missingness for that spe- pone.0049667 cific SR. The use of this feature is optional in the tool and 5. Pollock M, Fernandes RM, Becker LA, Pieper D, Hartling L. Chapter V: overviews of reviews. In: JPT H, Thomas J, still needs to be studied, but it might be advantageous for Chandler J, et al., eds. Cochrane Handbook for Systematic many reviewers and authors. Reviews of Interventions Version 6.1. John Wiley & Sons; 2020. The development of GROOVE raises some other future 6. Chen S-Y, Zhang X-R, Chen J, et al. An overview of meta-ana- challenges. First, there is guidance regarding which steps lyses of endovascular bridging therapies for acute ischemic to take in the development of an overview once the CCA stroke. Biomed Res Int. 2018;2018:9831210. doi:10.1155/ is calculated, but none about how to explicitly build the 2018/9831210 17592887, 2022, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1557 by Readcube (Labtiva Inc.), Wiley Online Library on [06/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 388  PEREZ-BRACCHIGLIONE ET AL. 7. Ioannidis JPA. The mass production of redundant, misleading, handling overlapping, discordant, and problematic data. J Clin and conflicted systematic reviews and meta-analyses. Milbank Epidemiol. 2020;118:69-85. doi:10.1016/j.jclinepi.2019.09.025 Q. 2016;94(3):485-514. doi:10.1111/1468-0009.12210 16. Pollock M, Fernandes RM, Pieper D, Tricco AC, Gates M, 8. Lunny C, Pieper D, Thabet P, Kanji S. Managing Overlap of Gates A, Hartling L Preferred reporting items for overviews of Primary Studies Results Across Systematic Reviews: Practical reviews (PRIOR): a protocol for development of a reporting Considerations for Authors of Overviews of Reviews 2021. doi: guideline for overviews of reviews of healthcare interventions. 10.21203/rs.3.rs-145573/v1 Syst Rev. 2019;8(1):335. doi:10.1186/s13643-019-1252-9 9. Lunny C, Brennan SE, McDonald S, McKenzie JE. Toward a 17. Bougioukas KI, Vounzoulaki E, Mantsiou CD, Savvides ED, comprehensive evidence map of overview of systematic review Karakosta C, Diakonidis T, Tsapas A, Haidich AB Methods for methods: paper 1-purpose, eligibility, search and data extrac- depicting overlap in overviews of systematic reviews: an intro- tion. Syst Rev. 2017;6(1):231. doi:10.1186/s13643-017-0617-1 duction to static tabular and graphical displays. J Clin 10. Lunny C, Brennan SE, McDonald S, McKenzie JE. Toward a Epidemiol. 2021;132:34-45. doi:10.1016/j.jclinepi.2020.12.004 comprehensive evidence map of overview of systematic review 18. Pieper D, Antoine S-L, Mathes T, Neugebauer EAM, methods: paper 2-risk of bias assessment; synthesis, presenta- Eikermann M. Systematic review finds overlapping reviews tion and summary of the findings; and assessment of the cer- were not mentioned in every other overview. J Clin Epidemiol. tainty of the evidence. Syst Rev. 2018;7(1):159. doi:10.1186/ 2014;67(4):368-375. doi:10.1016/j.jclinepi.2013.11.007 s13643-018-0784-8 19. Bijle MNA, Yiu CKY, Ekambaram M. Calcium-based caries 11. Gates M, Gates A, Guitard S, Pollock M, Hartling L. Guidance preventive agents: a meta-evaluation of systematic reviews and for overviews of reviews continues to accumulate, but impor- meta-analysis. J Evid Based Dent Pract. 2018;18(3):203-217.e4. tant challenges remain: a scoping review. Syst Rev. 2020;9(1): doi:10.1016/j.jebdp.2017.09.003 254. doi:10.1186/s13643-020-01509-0 20. Wong B. Color blindness. Nat Methods. 2011;8(6):441. doi:10. 12. Pollock M, Fernandes RM, Newton AS, Scott SD, Hartling L. A 1038/nmeth.1618 decision tool to help researchers make decisions about including 21. Pérez-Bracchiglione J, Meza N, Bangdiwala SI, et al. GROOVE systematic reviews in overviews of reviews of healthcare inter- - Graphical Representation of Overlap for OVErviews. doi:10. ventions. Syst Rev. 2019;8(1):29. doi:10.1186/s13643-018-0768-8 17605/OSF.IO/U2MS4 13. Hennessy EA, Johnson BT. Examining overlap of included studies in meta-reviews: guidance for using the corrected cov- ered area index. Res Synth Methods. 2020;11(1):134-145. doi:10. 1002/jrsm.1390 How to cite this article: Pérez-Bracchiglione J, 14. Cooper H, Koenka AC. The overview of reviews: unique chal- Meza N, Bangdiwala SI, et al. Graphical lenges and opportunities when research syntheses are the prin- Representation of Overlap for OVErviews: cipal elements of new integrative scholarship. Am Psychol. GROOVE tool. Res Syn Meth. 2022;13(3):381-388. 2012;67(6):446-462. doi:10.1037/a0027119 doi:10.1002/jrsm.1557 15. Lunny C, Brennan SE, Reid J, McDonald S, McKenzie JE. Overviews of reviews incompletely report methods for