PLoS MEDICINE

CONSORT for Reporting Randomized Controlled

Trials in Journal and Conference Abstracts:
Explanation and Elaboration
Sally Hopewell1,2*, Mike Clarke1,3, David Moher4,5, Elizabeth Wager6, Philippa Middleton7, Douglas G. Altman2,
Kenneth F. Schulz8, and the CONSORT Group
1 UK Cochrane Centre, Oxford, United Kingdom, 2 Centre for Statistics in Medicine, Wolfson College, Oxford University, Oxford, United Kingdom, 3 School of Nursing and
Midwifery, Trinity College Dublin, Dublin, Ireland, 4 Chalmers Research Group, Children’s Hospital of Eastern Ontario Research Institute, Ottawa, Canada 5 Department of
Epidemiology and Community Medicine, Faculty of Medicine, University of Ottawa, Ottawa, Canada, 6 Sideview, Princes Risborough, United Kingdom, 7 Discipline of
Obstetrics and Gynaecology, The University of Adelaide, Adelaide, Australia, 8 Family Health International, Research Triangle Park, North Carolina, United States of America

Funding: Financial support was

provided by the following sources to ABSTRACT
convene a meeting of the CONSORT
Group in Montebello, Canada, in
January 2007: the American Society Background
of Clinical Oncology, BMJ, Canadian
Institutes for Health Research, Clear, transparent, and sufficiently detailed abstracts of conferences and journal articles
Johnson & Johnson, The Lancet, related to randomized controlled trials (RCTs) are important, because readers often base their
Nordic Cochrane Centre, PLoS
Medicine, UK Cochrane Centre, and assessment of a trial solely on information in the abstract. Here, we extend the CONSORT
UK National Co-ordinating Centre for (Consolidated Standards of Reporting Trials) Statement to develop a minimum list of essential
Research Methodology. DM is items, which authors should consider when reporting the results of a RCT in any journal or
supported by a University of Ottawa
Research Chair. conference abstract.
Competing Interests: All authors
are involved in many initiatives in Methods and Findings
health care and healthcare research
which should benefit from a wide We generated a list of items from existing quality assessment tools and empirical evidence. A
uptake of the CONSORT for three-round, modified-Delphi process was used to select items. In all, 109 participants were
Abstracts statement.
invited to participate in an electronic survey; the response rate was 61%. Survey results were
Academic Editor: Erik von Elm, presented at a meeting of the CONSORT Group in Montebello, Canada, January 2007, involving
University of Bern, Switzerland
26 participants, including clinical trialists, statisticians, epidemiologists, and biomedical editors.
Citation: Hopewell S, Clarke M, Checklist items were discussed for eligibility into the final checklist. The checklist was then
Moher D, Wager E, Middleton P, et al. revised to ensure that it reflected discussions held during and subsequent to the meeting.
(2008) CONSORT for Reporting
Randomized Controlled Trials in CONSORT for Abstracts recommends that abstracts relating to RCTs have a structured format.
Journal and Conference Abstracts: Items should include details of trial objectives; trial design (e.g., method of allocation, blinding/
Explanation and Elaboration. PLoS masking); trial participants (i.e., description, numbers randomized, and number analyzed);
Med 5(1): e20. doi:10.1371/journal.
pmed.0050020 interventions intended for each randomized group and their impact on primary efficacy
outcomes and harms; trial conclusions; trial registration name and number; and source of
Received: October 3, 2007
Accepted: December 7, 2007
funding. We recommend the checklist be used in conjunction with this explanatory document,
Published: January 22, 2008 which includes examples of good reporting, rationale, and evidence, when available, for the
inclusion of each item.
Copyright: Ó 2008 Hopewell et al.
This is an open-access article
distributed under the terms of the Conclusions
Creative Commons Attribution
License, which permits unrestricted
use, distribution, and reproduction
CONSORT for Abstracts aims to improve reporting of abstracts of RCTs published in journal
in any medium, provided the original articles and conference proceedings. It will help authors of abstracts of these trials provide the
author and source are credited. detail and clarity needed by readers wishing to assess a trial’s validity and the applicability of its
Abbreviations: CONSORT, results.
Consolidated Standards of Reporting
Trials; CSE, Council of Science
Editors; ICMJE, International
Committee of Medical Journal
Editors; STARD, Standards for
Reporting Diagnostic Accuracy;
WAME, World Association of Medical
Editors

* To whom correspondence should

be addressed. E-mail: shopewell@
cochrane.co.uk

PLoS Medicine | www.plosmedicine.org 0001 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

Introduction tional Committee of Medical Journal Editors (ICMJE), and the

Council of Science Editors (CSE). Currently, however, the
Well-written abstracts of conferences and journal articles CONSORT Statement provides limited guidance about
reporting randomized controlled trials (RCTs) are important, preparing abstracts and, while it encourages the use of a
because readers will often base their initial assessment of a structured format, this is not a formal requirement. The
trial on the information reported in an abstract. They may ICMJE Uniform Requirements [23] also provide only limited
then use this information to decide whether or not to seek guidance on the format of abstracts for journal articles.
more knowledge about the trial, such as by reading the full We believe that instructions to authors from journals and
report if available. In some geographic areas, the abstract of a conference organizers should provide specific instructions
RCT may be all that health professionals have easy access to, about key elements of a trial that should be reported in an
and health-care decisions may be made solely on information abstract. Indeed, a recent study examining the content of 35
reported in it. Where the results of a trial are reported only as journals’ instructions to authors found that only 4% of all
a conference abstract, this abstract may provide the only words were devoted to the content or format of the abstract
permanent information about a study and the only way that [24]. Without a minimum amount of key information on a
its results can be accessed by most readers [1]. Journal and trial, it is difficult to assess the validity of its results or its
conference abstracts should contain sufficient information applicability.
about the trial to serve as an accurate record of its conduct
and findings, providing optimal information about the trial
Methods
within the space constraints of the abstract format. A
properly constructed and well-written abstract should also CONSORT for Abstracts: Development of the Checklist
help individuals to assess quickly the validity and applicability In collaboration with others in the CONSORT Group, we
of the findings and, in the case of abstracts of journal articles, have extended the current CONSORT Statement to develop a
aid the retrieval of reports from electronic databases [2]. checklist of essential items that authors should consider when
Conference abstracts, in particular, can provide valuable reporting the main (i.e., those reporting the pre-specified
information for systematic reviewers about studies that are primary outcome) results of a RCT in any journal or
not otherwise published, the exclusion of which from the conference abstract.
review might introduce bias [3]. First, we established a steering committee (MC, SH, DM,
PM, and EW). Second, we generated a list of items from
Incomplete and Inaccurate Reporting existing quality assessment and reporting tools, including the
A number of studies have highlighted the need for CONSORT Statement [22] and other guidance for the
improvements in the reporting of conference abstracts and structured reporting of journal abstracts and short reports
the abstracts of journal articles presenting the results of RCTs [25–28]. Third, additional items were generated as part of an
[4]. There are concerns over the accuracy and quality of trial empirical study assessing the quality of trials reported in
reports published in the proceedings of scientific meetings, conference proceedings and journal abstracts [29].
including the lack of information about the trial and the We then used a modified Delphi consensus method [30] to
robustness of the trial results, compared with results select and reduce the number of possible checklist items. A
published in a journal article [5–9]. Research has also shown total of 109 participants, who were known to have an interest
that trial information reported in conference abstracts may in the reporting of RCTs, the structure of abstracts, or both
differ from that reported in subsequent full publications of were invited (by e-mail) to participate in a Web-based survey
the same study [10–13]. and rate the importance of 27 suggested checklist items. The
The abstract of a journal article has similar limitations to response rate was 61% (n ¼ 63) for the first round of the
those of an abstract submitted to a scientific meeting. In Delphi survey. Respondents included journal editors (13%),
particular, print space limitations constrain the detail that health-care professionals (22%), methodologists (40%), sta-
authors may include on the trial’s methodology and results. A tisticians (5%), trialists (7%), and other individuals with
journal abstract should be an accurate reflection of what is expertise in the reporting of RCTs (13%). During three
included in the full journal article and should not include rounds of the survey, participants were asked about their
information that does not appear in the body of the paper. views on the relative importance of the possible checklist
Studies comparing the accuracy of information reported in a items. A more detailed discussion of the Delphi process is
journal abstract with that reported in the text of the full included in Text S1.
publication have found claims that are inconsistent with, or The results of the survey were presented at a one-day
missing from, the body of the full article [14–18]. Conversely, meeting (part of a three-day CONSORT Group meeting) in
omitting important contrary results from the abstract, such as January 2007, in Montebello, Canada, attended by 26
those concerning side effects, could seriously mislead a participants, several of whom also participated in the Delphi
reader’s interpretation of the trial findings [19,20]. survey. The meeting began with a review of the checklist
items proposed as a result of the Delphi process. Participants
Improving the Reporting of Randomized Trials in Journal then discussed in small groups whether proposed checklist
and Conference Abstracts items should be included, excluded, or modified in the final
The CONSORT (Consolidated Standards of Reporting checklist. These small-group deliberations were further
Trials) Statement, first published in 1996 [21] and updated discussed during plenary sessions.
in 2001 [22], provides recommendations for reporting RCTs Following the meeting, the checklist was revised and
in health-care journals. CONSORT has been endorsed by the circulated to the steering committee and meeting partic-
World Association of Medical Editors (WAME), the Interna- ipants to ensure that it reflected the discussions. The steering

PLoS Medicine | www.plosmedicine.org 0002 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

committee also developed this explanation and elaboration inadvertently misinterpreting data. For example, in a report
document, which was circulated through several iterations of a cluster trial, readers might misinterpret a small sample
among the authors. size as the number of participants rather than the number of
clusters, or vice versa [35].
CONSORT for Abstracts Checklist: Explanation and
Elaboration METHODS
We developed this document using the template used to Participants
develop the CONSORT and STARD (Standards for Reporting Item: Eligibility criteria for participants and the settings
Diagnostic Accuracy) explanatory articles [31,32]. Here each where the data were collected.
item (see Table 1) is stated, a recent example of good Example. ‘‘. . . conducted between June 2003 and January
reporting of the item is provided, followed by an explanation 2005, at 16 government district health centers in Senegal.
that includes the rationale and scientific background and, Patients older than 15 years with newly diagnosed sputum
where possible, discusses the evidence for the item as it smear-positive pulmonary TB were randomly assigned to the
relates to a trial reported in a journal or conference abstract. intervention or control group’’ [33].
Explanation. Every RCT addresses an issue relevant to a
Checklist Items particular population or group with the condition of interest.
Trialists may further restrict this sample by using eligibility
TITLE
criteria and by performing the trial in a particular setting (for
Item: Identification of the study as randomized. example primary, secondary, or tertiary care). Participant
Example. ‘‘Effectiveness of a strategy to improve adherence eligibility criteria may relate to demographics, clinical
to tuberculosis treatment in a resource poor setting: a cluster diagnosis, and comorbid conditions. A clear description of
randomized trial’’ [33]. the trial participants and setting in which they were studied is
Explanation. The ability to identify a relevant report in an needed so that readers may assess the external validity
electronic database depends to a large extent on how it was (generalisability) of the trial and determine its applicability to
indexed. Indexers may not classify a report as a RCT if the their own setting.
authors do not explicitly report this information [34]. To help
ensure that a study is appropriately indexed and identified as Interventions
a RCT, authors should state explicitly in the title that the Item: Interventions intended for each group.
participants were randomly assigned to their comparison Example. ‘‘Patients were randomized to receive either 100
groups. mg hydrocortisone or matching placebo as follows: the first
dose in the evening of the operative day, then 1 dose every 8
AUTHORS hours during the next 3 days. In addition, all patients received
Item: Contact details for the corresponding author. oral metoprolol (50–150 mg/d) titrated to heart rate’’ [36].
(This item is specific to conference abstracts) Explanation. The essential features of the experimental
Example. ‘‘Correspondence to: Dr Sally Hopewell, UK and comparison interventions should be described. Authors
Cochrane Centre, Summertown Pavilion, Middle Way, Ox- should report details about the interventions, e.g., dose, route
ford OX2 7LG, UK. Tel: þ44 1865 516300; Fax: þ44 1865 of administration, duration of administration, surgical
516311; Email: [email protected].’’ procedure, or manufacturer of inserted device.
Explanation. Adequate contact details for the correspond-
ing author are particularly important for RCTs reported in Objective
conference proceedings. These abstracts may be the only Item: Specific objective or hypothesis.
lasting source of information for many trials, as only half of Example. ‘‘To compare the effectiveness of an early switch
RCTs reported in conference proceedings are subsequently to oral antibiotics with the standard 7 day course of
published in full [1]. Adequate contact information would intravenous antibiotics in severe community acquired pneu-
enable readers to contact trialists for additional information monia’’ [37].
or clarifications regarding reported data. Adequate contact Explanation. The abstract should provide a clear statement
details should include the telephone number, postal, and of the specific objective or hypothesis addressed in the trial. If
email address of the principal investigator and, if available, more than one objective is addressed, the main objective (i.e.,
the trial Web site. based on the prespecified primary outcome) should be
indicated and only key secondary objectives stated [26].
TRIAL DESIGN
Item: Description of the trial design. Outcome
Example. ‘‘A cluster randomized controlled trial...’’ [33]. Item: Clearly defined primary outcome for this report.
Explanation. The design of the trial should be described, Example. ‘‘Main outcome measure: all-cause mortality at
for example, parallel group, cluster randomized, crossover, 180 days’’ [38].
factorial, superiority, equivalence or noninferiority, or some Explanation. RCTs assess outcomes for which the inter-
other combination of these designs. An important reason for ventions are being compared. Most trials have several
identifying the design of the trial is to ensure appropriate outcomes, some of which are deemed more important than
indexing in electronic databases, thus ensuring greater ease others. Such rankings are typically reported as primary and
of identification [34]. Alerting readers to the design of the secondary outcomes. There is evidence of selective reporting
trial also provides transparency as to the type of design used with significant or favourable outcomes being more likely to
to conduct the trial and should reduce the likelihood of be published than nonsignificant outcomes [39–41].

PLoS Medicine | www.plosmedicine.org 0003 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

Table 1. Items to Include when Reporting a Randomized Trial in a Journal or Conference Abstract

Item Description

Title Identification of the study as randomized

Authors * Contact details for the corresponding author
Trial design Description of the trial design (e.g. parallel, cluster, non-inferiority)
Methods
Participants Eligibility criteria for participants and the settings where the data were
collected
Interventions Interventions intended for each group
Objective Specific objective or hypothesis
Outcome Clearly defined primary outcome for this report
Randomization How participants were allocated to interventions
Blinding (masking) Whether or not participants, care givers, and those assessing the outcomes
were blinded to group assignment
Results
Numbers randomized Number of participants randomized to each group
Recruitment Trial status
Numbers analysed Number of participants analysed in each group
Outcome For the primary outcome, a result for each group and the estimated effect size
and its precision
Harms Important adverse events or side effects
Conclusions General interpretation of the results
Trial registration Registration number and name of trial register
Funding Source of funding

*This item is specific to conference abstracts.

doi:10.1371/journal.pmed.0050020.t001

Authors should explicitly state the primary outcome for the randomization; sequentially numbered identical containers
trial and when it was assessed (e.g., the time frame over which that are administered serially to participants; on-site com-
it was measured). The primary outcome is the prespecified puter system combined with allocations kept in a locked,
outcome considered of greatest importance and is usually the unreadable computer file that investigators can access only
one used in the sample size calculation [22]. In some instances after the characteristics of an enrolled participant are
a publication may report an outcome different from the entered; and sequentially numbered, opaque sealed envelopes
primary outcome. For example, conference abstracts are [46].
more likely to report interim analyses than are full The method of allocation concealment is generally poorly
publications [8,10], or to present different results for a single reported in conference abstracts and in abstracts of journal
trial in a series of abstracts. If the abstract focuses on a articles [7,47–49]. For example, in a review of 494 abstracts
secondary outcome of a trial, the abstract should identify presented at an oncology conference in 1992 and 2002, only
both this outcome and the primary outcome of the trial. nine (2%) abstracts reported the method of allocation
concealment. This information was missing from the remain-
Randomization
ing 485 conference abstracts, with no improvements seen
Item: How participants were allocated to interventions.
over the ten-year period [7].
Example. ‘‘Randomization was computer-generated, with
allocation concealment by opaque sequentially numbered Blinding (Masking)
sealed envelopes’’ [42]. Item: Whether or not participants, caregivers, and those
Explanation. It is important to conceal the allocation
assessing the outcomes were blinded to group
sequence from those assigning participants to the interven-
tion groups. Allocation concealment prevents investigators
assignment.
from influencing which participants are assigned to a given Example. ‘‘Children, parents, and the research assistants
intervention group (i.e., selection bias). Evidence shows that were blinded to group assignment’’ [50].
reports of trials reporting inadequate allocation concealment Explanation. Blinding refers to the practice of keeping the
are associated with exaggerated treatment effects [43,44]. trial participants, care providers, data collectors, and some-
Research suggests that adequate allocation concealment is times those analysing the data, unaware of which intervention
more important in preventing selection bias than are other is being administered to which participant, so that they will
components of the randomization process, such as the not be influenced by that knowledge. The term masking is
sequence generation (e.g., use of computer or random sometimes used instead of blinding [51,52] and might be
number table) [45]. preferable when reporting studies involving eyes and vision.
Authors should clearly describe the method for assigning It is important that authors describe whether or not
participants to interventions. Examples of approaches used participants, those administering the intervention (usually
to ensure adequate concealment include: centralised (e.g., health-care providers), and those assessing the outcome (the
allocation by a central office) or pharmacy-controlled data collectors and analysts) were blinded to the group

PLoS Medicine | www.plosmedicine.org 0004 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

allocation. Authors should avoid using terms such as ‘‘single’’ group. These data permit an assessment of whether partic-
or ‘‘double’’ blind as such terms are not well-understood [53]. ipants were analysed according to their original group
Information on the method of blinding is poorly reported assignment, which is important, because failure to include
in conference and journal abstracts [7,8,47–49]. Such report- all participants in the analysis may bias the results of the trial
ing is valuable as blinding may be important in protecting [22].
against bias [51]. Studies have shown that if investigators are Several studies have reported deficiencies in journal and
aware of the treatment, their attitudes for or against an conference abstracts in reporting the number of participants
intervention can directly affect whether or not they include, included in the analysis [6–8,13,48,59]. In a review of RCTs in
or treat, participants in a trial [45]. Furthermore, there is acute brain injury reported in journal abstracts, only 43%
evidence that participants who are aware of their assignment reported the number of participants included in the analysis
status are more likely to report symptoms, leading to biased [48]. In another evaluation of trials reported in abstracts for
results [51]. Perhaps most importantly, if outcome assessors an oncology conference, only 40% reported the number of
are not blinded to the intervention they are more likely to participants analysed, and only 6% indicated intention to
report favourable outcomes for the intervention which they treat analysis [8].
believe is better [54]. However, unlike allocation conceal-
ment, blinding of the participants, health-care providers, and Outcome
outcome assessors may not always be appropriate or possible, Item: For the primary outcome, a result for each group
such as in many surgical trials. In this case, authors should and the estimated effect size and its precision.
report if any form of blinding (such as blinding of data Example. ‘‘Treatment was successful for 682 (88%) of 778
analysts) was used. patients recruited in the intervention group, and for 563
(76%) of 744 patients recruited in the control group (adjusted
RESULTS risk ratio [RR], 1.18; 95% confidence interval [CI], 1.03–1.34)’’
Numbers Randomized [33].
Item: Number of participants randomized to each group. Explanation. For the primary outcome, authors should
Example. ‘‘Children (n ¼ 633) aged 1–3 randomly allocated report trial results as a summary of the outcome in each
to receive fortified milk (n ¼ 316) or control milk (n ¼ 317)’’ group (e.g., the number of participants with or without the
[55]. event, or the mean and standard deviation of measurements),
Explanation. The number of participants randomized to together with the contrast between groups known as the
each intervention group is an essential element of the results effect size. For binary outcomes, the effect size could be the
of a trial. This number defines the sample size, and readers relative risk, relative risk reduction, odds ratio, or risk
can use it to assess whether all randomized participants were difference. For survival time data, the measurement could
included in the data analysis. Again, this may be particularly be the hazard ratio or difference in median survival time. For
important for conference abstracts reporting interim analy- continuous data, the effect measure is usually the difference
ses, if a trial is still open to participant accrual or follow-up in means. Authors should present confidence intervals for the
[8,10]. Here authors should report the period of recruitment contrast between groups and as a measure of the precision
on which the data are based. (uncertainty) of the estimate of the effect [22]. For abstracts
not reporting the ‘‘primary’’ outcome of the trial (e.g.,
Recruitment abstracts focusing on safety data or economic impacts), the
Item: Trial status. secondary nature of the outcomes should be indicated, and,
Example. ‘‘An interim analysis was performed because of where possible, sufficient details of the primary outcome
slow accrual’’ [56]. should be included to allow other findings to be taken in the
Explanation. Authors should describe the status of the trial proper context.
and whether it is still ongoing, closed to recruitment, or Several studies have observed deficiencies in the reporting
closed to follow-up. This information is particularly impor- of statistical results in journal abstracts [57,60–62]. For
tant for conference abstracts, which are more likely than full example, Pocock and colleagues [57] found that journal
articles to report interim analyses [10]. abstracts of RCTs tended to overemphasize statistically
If the trial has stopped earlier than planned it is important significant outcomes compared to the full journal article,
to say why. Possible reasons for early termination include: leading to problems in interpretation of the results. Poor
slow accrual rates, poor data quality, poor adherence, reporting of results is also a problem for trials presented in
resource deficiencies, unacceptable harms or large benefits, conference abstracts [7,8,59]. A study of 494 reports of RCTs
or emerging information that makes the trial irrelevant, in oncology found that only 26% of conference abstracts
unnecessary, or unethical. If a trial stops early for apparent reported the size of the effect and significance of the result
benefit, the estimates of treatment effect are more likely to be [7].
biased and prone to exaggeration [57,58].
Harms
Numbers Analysed Item: Important adverse events or side effects.
Item: Number of participants analysed in each group. Example. ‘‘Adverse events were more common with top-
Example. ‘‘. . . 300 were included in the analysis of the iramate vs placebo, respectively, including paresthesia (50.8%
primary outcome (100 in the acetaminophen group, 100 in vs 10.6%), taste perversion (23.0% vs 4.8%), anorexia (19.7%
the ibuprofen group, and 100 in the codeine group)’’ [50]. vs 6.9%), and difficulty with concentration (14.8% vs 3.2%)’’
Explanation. Authors should report the number of [63].
participants included in the analysis for each intervention Explanation. Most interventions have unintended and

PLoS Medicine | www.plosmedicine.org 0005 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

often undesirable effects as well as intended and beneficial of them are subsequently published [1]. Such trial registration
effects. In order to make rational and balanced decisions, provides readers with a way to obtain more information
readers need information about the relative benefits and about the trial and its results. Registration information will
harms of an intervention. Authors should describe any also help to link abstracts with subsequent full publications
important adverse (or unexpected) effects of an intervention (or multiple abstracts from the same trial) and thus reduce
in the abstract. If no important adverse events have occurred, the risk of inadvertent double-counting in systematic reviews.
the authors should state this explicitly [20].
Explicit reference to the reporting of harms in the title or FUNDING
abstract is also important for appropriate database indexing Item: Source of funding.
and information retrieval. Derry and colleagues [64] found Example. ‘‘Funded by The Breast Cancer Research Foun-
that only 66 of 107 RCTs that reported data on adverse events dation’’ [72].
in the full publication mentioned harms in the title or Explanation. Authors should report the source of funding
abstract; thus, harms could not have been identified for many for the trial as this is important information for readers
of the articles in a search of titles and abstracts in an assessing a trial. A recent systematic review showed that
electronic bibliographic database. studies funded by the pharmaceutical industry had four times
Harms are also poorly reported in conference abstracts. A (odds ratio 4.05; 95% confidence interval 2.98–5.51) the odds
recent examination of over 800 ophthalmology conference of having outcomes favouring the sponsor than studies
abstracts reporting trials found that the majority (71%) did funded by other sources [73]. Similarly, authors should report
not report harms related to the treatment intervention, and any other sources of support, such as in the preparation of
harms were reported as a primary outcome measure in only the abstract, presentation, or manuscript [74].
6% of abstracts [9].
Discussion
CONCLUSIONS
Item: General interpretation of the results. CONSORT for Abstracts strongly recommends the use of
Example. ‘‘Multivitamin supplementation reduced the structured abstracts for reporting RCTs [75]; the full
incidence of low birth weight and small-for-gestational-age CONSORT Statement also supports their use [22]. Since
births but had no significant effects on prematurity or fetal 1987 when the Ad Hoc Working Group for Critical Appraisal
death’’ [65]. of the Medical Literature [25–27] first published recommen-
Explanation. The conclusions of the trial, consistent with dations for the adoption of structured abstracts, many
the results reported in the abstract, should be clearly stated journals have promoted their use, and many different formats
along with their clinical application (avoiding over-general- for structured abstracts now exist. We recognise that journals
isation). Authors should balance the benefits and harms in may have developed their own set of headings for abstracts
their conclusions. Where applicable, authors should also note [76,77]. It is not the intention of this reporting guide to
whether additional studies are required before the results are suggest changes to these headings but to recommend what
used in clinical settings [26]. information should be reported within them when describing
a RCT.
TRIAL REGISTRATION It is important to note that, because of the space
Item: Registration number and name of trial register. limitations of an abstract, it will only ever be possible to
Example. ‘‘Trial Registry: www.clinicaltrials.gov; Identifier: provide limited information about a trial report. CONSORT
NCT00412009’’ [33]. for Abstracts sets out to recommend what information should
Explanation. Nonpublication of entire trials and selective be reported within these constraints when describing a RCT.
reporting of outcomes within trials has been well-docu- Readers of abstracts should always try to obtain more
mented [39,41,66]. Covert redundant publication can also information about a trial and its results, either by accessing
cause problems in systematic reviews when results from the the full publication or, in the case of unpublished conference
same trial are inadvertently included more than once [67]. To abstracts [1], by contacting the authors for more information.
minimize or avoid these problems there have been many calls With the aim of greatly improving access to information
for trial registration [68]. Due to more recent serious about clinical trials and their results, the World Health
problems of withholding data [69] there has been a renewed Organization (WHO) recently established an International
effort to register RCTs. By registering a RCT, authors Clinical Trials Registry Platform. Their goal is to produce a
typically report a minimal set of information and obtain a single minimum standard for information that trialists should
unique trial registration number. disclose before the trial begins [78,79]. Moreover, as registra-
In September 2004 the International Committee of Medical tion of the trial methods has become more common, several
Journal Editors (ICMJE) indicated a change in their policy for forces have begun to advocate for the disclosure of trial
publishing RCTs, saying that they would consider trials for results in specially designed repositories linked to trial
publication only if they had been registered before the registers. In June 2007, endorsing the WHO’s International
enrolment of the first patient (as of 1 July 2005) [70]. This Clinical Trials Registry Platform, the ICMJE published an
position has resulted in a dramatic increase in the number of editorial recommending a standard abstract format for
trials being registered [71]. reporting results. The ICMJE suggested that CONSORT for
In an abstract reporting a trial, authors should provide Abstracts may be one such option [80]. At present, there is no
details of the trial registration number and name of trial formal consensus on international norms and standards for
register. Registration information will be particularly impor- results reporting. The WHO International Clinical Trials
tant for abstracts reported in conference meetings, as not all Registry Platform has therefore established a Study Group on

PLoS Medicine | www.plosmedicine.org 0006 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

the Reporting of Findings of Clinical Trials to advise the Johnson & Johnson Pharmaceutical Research and Development,
WHO Registry Platform on matters related to the reporting United States; Isabelle Boutron, University Paris 7 Denis Diderot,
Assistance Publique des Hôpitaux de Paris, INSERM, France; P.J.
of the findings of clinical trials. Full transparency and Devereaux, McMaster University, Canada; Kay Dickersin, Johns
accountability require that all results of all trials are made Hopkins Bloomberg School of Public Health, United States; Diana
available to the public in a timely manner. Elbourne, London School of Hygiene & Tropical Medicine, United
Kingdom; Susan Ellenberg, University of Pennsylvania School of
Like the CONSORT Statement, CONSORT for Abstracts Medicine, United States; Val Gebski, University of Sydney, Australia;
has been developed primarily for reporting the main results Steven Goodman, Journal of the Society for Clinical Trials, United States;
of parallel group RCTs (i.e., those relating to the prespecified Peter C. Gøtzsche, Nordic Cochrane Centre, Denmark; Trish Groves,
BMJ, United Kingdom, Steven Grunberg, American Society of
primary outcome). There may well be instances where Clinical Oncology, United States; Brian Haynes, ACP Journal Club,
different types of trial information, such as composite Canada; Astrid James, The Lancet, United Kingdom; Peter Juhn,
outcomes, or different designs, such as cluster trials or Johnson & Johnson, United States; Don Minckler, University of
noninferiority and equivalence trials, will require additional California Irvine, United States; Victor M. Montori, Knowledge and
Encounter Research Unit, Mayo Clinic College of Medicine, United
information not covered in this explanation and elaboration States; Cynthia Mulrow, Annals of Internal Medicine, United States;
document. Possible additional abstract extensions may be Stuart Pocock, London School of Hygiene & Tropical Medicine,
warranted, as has been done for the CONSORT Statement for United Kingdom; Drummond Rennie, JAMA, United States; David
Schriger, Annals of Emergency Medicine, United States; Iveta Simera,
full reports [35,81]. EQUATOR Network, United Kingdom.
The length of an abstract reporting a RCT using the Author contributions. SH, MC, DM, EW, and PM were involved in
CONSORT for Abstracts checklist is difficult to estimate. In the design, implementation, analysis of the study, and in writing and
developing the checklist 250 to 300 words were found to be commenting on drafts on the final manuscript. DGA and KFS were
involved in the analysis of the study and commenting on drafts on the
sufficient to address all of the items in the checklist. Worked final manuscript.
examples of using the CONSORT for Abstracts checklist are
References
available on the CONSORT website at http://www.
1. Scherer RW, Langenberg P, von Elm E (2007) Full publication of results
consort-statement.org/. In the past, MEDLINE has truncated initially presented in abstracts. Cochrane Database of Systematic Reviews,
journal abstracts at 250 words [82]. This has resulted in many Art. No.: MR000005. doi:10.1002/14651858.MR000005.pub3. Available: http://
www.mrw.interscience.wiley.com/cochrane/clsysrev/articles/MR000005/
journals setting word limits for abstracts in their journals at frame.html. Accessed 1 May 2007.
250 words. However, since 2000 the National Library of 2. Harbourt AM, Knecht LS, Humphreys BL (1995) Structured abstracts in
Medicine increased the word limit for an abstract appearing MEDLINE, 1989–1991. Bull Med Libr Assoc 83: 190–195.
3. Hopewell S, McDonald S, Clarke M, Egger M (2007) Grey literature in meta-
in MEDLINE to 10,000 characters, which equates to over analyses of randomized trials of health care interventions. Cochrane
1,000 words. While most abstract reports will not require Database of Systematic Reviews, Art. No.: MR000010. doi:10.1002/14651858.
MR000010.pub3. Available: http://www.mrw.interscience.wiley.com/
anywhere near 1,000 words, such a word length will be cochrane/clsysrev/articles/MR000010/frame.html. Accessed 1 May 2007.
sufficient to report even the most complex of trials in 4. Hopewell S, Eisinga A, Clarke M (2007) Better reporting of randomized
abstract form [82]. trials in biomedical journal and conference abstracts. J Info Sci doi:10.1177/
0165551507080415.
Clear, transparent, and accurate reporting of research is 5. Chalmers I, Adams M, Dickersin K, Hetherington J, Tarnow-Mordi W, et al.
important because it enables readers to understand what was (1990) A cohort study of summary reports of controlled trials. JAMA 263:
done and hence to evaluate the reliability and relevance of 1401–1405.
6. Herbison P (2005) The reporting quality of abstracts of randomised
the findings. This extension of the CONSORT Statement aims controlled trials submitted to the ICS meeting in Heidelberg. Neurourol
to improve the reporting of RCTs in both the abstracts of Urodyn 24: 21–24.
7. Hopewell S, Clarke M (2005) Abstracts presented at the American Society
journal articles and conference proceedings [83]. When using of Clinical Oncology conference: how completely are trials reported? Clin
the CONSORT for Abstracts checklist we encourage authors Trials 2: 265–268.
to use it in conjunction with this explanation and elaboration 8. Krzyzanowska MK, Pintilie M, Tannock IF (2003) Factors associated with
failure to publish large randomized trials presented at an oncology
document. We encourage journals and conference organisers meeting. JAMA 290: 495–501.
to endorse the use of CONSORT for Abstracts by modifying 9. Scherer R (2006) Are harms reported in abstracts of trial results from
conference proceedings? XIV Cochrane Colloquium, 23–26 October,
their ‘‘Instructions to Authors’’ and drawing their readers’ Dublin, Ireland: 63.
attention to this reporting guidance, perhaps through an 10. Hopewell S, Clarke M, Askie L (2006) Reporting of trials presented in
editorial or by including a link to the checklist on the conference abstracts needs to be improved. J Clin Epidemiol 59:: 681–684.
11. Toma M, McAlister FA, Bialy L, Adams D, Vandermeer B, et al. (2006)
conference website. The most important benefit will be to Transition from meeting abstract to full-length journal article for
enable readers to use abstracts more effectively and to assess randomized controlled trials. JAMA 295: 1281–1287.
the validity of the research more precisely. When key aspects 12. Dundar Y, Dodd S, Dickson R, Walley T, Haycox A, et al. (2006) Comparison
of conference abstracts and presentations with full-text articles in the
of study methods are omitted, reader assessments are less health technology assessments of rapidly evolving technologies. Health
certain, and might well take longer to make. Technol Assess 10: 1–145.
13. Chokkalingam A, Scherer R, Dickersin K (1998) Agreement of data
abstracts compared to full publications. Control Clin Trials 19: 61S–62S.
Supporting Information 14. Estrada CA, Bloch RM, Antonacci D, Basnight LL, Patel SR, et al. (2000)
Reporting and concordance of methodologic criteria between abstracts
Text S1. Results of the Delphi Consensus Survey and articles in diagnostic test studies. J Gen Intern Med 15: 183–187.
15. Froom P, Froom J (1993) Deficiencies in structured medical abstracts. J Clin
Found at doi:10.1371/journal.pmed.0050020.sd001 (69 KB RTF). Epidemiol 46: 591–594.
16. Harris AH, Standard S, Brunning JL, Casey SL, Golderg JH, et al. (2002) The
accuracy of abstracts in psychology journals. J Psychol 136: 141–148.
Acknowledgments 17. Pitkin RM, Branagan MA, Burmeister LF (1999) Accuracy of data in
abstracts of published research articles. JAMA 281: 1110–1111.
We are grateful to everyone who contributed to the development of 18. Ward LG, Kendrach MG, Price SO (2004) Accuracy of abstracts for original
CONSORT for Abstracts and to the Delphi survey. We are also research articles in pharmacy journals. Ann Pharmacother 38: 1173–1177.
grateful to those who participated in the meeting of the CONSORT 19. Ioannidis JP, Lau J (2001) Completeness of safety reporting in randomized
Group in January 2007 and commented on drafts of the manuscript: trials: an evaluation of 7 medical areas. JAMA 285: 437–443.
Virginia Barbour, PLoS Medicine, United Kingdom; Jesse Berlin, 20. Ioannidis JP, Evans SJ, Gøtzsche PC, O’Neill RT, Altman DG, et al. (2004)

PLoS Medicine | www.plosmedicine.org 0007 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

Better reporting of harms in randomized trials: an extension of the 50. Clark E, Plint AC, Correll R, Gaboury I, Passi B (2007) A randomized,
CONSORT statement. Ann Intern Med 141: 781–788. controlled trial of acetaminophen, ibuprofen, and codeine for acute pain
21. Begg C, Cho M, Eastwood S, Horton R, Moher D, et al. (1996) Improving the relief in children with musculoskeletal trauma. Pediatrics 119: 460–467.
quality of reporting of randomized controlled trials. The CONSORT 51. Schulz KF, Grimes DA (2006) The Lancet handbook of essential concepts in
statement. JAMA 276: 637–639. clinical research. London: Elsevier.
22. Moher D, Schulz KF, Altman DG, Lepage L (2001) The CONSORT 52. Schulz KF, Altman DG, Moher D (2007) Blinding is better than masking.
statement: revised recommendations for improving the quality of reports BMJ 334: 918.
of parallel-group randomised trials. Lancet 357: 1191–1194. 53. Devereaux PJ, Manns BJ, Ghali WA, Quan H, Lacchetti C, et al. (2001)
23. International Committee of Medical Journal Editors (2006) Uniform Physician interpretations and textbook definitions of blinding terminology
requirements for manuscripts submitted to biomedical journals: writing in randomized controlled trials JAMA 285: 2000–2003.
and editing for biomedical publication. [Updated February 2006.] 54. Poolman RW, Struijs PA, Krips R, Sierevelt IN, Marti RK, et al. (2007)
Available: http://www.icmje.org. Accessed 1 December 2006. Reporting of outcomes in orthopaedic randomized trials: does blinding of
24. Schriger DL, Arora S, Altman DG (2006) The content of medical journal outcome assessors matter? J Bone Joint Surg 89-A: 550–558.
Instructions for authors. Ann Emerg Med 48: 743–749. 55. Sazawal S, Dhingra U, Dhingra P, Hiremath G, Kumar J, et al. (2007) Effects
25. Ad Hoc Working Group for Critical Appraisal of the Medical Literature of fortified milk on morbidity in young children in north India:
(1987) A proposal for more informative abstracts of clinical articles. Ann community based, randomised, double masked placebo controlled trial.
Intern Med 106: 598–604. BMJ 334: 140.
26. Haynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ (1990) More 56. Kim KB, Legha SS, Gonzalez R, Anderson C, Papadopoulos NE, et al. (2006)
informative abstracts revisited. Ann Intern Med 113: 69–76. A phase III randomized trial of adjuvant biochemotherapy (BC) versus
27. Haynes RB, Mulrow CD, Huth EJ, Altman DG, Gardner MJ (1996) More interferon-a-2b (IFN) in patients (pts) with high risk for melanoma
informative abstracts revisited. Cleft Palate-Craniofac J 33: 1–9. recurrence. J Clin Oncol, ASCO Annual Meeting Proceedings Part I. Vol
28. Deeks JJ, Altman DG (1998) Inadequate reporting of controlled trials as 24: 8003.
short reports. Lancet 318: 193–194. 57. Pocock SJ, Hughes MD, Lee RJ (1987) Statistical problems in the reporting
29. Hopewell S (2004) Impact of grey literature on systematic reviews of of clinical trials. N Engl J Med 317: 426–432.
randomized trials [PhD dissertation]. Oxford (UK): Wolfson College, 58. Montori VM, Devereaux PJ, Adhikari NK, Burns KE, Eggert CH, et al. (2005)
University of Oxford. Randomized trials stopped early for benefit: a systematic review. JAMA 294:
30. Hasson F, Keeney S, McKenna H (2000) Research guidelines for the Delphi 2203–2209.
survey method. J Adv Nurs 32: 1008–1015. 59. Bhandari M, Devereaux PJ, Guyatt GH, Cook DJ, Swiontkowski MF, et al.
31. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, et al. (2001) The (2002) An observational study of orthopaedic abstracts and subsequent full-
revised CONSORT statement for reporting randomized trials: explanation text publications. J Bone Joint Surg 84-A: 615–621.
and elaboration. Ann Intern Med 134: 663–694. 60. Dryver E, Hux JE (2002) Reporting of numerical and statistical differences
32. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, et al. (2003) in abstracts: improving but not optimal. J Gen Intern Med 17: 203–206.
The STARD statement for reporting studies of diagnostic accuracy: 61. Gøtzsche PC (2006) Believability of relative risks and odds ratios in
explanation and elaboration. Ann Intern Med 138: W1–12. abstracts: cross sectional study. BMJ 333: 231–234.
33. Thiam S, LeFevre AM, Hane F, Ndiaye A, Ba F, et al. (2007) Effectiveness of 62. Schwartz LM, Woloshin S, Dvorin EL, Welch HG (2006) Ratio measures in
a strategy to improve adherence to tuberculosis treatment in a resource- leading medical journals: structured review of accessibility of underlying
poor setting: a cluster randomized controlled trial. JAMA 297: 380–386. absolute risks. BMJ 333: 1248–1250.
34. Dickersin K, Manheimer E, Wieland S, Robinson KA, Lefebvre C, et al. 63. Johnson BA, Rosenthal N, Capece JA, Wiegand F, Mao L, et al. (2007)
(2002) Development of the Cochrane Collaboration’s CENTRAL register of Topiramate for treating alcohol dependence: a randomized controlled
controlled clinical trials. Eval Health Prof 25: 38–64. trial. JAMA 298: 1641–1651.
35. Campbell MK, Elbourne DR, Altman DG; CONSORT group (2004) 64. Derry S, Kong Loke Y, Aronson JK (2001) Incomplete evidence: the
CONSORT statement: extension to cluster randomised trials. BMJ 328: inadequacy of databases in tracing published adverse drug reactions in
702–708. clinical trials. BMC Med Res Methodol 1: 7.
36. Halonen J, Halonen P, Jarvinen O, Taskinen P, Auvinen T, et al. (2007) 65. Fawzi WW, Msamanga GI, Urassa W, Hertzmark E, Petraro P, et al. (2007)
Corticosteroids for the prevention of atrial fibrillation after cardiac Vitamins and perinatal outcomes among HIV-negative women in Tanzania.
surgery: a randomized controlled trial. JAMA 297: 1562–1567. N Engl J Med 356: 1423–1431.
37. Oosterheert JJ, Bonten MJ, Schneider MM, Buskens E, Lammers JW, et al. 66. Dickersin K (1997) How important is publication bias? A synthesis of
(2006) Effectiveness of early switch from intravenous to oral antibiotics in available data. AIDS Educ Prev 9: 15–21
severe community acquired pneumonia: multicentre randomised trial. BMJ 67. Tramer MR, Reynolds DJ, Moore RA, McQuay HJ (1997) Impact of covert
33: 1193. duplicate publication on meta-analysis: a case study. BMJ 315: 635–640.
38. Mebazaa A, Nieminen MS, Packer M, Cohen-Solal A, Kleber FX, et al. (2007) 68. Simes RJ (1986) Publication bias: the case for an international registry of
Levosimendan vs dobutamine for patients with acute decompensated heart clinical trials. J Clin Oncol 4: 1529–1541.
failure: the SURVIVE randomized trial. JAMA 297: 1883–1891. 69. Whittington CJ, Kendall T, Fonagy P, Cottrell D, Cotgrove A, et al. (2004)
39. Chan AW, Hrobjartsson A, Haahr MT, Gøtzsche PC, Altman DG (2004A) Selective serotonin reuptake inhibitors in childhood depression: systematic
Empirical evidence for selective reporting of outcomes in randomized review of published versus unpublished data. Lancet 363: 1341–1345.
trials: comparison of protocols to published articles. JAMA 291: 2457– 70. De Angelis CD, Drazen JM, Frizelle FA, Haug C, Hoey J, et al. (2005) Is this
2465. clinical trial fully registered? A statement from the International
40. Chan AW, Altman DG (2004B) Identifying outcome reporting bias in Committee of Medical Journal Editors. Lancet 365: 1827–1829.
randomised trials on PubMed: review of publications and survey of authors. 71. Zarin DA, Ide NC, Tse T, Harlan WR, West JC, et al. (2007) Issues in the
BMJ 330: 753. registration of clinical trials. JAMA 297: 2112–2120.
41. Williamson PR, Gamble C (2005) Identification and impact of outcome 72. Zellars RC, Frassica D, Stearns V, Fetting JH, Armstrong DK, et al. (2006)
selection bias in meta-analysis. Stat Med 24: 1547–1561. Partial breast irradiation (PBI) concurrent with adjuvant dose-dense
42. Johnson NP, Farquhar CM, Hadden WE, Suckling J, Yu Y, et al. (2004) The doxorubicin and dyclophosphamide (ddAC) chemotherapy in early-stage
FLUSH trial: flushing with lipiodol for unexplained (and endometriosis- breast cancer: Preliminary safety results from a feasibility trial. J Clin
related) subfertility by hysterosalpingography: a randomized trial. Hum Oncol, ASCO Annual Meeting Proceedings Part I. Vol 24, No. 18S (June 20
Reprod 19: 2043–2051. Supplement), 2006: 10675.
43. Gluud LL (2006) Bias in clinical intervention research. Am J Epidemiol 163: 73. Lexchin J, Bero LA, Djulbegovic B, Clark O (2003) Pharmaceutical industry
493–501. sponsorship and research outcome and quality: systematic review. BMJ 326:
44. Pildal J, Hrobjartsson A, Jorgensen K, Hilden J, Altman D, et al. (2007) 1167–1170.
Impact of allocation concealment on conclusions drawn from meta- 74. Bero L, Oostvogel F, Bacchetti P, Lee K (2007) Factors associated with
analyses of randomized trials. Int J Epidemiol 36: 847–857. findings of published trials of drug-drug comparisons: why some statins
45. Juni P, Altman DG, Egger M (2001) Systematic reviews in health care: appear more efficacious than others. PLoS Med 4: e184. doi: 10.1371/
assessing the quality of controlled clinical trials. BMJ 323: 42–46. journal.pmed.0040184
46. Schulz KF, Grimes DA (2002) Allocation concealment in randomised trials: 75. Hartley J (2004) Current findings from research on structured abstracts. J
defending against deciphering. Lancet 359: 614–618. Med Libr Assoc 92: 368–371.
47. Scherer RW, Crawley B (1998) Reporting of randomized clinical trial 76. Guimaraes CA (2006) Structured abstracts: narrative review. Acta Cir Bras
descriptors and use of structured abstracts. JAMA 280: 269–272. 21: 263–268.
48. Burns KEA, Adhikari NKJ, Kho M, Meade MO, Patel RV, et al. (2005) 77. Sollaci LB, Pereira MG (2004) The introduction, methods, results, and
Abstract reporting in randomized clinical trials of acute lung injury: an discussion (IMRAD) structure: a fifty-year survey. J Med Libr Assoc 92: 364–
audit and assessment of a quality of reporting score. Crit Care Med 33: 367.
1937–1945. 78. Gulmezoglu AM, Pang T, Horton R, Dickersin K (2005) WHO facilitates
49. Taddio A, Pain T, Fassos FF, Boon H, Ilersich AL, et al. (1994) Quality of international collaboration in setting standards for clinical trial registra-
nonstructured and structured abstracts of original research articles in the tion. Lancet 365: 1829–1831.
British Medical Journal, the Canadian Medical Association Journal and the 79. World Health Organisation (2007) International Clinical Trials Registry
Journal of the American Medical Association. CMAJ 150: 1611–1618. Platform. http://www.who.int/ictrp/en/. Accessed 12 June 2007.

PLoS Medicine | www.plosmedicine.org 0008 January 2008 | Volume 5 | Issue 1 | e20

 CONSORT for Abstracts

80. Laine C, Horton R, Deangelis CD, Drazen JM, Frizelle FA, et al. (2007) 82. National Library of Medicine (2007) MEDLINE / PubMed data element
Clinical trial registration: looking back and moving ahead. BMJ 334: 1177– (field) descriptions. Available: http://www.nlm.nih.gov/bsd/mms/
1178. medlineelements.html. Accessed 16 April 2007.
81. Piaggio G, Elbourne DR, Altman DG, Pocock SJ, Evans SJW (2006) 83. Hill CL, Buchbinder R, Osborne R (2007) Quality of reporting of
Reporting of noninferiority and equivalence randomized trials: an randomized clinical trials in abstracts of the 2005 Annual Meeting of the
extension of the CONSORT statement. JAMA 295: 1152–1160. American College of Rheumatology. J Rheumatol 34: 2476–2480.

PLoS Medicine | www.plosmedicine.org 0009 January 2008 | Volume 5 | Issue 1 | e20