In the mid 1990s, in response to concerns about quality of reporting of randomised controlled trials (RCTs), an international group developed the Consolidated Standards of Reporting Trials (CONSORT) statement.1,2 CONSORT is intended to improve the reporting of an RCT, enabling readers to understand its conduct and to gauge the validity of its results.
The original CONSORT statement, developed for simple two-group parallel RCTs, comprises a checklist and flow diagram.1 Presently, CONSORT includes a 22-item checklist and four-stage flow diagram.2 The items included in the checklist were selected, whenever possible, because empirical evidence indicates that not reporting the information is associated with biased estimates of treatment effectiveness, or because the information is essential to judge the reliability or relevance of the findings.2-6 CONSORT has been endorsed by prominent medical journals and by editorial groups, such as the International Committee of Medical Journal Editors.
Since the publication of CONSORT, several evaluations of its effectiveness have been published.7-10 Two articles have suggested that journal endorsement of CONSORT is associated with improved quality of reporting.7,8 However, other study results have been equivocal.9,10 To gain a broader global perspective of CONSORT’s effectiveness, we conducted a systematic review, examining effectiveness in journals that have formally endorsed CONSORT.
Studies were eligible if they were comparative studies evaluating the quality of RCT reporting; one of the comparator groups included RCTs reported in CONSORT-adopting journals; and outcomes reported included any of the 22 items on the CONSORT checklist, summary scores based on the CONSORT checklist, or any measures of overall trial quality.11 Study quality was not an exclusion criterion, and studies were not excluded based on publication status or language of publication.
The main search strategy was developed in MEDLINE (1996 to 2005 week 28) and customised for EMBASE (1996 to 2005 week 32); both were executed through the Ovid interface. In addition, the Cochrane Methodology Register (up to 2nd Quarter 2005) and the Cochrane Database of Methodology Reviews (up to 2nd Quarter 2005) were initially searched via Update Software in 2004 and then via Wiley InterScience for the 2005 update. The Science Citation Index, Social Sciences Citation Index and Arts & Humanities Citation Index (June 2005) were searched through the ISI Web of Knowledge interface.
Two reviewers (A C P and A M) independently reviewed titles and abstracts using broad screening criteria (comparative study, and outcomes included some measure of trial reporting quality). Articles of potential relevance were further assessed (by A C P and D M) for eligibility. Disagreements were resolved by consensus. The reviewers were not blinded to authors or journals.
When a comparative study with appropriate outcomes was identified, a detailed assessment was undertaken to determine eligibility. First, the study was reviewed to determine the journals and time frame in which the individual RCTs forming the basis for the comparison were published.
A single non-blinded reviewer (A C P) assessed reporting quality using a five-item checklist based on principles of internal and external validity.12 Quality assessment included the use of a control group, RCT cohort ascertainment, how the authors reached agreement on whether a checklist item was included, and whether the reviewers were blinded when assessing outcomes.
Data were extracted by one reviewer (A C P) using a standardised data form, and then checked for accuracy by a second reviewer (D M). Only “consensus” data were used in the review. All data were entered in Review Manager 4.2 (Nordic Cochrane Centre, Copenhagen, Denmark). Dichotomous data were expressed as risk ratios (RR) with 95% confidence intervals. Continuous data were expressed as standardised mean differences with 95% confidence intervals. Consistency across the studies was examined using the I-squared test. Data were pooled using fixed effect models. A priori subgroup analyses included a comparison of RCTs published in journals that specifically require submission of a completed CONSORT checklist with journals that are not CONSORT adopters.
Our search strategy identified 1128 studies; 248 were potentially relevant and eight of them7,8,13-18 met our inclusion criteria (Box 1). One study was included after the author provided unpublished data.17 Six studies that examined changes in the quality of trial reporting over time or compared quality of trial reporting between journals were excluded, either because they did not include a comparator group or were not published in CONSORT-adopting journals,9,19-21 or because insufficient detail was available to determine eligibility.10,22
The number of RCTs in individual studies ranged from 13 to 240, and the number of journals in individual studies ranged from 1 to 68 (Box 2). All studies were published in English. One study used a different definition of CONSORT adoption (journals in which the editor applied the CONSORT checklist were considered CONSORT adopters),7 but 88% of RCTs in the study met our definition of CONSORT adoption. For all other included studies, the definition of CONSORT adopters matched our definition.
Overall, 37 different outcomes were reported (range, 1–16 per study) (Box 3). This variability resulted in most outcomes being reported in only one study. Outcomes used in more than one study were reporting of allocation concealment (five studies), sequence generation (three studies), blinding (four studies), participant flow (two studies) and overall number of CONSORT checklist items (two studies, one using a total of 30 items, the other 40 items). Blinding as an outcome was assessed as reported or not in one study,17 and blinding of participants and data analysts in three studies.7,17,18
Box 2 summarises the methodological quality of the studies. All studies were quasi-experimental, with only one study having an a priori defined control group.8
CONSORT-adopting journals had better reporting of the method of sequence generation and allocation concealment (Box 4A). For the two studies that reported total number of CONSORT items, the standardised mean difference was 0.83 (95% CI, 0.46–1.19). CONSORT adoption appeared to have less effect on reporting of participant flow and reporting of blinding of participants or data analysts.
The descriptions of the method of sequence generation, participant flow and total CONSORT items (standardised mean difference, 3.67 items; 95% CI, 2.09–5.25) were better after adoption of CONSORT (Box 4B). CONSORT adoption appeared to have less effect on allocation concealment. The paucity of data precluded meaningful examination of publication bias. There was no evidence of heterogeneity.
Studies evaluating the effectiveness of the CONSORT checklist are methodologically weak. The eight studies we identified used different quasi-experimental designs, which (with one exception) did not include a control group, addressed different questions, and used a large number of discrete outcomes, negating the possibility of combining them — a strength of the systematic review process. Further, analysis of studies that sample journals should take account of the clustering effect in the analysis, which would widen the confidence intervals. Only one study included in the review considered this effect. Nevertheless, our review suggests some improvement in the quality of reporting RCTs when the CONSORT checklist is used. Although the degree of improvement in quality of reporting of particular trial items differs somewhat between our comparison groups (CONSORT adopters v non-adopters after CONSORT publication, and CONSORT adopters before and after CONSORT publication), the direction of the effect remains the same. The difference in magnitude may be related to the small number of studies.
Our results are encouraging, but provide no definitive answer as to whether the CONSORT checklist improves reporting of RCTs. Perhaps unique to our review is ascertaining whether the RCTs were published in journals that adhered to CONSORT. If journals do not enforce the use of the checklist, it is likely that the effects we observed underestimate the possible benefits with proper enforcement of CONSORT. A review of 167 high-impact journals found that only 22% mentioned CONSORT in their instructions to authors, and 25% of these referred to the obsolete 1996 version.24 In another study involving 15 high-impact journals (five of which were not included in the previously mentioned review) that have reported endorsing CONSORT, only eight referred to the statement in their instructions to authors.25 CONSORT-adopting journals should be more proactive in enforcing adherence to CONSORT.
Although we did not find a strong influence of CONSORT on the quality of reporting of blinding, a recent study has suggested that the CONSORT checklist may have had a positive influence. In a study of RCTs in journals that adopted the original and then the revised CONSORT checklist, the authors found that the quality of reporting of blinding improved by 23%–55% between publication of the two checklists.26 This is particularly important in light of emerging data suggesting investigators, educators and readers vary greatly in their interpretations and definitions of types of blinding.27
2 Characteristics and quality of included studies
Jefferson Smurfit Foundation, Ireland; US National Library of Medicine; Merck; Glaxo Wellcome; CIHR |
|||||||||||||||
Received 7 December 2005, accepted 22 May 2006
Abstract
Objective: To determine whether the adoption of the CONSORT checklist is associated with improvement in the quality of reporting of randomised controlled trials (RCTs).
Data sources: MEDLINE, EMBASE, Cochrane CENTRAL, and reference lists of included studies and of experts were searched to identify eligible studies published between 1996 and 2005.
Study selection: Studies were eligible if they (a) compared CONSORT-adopting and non-adopting journals after the publication of CONSORT, (b) compared CONSORT adopters before and after publication of CONSORT, or (c) a combination of (a) and (b). Outcomes examined included reports for any of the 22 items on the CONSORT checklist or overall trial quality.
Data synthesis: 1128 studies were retrieved, of which 248 were considered possibly relevant. Eight studies were included in the review. CONSORT adopters had significantly better reporting of the method of sequence generation (risk ratio [RR], 1.67; 95% CI, 1.19–2.33), allocation concealment (RR, 1.66; 95% CI, 1.37–2.00) and overall number of CONSORT items than non-adopters (standardised mean difference, 0.83; 95% CI, 0.46–1.19). CONSORT adoption had less effect on reporting of participant flow (RR, 1.14; 95% CI, 0.89–1.46) and blinding of participants (RR, 1.09; 95% CI, 0.84–1.43) or data analysts (RR, 5.44; 95% CI, 0.73–36.87). In studies examining CONSORT-adopting journals before and after the publication of CONSORT, description of the method of sequence generation (RR, 2.78; 95% CI, 1.78–4.33), participant flow (RR, 8.06; 95% CI, 4.10–15.83), and total CONSORT items (standardised mean difference, 3.67 items; 95% CI, 2.09–5.25) were improved after adoption of CONSORT by the journal.
Conclusions: Journal adoption of CONSORT is associated with improved reporting of RCTs.