Improving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research Carol Kilkenny William J. Browne Innes C. Cuthill Michael Emerson Douglas G. Altman

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

 Improving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research 2010 PLoS Biology 10.1371/journal.pbio.1000412 kilkenny2010-arrive-guidelines-reporting-animal-research kilkenny2010-arrive-guidelines-reporting-animal-research.txt, the project's staged copy f1fe875d64d9f970e94a3e86dbb6dfeca3aa5600 cc-by laboratory-animal-science

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

 143 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them inadequate reporting of animal experiments and the minimum information needed for critical review, synthesis, and repetition how reporting guidance can make published animal research complete, transparent, assessable, reusable, and less wasteful the ARRIVE guidelines, a consensus-developed checklist of twenty minimum reporting items for animal research method scientific, ethical, and economic loss caused by incomplete descriptions of animal-study methods and results expert-consensusexpert-consensus mechanismsincomplete information makes publications unusable for policy and clinical or scientific practice mechanismsmethod and result detail enables peer critique, relevance assessment, validity judgment, and experimental repetition mechanismsreporting guidelines specify the information researchers and reviewers should expect in an article mechanismsanimal-number reporting supports significance assessment, data reanalysis, and procedural repetition mechanismsARRIVE packages minimum information into a twenty-item reporting checklist mechanismsthe checklist joins animal characteristics, husbandry, and experimental, statistical, and analytical method details mechanismsexpert working-group consensus produces a draft that wider scientific consultation then refines mechanismsauthors use ARRIVE during manuscript preparation and reviewers use it for quality assurance mechanismswell-designed and well-reported studies supply the evidence units from which systematic reviews are constructed mechanismsjournal endorsement and coordinated stakeholder dissemination extend guideline adoption sound underlying science can remain unusable when its publication omits relevant information the reporting survey covered 271 animal-research articles, so its numerical estimates describe that sampled literature most surveyed papers did not report randomisation or blinding, and many did not fully report statistical methods the paper relies on emerging evidence from CONSORT rather than a direct trial of ARRIVE itself ARRIVE specifies minimum information but does not mandate or prescribe the structure of reporting ARRIVE applies broadly across animal bioscience, but the meaning of laboratory animal is confined to experimental laboratory or formal test settings reporting omissions and limitations of animal models can jointly prevent systematic reviews from reaching useful efficacy conclusions improving animal-research reporting has no single simple or rapid solution and requires coordinated stakeholder effort biomedical researchbiomedical research reporting of biomedical researchreporting of biomedical research incomplete reportingincomplete reporting limited valuelimited value poor reportingpoor reporting animal researchanimal research research methodsresearch methods scientific, ethical, and economic implicationsscientific, ethical, and economic implications hypothesis or objectivehypothesis or objective number and characteristics of the animals usednumber and characteristics of the animals used randomisationrandomisation blindingblinding statistical methodsstatistical methods measure of precision or variabilitymeasure of precision or variability peer review processpeer review process quality controlquality control level of detail of methodslevel of detail of methods presentation of key resultspresentation of key results knowledge baseknowledge base biological relevancebiological relevance reliability and validity of the findingsreliability and validity of the findings experiment to be repeatedexperiment to be repeated reporting guidelinesreporting guidelines CONSORT StatementCONSORT Statement randomised controlled clinical trialsrandomised controlled clinical trials quality and transparency of reports of clinical trialsquality and transparency of reports of clinical trials health researchhealth research bioscience research areasbioscience research areas number of animals usednumber of animals used biological and statistical significancebiological and statistical significance data reanalyseddata reanalysed experimental methodsexperimental methods availability and utility of the informationavailability and utility of the information unnecessary animal useunnecessary animal use guidelines for reporting animal researchguidelines for reporting animal research ARRIVEARRIVE Animals in Research: Reporting In Vivo ExperimentsAnimals in Research: Reporting In Vivo Experiments checklist of 20 itemschecklist of 20 items minimum informationminimum information specific characteristics of animals usedspecific characteristics of animals used housing and husbandryhousing and husbandry experimental, statistical, and analytical methodsexperimental, statistical, and analytical methods high-quality, comprehensive reportinghigh-quality, comprehensive reporting accurate critical reviewaccurate critical review Consensus and consultationConsensus and consultation guideline development processguideline development process expert working groupexpert working group draft set of guidelinesdraft set of guidelines wider consultation with the scientific communitywider consultation with the scientific community final version of the checklistfinal version of the checklist laboratory animalslaboratory animals comparative experimentscomparative experiments other study designsother study designs mandatory or absolutely prescriptivemandatory or absolutely prescriptive structure of reportingstructure of reporting quality assurancequality assurance completeness and transparencycompleteness and transparency output from research using animalsoutput from research using animals design, conduct, and analysis of the experimentsdesign, conduct, and analysis of the experiments systematic reviews of animal researchsystematic reviews of animal research Well-designed and -reported animal studiesWell-designed and -reported animal studies essential building blocksessential building blocks reporting omissionsreporting omissions limitations of the animal modelslimitations of the animal models barrier to reaching any useful conclusionbarrier to reaching any useful conclusion collective effortscollective efforts practical resourcepractical resource journal Instructions to Authorsjournal Instructions to Authors Failure to describe research methodsFailure to describe research methods Improved reportingImproved reporting bias in animal selectionbias in animal selection outcome assessmentoutcome assessment improvements in reporting research using animalsimprovements in reporting research using animals in-depth critiquein-depth critique research articlesresearch articles waste of financial resourceswaste of financial resources CONSORT Statement → reporting guidelinesThe CONSORT Statement was one of the first reporting guidelines. Well-designed and -reported animal studies → systematic reviews of animal researchWell-designed and reported animal studies are building blocks from which a systematic review is constructed. poor reporting → waste of financial resourcesPoor reporting makes publications unusable and contributes to cumulative research waste. Failure to describe research methods → scientific, ethical, and economic implicationsFailure to describe methods and results has potential scientific, ethical, and economic implications. randomisation → bias in animal selectionRandomisation is reported as a method used to reduce bias in animal selection. blinding → outcome assessmentBlinding is reported as a method used to reduce bias in outcome assessment. level of detail of methods → peer review processThe detail of methods is crucial to peer review and later use of the knowledge base. presentation of key results → knowledge basePresentation of key results affects the utility and validity of the knowledge base. research articles → in-depth critiqueResearch articles should include enough relevant information for in-depth critique. number of animals used → biological and statistical significanceReporting animal numbers enables assessment of biological and statistical significance. number of animals used → data reanalysedReporting animal numbers is necessary for data reanalysis. Improved reporting → availability and utility of the informationImproved reporting maximises the availability and utility of information from every animal and experiment. Improved reporting → unnecessary animal useImproved reporting prevents unnecessary animal use in future. CONSORT Statement → ARRIVEARRIVE was developed using the CONSORT Statement as its foundation. minimum information → accurate critical reviewThe twenty-item minimum information set permits accurate critical review of what was done and found. Consensus and consultation → guideline development processConsensus and consultation are described as the corner-stones of guideline development. wider consultation with the scientific community → final version of the checklistFeedback from wider consultation was incorporated into the final checklist. ARRIVE → comparative experimentsARRIVE applies to comparative animal experiments and other study designs. ARRIVE → mandatory or absolutely prescriptiveARRIVE is not intended to be mandatory or absolutely prescriptive. ARRIVE → quality assuranceAuthors and peer reviewers can use the checklist for quality assurance. reporting omissions → barrier to reaching any useful conclusionReporting omissions combine with animal-model limitations to block useful efficacy conclusions. collective efforts → improvements in reporting research using animalsReporting improvement requires collective effort by authors, editors, reviewers, and funders. ARRIVE → practical resourceARRIVE provides a practical resource to aid reporting improvements. fit for purpose - containing enough relevant information to inform policy or clinical and scientific practice good reporting - sufficient methods and results detail for critique, interpretation, reliability assessment, and repetition reporting guidelines - guidance specifying which information belongs in a research article ARRIVE - Animals in Research: Reporting In Vivo Experiments minimum information - twenty classes of detail required for comprehensive animal-research reporting laboratory animal - any animal species undergoing an experimental procedure in a research laboratory or formal test setting quality assurance - peer-review checking for completeness and transparency using the ARRIVE checklist essential building blocks - well-designed and well-reported animal studies composing an evidence synthesis critiques → inadequate biomedical reporting that leaves otherwise sound science unusablenote: The opening problem statement. supports → complete reporting as a prerequisite for peer review, reuse, validity assessment, and replicationnote: The stated rationale for reporting detail. supports → reporting guidelines as an empirically promising interventionnote: The paper cites the emerging CONSORT evidence. proposes → the ARRIVE animal-research reporting guidelinesnote: The paper explicitly leads the guideline initiative. defines → a twenty-item minimum-information checklist for animal-research publicationsnote: The central artifact is enumerated and scoped. supports → consensus and multidisciplinary consultation in guideline developmentnote: The development process is explicitly justified. qualifies → ARRIVE as guidance rather than a mandatory prescriptive reporting structurenote: The scope paragraph expressly limits the guideline. recommends → coordinated adoption by authors, editors, reviewers, funders, journals, and the scientific communitynote: The implementation recommendation is explicit. incomplete reporting —causes→ limited valuethe relation itself is separately evidenced · strength of the relation: 0.85 · note: Direct positive production of reduced publication utility. · the relation stated: Incomplete reporting renders many publications of limited value.incomplete reportinglimited value CONSORT Statement —causes→ quality and transparency of reports of clinical trialsthe relation itself is separately evidenced · strength of the relation: 0.9 · note: Explicit measured improvement. · the relation stated: Evidence shows that CONSORT improves the quality and transparency of clinical-trial reports.CONSORT Statementquality and transparency of reports of clinical trials ARRIVE —same-as→ Animals in Research: Reporting In Vivo Experimentsthe relation itself is separately evidenced · strength of the relation: 1 · note: Explicit referred-to-as naming equivalence. · the relation stated: The animal-research reporting guidelines are referred to as ARRIVE, meaning Animals in Research: Reporting In Vivo Experiments.ARRIVEAnimals in Research: Reporting In Vivo Experiments checklist of 20 items —part-of→ ARRIVEthe relation itself is separately evidenced · strength of the relation: 1 · note: Direct consists-of containment. · the relation stated: The ARRIVE guidelines consist of a checklist of 20 items.checklist of 20 itemsARRIVE

Improving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research

In the last decade the number of bioscience journals has increased enormously, with many filling specialised niches reflecting new disciplines and technologies. The emergence of open-access journals has revolutionised the publication process, maximising the availability of research data. Nevertheless, a wealth of evidence shows that across many areas, the reporting of biomedical research is often inadequate, leading to the view that even if the science is sound, in many cases the publications themselves are not “fit for purpose,” meaning that incomplete reporting of relevant information effectively renders many publications of limited value as instruments to inform policy or clinical and scientific practice [1] – [21] . A recent review of clinical research showed that there is considerable cumulative waste of financial resources at all stages of the research process, including as a result of publications that are unusable due to poor reporting [22] . It is unlikely that this issue is confined to clinical research [2] – [14] , [16] – [20] .

Failure to describe research methods and to report results appropriately therefore has potential scientific, ethical, and economic implications for the entire research process and the reputation of those involved in it. This is particularly true for animal research, one of the most controversial areas of science. The largest and most comprehensive review of published animal research undertaken to date, to our knowledge, has highlighted serious omissions in the way research using animals is reported [5] . The survey, commissioned by the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs), a UK Government-sponsored scientific organisation, found that only 59% of the 271 randomly chosen articles assessed stated the hypothesis or objective of the study, and the number and characteristics of the animals used (i.e., species/strain, sex, and age/weight). Most of the papers surveyed did not report using randomisation (87%) or blinding (86%) to reduce bias in animal selection and outcome assessment. Only 70% of the publications that used statistical methods fully described them and presented the results with a measure of precision or variability [5] . These findings are a cause for concern and are consistent with reviews of many research areas, including clinical studies, published in recent years [2] – [22] .

Good Reporting Is Essential for Peer Review and to Inform Future Research

Scrutiny by scientific peers has long been the mainstay of “quality control” for the publication process. The way that experiments are reported, in terms of the level of detail of methods and the presentation of key results, is crucial to the peer review process and, indeed, the subsequent utility and validity of the knowledge base that is used to inform future research. The onus is therefore on the research community to ensure that their research articles include all relevant information to allow in-depth critique, and to avoiding duplicating studies and performing redundant experiments. Ideally scientific publications should present sufficient information to allow a knowledgeable reader to understand what was done, why, and how, and to assess the biological relevance of the study and the reliability and validity of the findings. There should also be enough information to allow the experiment to be repeated [23] . The problem therefore is how to ensure that all relevant information is included in research publications.

Using Reporting Guidelines Measurably Improves the Quality of Reporting

Evidence provided by reviews of published research suggests that many researchers and peer reviewers would benefit from guidance about what information should be provided in a research article. The CONSORT Statement for randomised controlled clinical trials was one of the first guidelines developed in response to this need [24] , [25] . Since publication, an increasing number of leading journals have supported CONSORT as part of their instructions to authors [26] , [27] . As a result, convincing evidence is emerging that CONSORT improves the quality and transparency of reports of clinical trials [28] , [29] .

Following CONSORT, many other guidelines have been developed—there are currently more than 90 available for reporting different types of health research, most of which have been published in the last ten years (see http://www.equator-network.org and references [30] , [31] ). Guidelines have also been developed to improve the reporting of other specific bioscience research areas including metabolomics and gene expression studies [32] – [37] . Several organisations support the case for improved reporting and recommend the use of reporting guidelines, including the International Committee of Medical Journal Editors, the Council of Science Editors, the Committee on Publication Ethics, and the Nuffield Council for Bioethics [38] – [41] .

Improving the Reporting of Animal Experiments—The ARRIVE Guidelines

Most bioscience journals currently provide little or no guidance on what information to report when describing animal research [42] – [50] . Our review found that 4% of the 271 journal articles assessed did not report the number of animals used anywhere in the methods or the results sections [5] . Reporting animal numbers is essential so that the biological and statistical significance of the experimental results can be assessed or the data reanalysed, and is also necessary if the experimental methods are to be repeated. Improved reporting of these and other details will maximise the availability and utility of the information gained from every animal and every experiment, preventing unnecessary animal use in the future. To address this, we led an initiative to produce guidelines for reporting animal research. The guidelines, referred to as ARRIVE (Animals in Research: Reporting In Vivo Experiments), have been developed using the CONSORT Statement as their foundation [24] , [25] .

The ARRIVE guidelines consist of a checklist of 20 items describing the minimum information that all scientific publications reporting research using animals should include, such as the number and specific characteristics of animals used (including species, strain, sex, and genetic background); details of housing and husbandry; and the experimental, statistical, and analytical methods (including details of methods used to reduce bias such as randomisation and blinding). All the items in the checklist have been included to promote high-quality, comprehensive reporting to allow an accurate critical review of what was done and what was found.

Consensus and consultation are the corner-stones of the guideline development process [51] . To maximise their utility, the ARRIVE guidelines have been prepared in consultation with scientists, statisticians, journal editors, and research funders. We convened an expert working group, comprising researchers and statisticians from a range of disciplines, and journal editors from Nature Cell Biology , Science , Laboratory Animals , and the British Journal of Pharmacology (see Acknowledgments). At a one-day meeting in June 2009, the working group agreed the scope and broad content of a draft set of guidelines that were then used as the basis for a wider consultation with the scientific community, involving researchers, and grant holders and representatives of the major bioscience funding bodies including the Medical Research Council, Wellcome Trust, Biotechnology and Biological Sciences Research Council, and The Royal Society (see Table 1 ). Feedback on the content and wording of the items was incorporated into the final version of the checklist. Further feedback on the content utility of the guidelines is encouraged and sought.

Funding bodies consulted.

The ARRIVE guidelines (see Table 2 ) can be applied to any area of bioscience research using laboratory animals, and the inherent principles apply not only to reporting comparative experiments but also to other study designs. Laboratory animal refers to any species of animal undergoing an experimental procedure in a research laboratory or formal test setting. The guidelines are not intended to be mandatory or absolutely prescriptive, nor to standardise or formalise the structure of reporting. Rather they provide a checklist that can be used to guide authors preparing manuscripts for publication, and by those involved in peer review for quality assurance, to ensure completeness and transparency.

Animal Research: Reporting In Vivo experiments: The ARRIVE guidelines.

Improved Reporting Will Maximise the Output of Published Research

These guidelines were developed to maximise the output from research using animals by optimising the information that is provided in publications on the design, conduct, and analysis of the experiments. The need for such guidelines is further illustrated by the systematic reviews of animal research that have been carried out to assess the efficacy of various drugs and interventions in animal models [8] , [9] , [13] , [52] – [55] . Well-designed and -reported animal studies are the essential building blocks from which such a systematic review is constructed. The reviews have found that, in many cases, reporting omissions, in addition to the limitations of the animal models used in the individual studies assessed in the review, are a barrier to reaching any useful conclusion about the efficacy of the drugs and interventions being compared [2] , [3] .

Driving improvements in reporting research using animals will require the collective efforts of authors, journal editors, peer reviewers, and funding bodies. There is no single simple or rapid solution, but the ARRIVE guidelines provide a practical resource to aid these improvements. The guidelines will be published in several leading bioscience research journals simultaneously [56] – [60] , and publishers have already endorsed the guidelines by including them in their journal Instructions to Authors subsequent to publication. The NC3Rs will continue to work with journal editors to extend the range of journals adopting the guidelines, and with the scientific community to disseminate the guidelines as widely as possible ( http://www.nc3rs.org.uk/ARRIVE ).