SEARCH

SEARCH BY CITATION

Keywords:

  • Immunosuppression;
  • kidney transplantation;
  • outcome reporting;
  • randomized trial;
  • reporting bias;
  • systematic review

Abstract

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Inconsistent and incomplete outcome reporting may make estimates of treatment effects from published randomized trials unreliable. We aimed to determine outcome reporting practices and source of differences in reporting quality among randomized trials of primary immunosuppression in kidney transplantation. We searched the Cochrane Renal Group's Specialized Register, 2000–2012, specified four core outcomes we expected trials to report, and recorded if and how completely each was reported. We identified 179 trials. One hundred sixty-eight (94%) reported death, 145 (81%) as number dead and 119 (66%) as time to death. One hundred sixty-five (92%) reported graft loss, 158 (88%) as number with graft loss and 127 (71%) as time to graft loss. One hundred twenty-one (68%) reported creatinine and 114 (64%) estimated GFR (eGFR). One hundred forty-one (79%) provided complete reports of number dead, 95 (53%) censored and 99 (55%) uncensored number with graft loss. Seventy-three (41%) provided complete reports of time to death, 67 (37%) censored and 31 (17%) uncensored time to graft loss. Complete reporting of graft function was infrequent: 62 (35%) eGFR and 50 (28%) creatinine. All four outcomes were reported in any form in 61 (34%) and completely in 28 (16%) trials. No single trial or journal characteristic was consistently associated with complete outcome reporting. Outcome reporting in kidney transplant trials is inconsistent and frequently incomplete, and published estimates of treatment effects may be unreliable.


Abbreviations
CI

confidence interval

CNI

calcineurin inhibitor

COMET

core outcome measures in effectiveness trial

CONSORT

consolidated standards of reporting trials

eGFR

estimated GFR

OMERACT

outcome measures in rheumatology

OR

odds ratio

ORBIT

outcomes reporting bias in trials

Background

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Evidence-informed decision-making requires robust estimates of the relative benefits and harms of different treatments based upon publication of consistent, complete and comparable trial data. Two main sources of outcome reporting bias exist, both of which may overestimate treatment effects: selective reporting of outcomes on the basis of their statistical significance, and incomplete outcome reporting, which occurs when an outcome is reported as having been measured and analyzed but is then either not reported (because of unfavorable result) or reported in insufficient detail to allow estimation of the precision and size of treatment effect [1, 2]. Outcome reporting bias is prevalent in a wide range of specialty-specific interventional trials where both unappreciated harms and overestimated benefits of treatments are reported consequences [3-5].

In kidney transplant trials, outcome reporting has received little attention. Use of inconsistent kidney function end points and variable definitions of acute rejection episodes and graft survival have been described, but the extent of selective and incomplete reporting and consequences of missing outcome data on the precision and magnitude of estimates of effect of different treatments are unknown [6-8].

Immunosuppression in kidney transplant recipients involves the selection and use of induction and maintenance phase drugs in a coordinated way. Each drug has specific benefits and harms, their use varying among patient groups depending upon healthcare resources, clinicians' perceptions of treatment effects and estimates of recipient risk of drug-related complications [4]. Consistent and complete reporting of core outcomes in trials of primary immunosuppression would allow generation of reliable estimates of the relative effects of different treatments and help provide clarity to clinicians in such a complex therapeutic area.

We aimed to describe the consistency and completeness of reporting for four core outcomes reported in trials of primary immunosuppression in kidney transplantation and to identify possible sources of differences in reporting completeness.

Methods

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Selection criteria

We searched the Specialized Register of the Cochrane Renal Group for reports of contemporary randomized trials in kidney transplantation published in peer-reviewed journals from 2000 until July 2012. We limited to English language reports, and included only trial reports of primary immunosuppression interventions in kidney recipients, excluding reports where recipients received another organ in addition to a kidney, or where the unit of randomization was the donor organ and not the recipient. We also excluded switch trials where recipients were randomized at some time distant from time of transplantation to altered immunosuppression. These exclusion criteria were applied to establish a cohort of trials that could be reasonably expected to be concerned with reporting similar outcomes within broadly similar time frames. We selected the earliest report of the full clinical results of a trial for inclusion.

Identification of trial reports

The Specialized Register is a register of all known reports of randomized trials in nephrology organized by trial. It contains reports identified from monthly searches of the Cochrane Central Register of Controlled Trials CENTRAL, weekly searches of MEDLINE OVID SP and searches of the current year of Embase OVID SP, and has been used in multiple published systematic reviews of immunosuppression. Hand searching of renal-related journals and the proceedings of major renal conferences, weekly current awareness alerts for selected renal journals and searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov are also incorporated.

We judged that from the year 2000 trials should be reported in accordance with international guidelines including the consolidated standards of reporting trials (CONSORT) statement (http://www.consort-statement.org/) and International Committee of Medical Journal Editors' uniform requirements for manuscripts (http://www.icmje.org/).

Outcome analysis

Using a standardized data abstraction spread sheet, two reviewers independently examined each trial report and recorded key descriptive data. We collected details of year and journal of publication, drug-class comparison, trial location, phase, size, duration of follow-up and sponsorship, and journal impact factor and CONSORT endorsement at date of publication. Journal CONSORT endorsement required adoption of the statement at least 6 months before trial publication so as to allow for lag between manuscript submission and publication [9]. We contacted journal editorial offices to determine date of CONSORT adoption. Where no reply was obtained we reviewed previous journal issues to determine author guidelines.

We recorded absolute measures of function: patient death and time to death, graft loss and time to loss, and measures of relative function; GFR measured directly, or by approximation of GFR with urinary creatinine clearance, or estimated by formulae from serum creatinine (estimated GFR [eGFR]). Relative measures of function are increasingly being used as surrogate markers of absolute end points in kidney transplant trials despite lacking proper validation, and so a selection of both types of measures was made [10]. Where graft loss or time to loss was reported, we determined whether it included recipients who lost their graft and returned to dialysis together with those who died with a functioning transplant, or included only those who lost their graft and returned to dialysis, but not those dying with a functioning transplant (“censoring” for death with a functioning graft). Where GFR was directly measured, we recorded which chemical method was used, and where GFR was estimated, we recorded which formula was used to calculate eGFR.

Where outcomes were reported, we examined completeness of the published data using the outcomes reporting bias in trials (ORBIT) classification system, adapted for our study [1]. Specifically, where outcomes were calculated using time-to-event data, to be graded as providing a complete report, the number in each group at the start of analysis, the proportion surviving at a fixed time point and either the number remaining at risk or a test of difference between the groups had to be stated, or alternatively, a hazard ratio with confidence interval (CI) or p-value presented. For continuous outcomes such as creatinine and eGFR, we expected that the number of participants contributing measurements, the point estimate result (e.g. mean or median) and an estimate of precision (e.g. standard deviation or error) would be stated. For dichotomous outcomes, such as number of patient deaths, we expected the numerator and denominator in each group to be stated. We then assessed the association between the frequency of core outcome reporting and the statistical significance of the results reported in included trials. Specifically, where a core outcome was reported, we determined the proportion of trials where a statistically significant difference between treatment arms was reported for that outcome (either in words “a statistically significant result was found” or quantitatively with a p-value of p < 0.05), and compared this with the proportion of all trials reporting that outcome. We hypothesized that we would find an inverse relationship, with the proportion of trials reporting a statistically significant association increasing with decreasing likelihood of complete data reporting, suggesting outcome reporting bias [11].

We assessed potential predictors of complete reporting using logistic regression for the outcomes of patient death or time to death, graft loss or time to loss, serum creatinine and eGFR. Potential predictors considered included publication year, journal impact factor (categorized as <3, 3–5.9, ≥6, adjusted for year of publication), journal endorsement of CONSORT guidelines at date of publication, geographical location of trial, number of trial participants (categorized <100, 100–199, ≥200) and funding source of the trial. Analysis was undertaken using STATA version 11.2 (StataCorp, College Station, TX) using the backwards stepwise approach, a likelihood ratio test significance of <0.05 warranting inclusion in the final model.

Results

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

We identified 179 eligible trials of primary immunosuppression in kidney transplant recipients during the study period (Figure 1). Trial characteristics are described in Table 1. All were parallel design. Sixty-three (35%) did not declare their source of funding or were unfunded, 102 (57%) were funded by pharmaceutical sponsorship and 14 (8%) declared funding from a public body. Thirty (17%) were multinational, recruiting participants from more than one continent, and 46 (26%) recruited participants from North America, 79 (44%) from Europe, 13 (7%) the Asia-Pacific region, and 11 (6%) from elsewhere. One hundred forty-one (79%) were published in transplantation journals, 22 (12%) in nephrology journals and 16 (9%) in general medicine or other journals. One hundred twenty (67%) were published in journals that currently endorse the CONSORT statement. Sixty-two trials (34%) compared a calcineurin inhibitor (CNI) with another immunosuppressive agent, the most frequent drug class comparison. Trial participant numbers varied from 15 to 1645. There were only four (2%) trials that reported results at 1 month or earlier, and 13 (7%) that did so before 6 months with duration of follow-up ranging from 0.25 to 240 months. One hundred sixty-three (91%) were phase III clinical trials.

image

Figure 1. Identification and inclusion of trial reports.

Download figure to PowerPoint

Table 1. Characteristics of included trials and publishing journal
CharacteristicNumber of trials, total = 179
N%
  1. CONSORT, consolidated standards of reporting trials.

Trial
Year of publication
2000–20011911
2002–20032112
2004–20053419
2006–20073720.5
2008–20093218
2010–20113419
201220.5
Number of participants
<1006436
100–2005330
>2006234
Drug class investigated
Antibody induction4626
Antimetabolite3017
Calcineurin inhibitor (CNI)6234
Other4123
Location
North America4626
Europe7944
Asia-Pacific137
Multinational3017
Other116
Funding source
Not stated/unfunded6335
Pharmaceutical industry10257
Public funder148
Phase of trial
II127
III16391
IV42
Duration of follow-up
≤1 month42
>1 to <6 months95
6 to ≤12 months12570
>12 months4133
Statistical significance of reported outcomes
Significant4022
Not significant10659
No test of significance3318
Journal
Type
Transplantation14179
General nephrology2212
Other169
Impact factor
<35330
3–5.98246
≥64424
Current endorser of CONSORT guidelines
Yes12067
No5933

Number and type of outcomes reported

Death and graft loss

Overall, 168 (94%) reported a measure of death or time to death, 119 (66%) reported time to death, 145 (81%) reported number dead and 96 (53%) reported both (Figure 2, Table S1).

image

Figure 2. Completeness of reporting of core outcomes among included trial reports.

Download figure to PowerPoint

One hundred sixty–five (92%) reported at least one measure of graft loss or time to loss, 143 (80%) reported number with graft loss and 127 (71%) reported time to graft loss. Of trials reporting number with graft loss, 123 (69%) clearly defined whether graft loss included or was censored for death with a functioning graft. Another 15 (8%) reports provided sufficient information in text or tables for the definition of graft loss to be assumed, the definition of graft loss remaining completely unclear in 5 (3%) reports. Fifty-eight (32%) reported number with graft loss both including death with a functioning graft and censored for death with function. Of trials reporting time to graft loss, whether or not graft loss was censored for death with a functioning graft was clear in 72 (40%), able to be assumed in 46 (26%) but completely unclear in 9 (5%) reports. Twenty-eight (16%) reported time to graft loss both including death with a functioning graft and censored for death with function. Fifteen (8%) reported all four possible measures of graft loss and time to loss, both censored and uncensored for death with a functioning graft.

Graft function

Fourteen (8%) trials reported measured approximations of GFR. Seven (4%) trials reported 24 h measured urinary creatinine clearance, and eight (4.5%) used an exogenous filtration marker.

One hundred twenty-one (68%) reported any measure of serum creatinine, most often as the mean end point value (98 trials, 55%). Thirteen (7%) trials reported more than one measure of creatinine, with seven (4%) reporting both the mean and median end point creatinine values. Two (1%) trials reported three measures: mean and median end point, and rate of change.

eGFR was reported in 114 (64%) trials. Ten (5.5%) trials did not make explicit the equation used for calculation of eGFR. Sixty-six (37%) reported the Cockcroft–Gault formula, 23 (13%) the MDRD, 26 (15%) the Nankivell and five (3%) the Schwartz formula. Two (1%) trials reported eGFR using all of the Nankivell, MDRD and Cockcroft–Gault formulae. Five (3%) further trials reported eGFR by both Cockcroft–Gault and MDRD, and five (3%) more by Cockcroft–Gault and Nankivell. Trials most frequently reported the mean end point eGFR (90 trials, 50%).

Completeness of outcomes reported

Any report of all four core outcomes was found in 61 (34%) trials, though only 28 (16%) reported this number completely (Figure 3). Only one (0.5%) trial failed to provide any report of at least one core outcome, though 10 (5.5%) reported none of the core outcomes completely.

image

Figure 3. Prevalence of any reports and complete reports of core outcomes among included trial reports.

Download figure to PowerPoint

Overall, dichotomous outcomes—number dead and number with graft loss—were more completely reported than the time-to-event outcomes time to death and time to graft loss (Figure 2). One hundred forty-one (79%) presented complete reports of number dead, four (2%) presented incomplete data and one (0.5%) stated in methods that the outcome was analyzed but subsequently gave no result (Table S2). In 33 (18%) trials the number dead was not mentioned even though likely measured. For time to death, 73 (41%) gave complete reports, 46 (26%) incomplete reports and six (3%) stated that the outcome was analyzed but provided no subsequent results. In 47 (26%), time to death was not mentioned even though likely to have been measured. A further seven (39%) must have measured time to death because the related outcome of time to graft loss including death with function was reported, thereby implying that a record of deaths was kept. Incomplete data typically arose as although the number in each trial arm and the proportion of patients surviving at a given time point were quoted, either the number of people remaining alive (uncensored from the survival analysis) was not clear (three trials, 1.5%), or there were no CIs given or no statistical test for difference between the groups (nine trials, 5%), or both (32 trials, 18%).

Similar patterns of reporting were observed when examining number with graft loss, and time to graft loss for both censored and uncensored end points (Figure 2).

Where continuous measures of serum creatinine and eGFR were used, only a minority of trials reported complete data: 50 (28%) and 63 (35%) trials, respectively. Twenty-nine (16%) trials reported calculated eGFR, yet did not report creatinine that must have been collected for estimation of GFR. Data were incomplete for 72 (40%) trials reporting creatinine, and 52 (29%) trials reporting eGFR. Most commonly, the number of participants contributing measurements was not reported (65 trials, 36% for creatinine, 34 trials, 19% for eGFR), which cannot be assumed from the number of participants randomized to each arm as these outcomes are dependent upon the participant still being alive and available to provide a blood sample at the nominated time point. In 34 (19%) trials for creatinine and 24 (13%) for eGFR, no measure of precision (standard deviation or standard error) was presented. Both of these—number of participants contributing data and the estimate of variability—were missing in 26 (15%) for creatinine and 18 (10%) for eGFR.

Predictors of complete reporting

Overall, 157 (88%) trials reported complete data for either death or time to death (Table S3). Larger trials more commonly completely reported death (p = 0.01). Trials sponsored by pharmaceutical industry more commonly reported death completely than those not stating their source of funding or explicitly unfunded (odds ratio [OR] 5.45 [1.81–16.37]) (Figure 4). Trials funded by a public funder all reported death completely. One hundred twenty trials (67%) reported complete data for either number with graft loss or time to graft loss. Larger trials were more likely to completely report graft loss (p = 0.005). Serum creatinine was completely reported in 50 (28%) trials. Trials reported in higher impact factor journals were more likely to report creatinine completely (p = 0.02). eGFR was completely reported in 62 (35%) trials, more commonly in recently published reports (OR 1.22 [1.11–1.36] per year after 2000). Analyzing the effect of time as a categorical variable stratified into two eras (2000–2005 and 2006–2012) did not change the results of our analyses of the predictors of complete reporting for any core outcome (data not shown).

image

Figure 4. Predictors of complete data reporting of patient and graft function outcomes.

Download figure to PowerPoint

Statistical significance of outcomes reported

We observed an inverse relationship between the likelihood of reporting a statistically significant result and likelihood of complete reporting of outcomes. Despite being reported in a smaller proportion of trials, a statistically significant result was reported in a larger proportion of trials giving any report of creatinine (19%) or eGFR (25%) than trials reporting death (0.5%) or graft loss (4%). Complete reports of core outcomes were more likely to report a statistically significant difference between treatment arms than any reports of core outcomes (Figure 5).

image

Figure 5. Association between frequency of outcome reporting and statistical significance of results.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Summary of findings

We have demonstrated a wide variation in the consistency of reported outcome measures, and in the completeness of data provided in contemporary kidney transplant trials of primary immunosuppression. Reporting varied in completeness and consistency by which core outcome was measured, and by whether it was reported as dichotomous, continuous or time-to-event data. Death and graft loss were more frequently and more completely reported as dichotomous outcomes than time-to-event data. Graft function was estimated or measured using numerous algorithms and methods and reported by an even greater number of end points. Reasons for incomplete outcome reporting varied. For time-to-event data, the number of people remaining “at risk” was frequently unclear. For continuous data, the number of participants contributing data to specific outcomes measured at defined time points was often lacking. Incomplete reporting of CIs or tests of statistical significance of between-group differences was common. No single source of difference in completeness of outcome reporting among trials was evident. The inverse relationship we observed between the likelihood of reporting a statistically significant result and the likelihood of complete reporting of core outcomes suggests that the statistical significance of a result may bias its reporting among our cohort of trials.

We propose that current reporting guidelines are ineffective at ensuring complete outcome data reporting and that other initiatives are needed to improve the reliability of the evidence base for treatment decisions for kidney transplant recipients.

Strengths and limitations of this study

In identifying included trials, we assumed that all trials testing one immunosuppressive regimen against another should be concerned with similar outcomes. The core outcome measures selected are the most fundamental measures of transplantation success, and propose that it would be hard to justify conducting any trial of immunosuppression without collecting them. We included only the first full publication of a trial's results, as we thought that this should provide full reports of core outcomes, but did not check subsequent publications for supplementary reporting. The tool we used to classify selective outcome reporting—the ORBIT classification system—is a sensitive and specific validated tool for classifying missing or incomplete outcome data in reports of randomized controlled trials [1]. We did not compare published reports with trial protocols and so could not quantify selective outcome reporting bias within trials, as our research question was more concerned with comparing outcome reporting patterns among trials. In addition, we did not investigate outcome reporting in relation to how well other aspects of the trial were reported, for example the study design or methodology. Some reports were likely published in journals that adhered to reporting guidelines—even if not explicitly so at the time of publication, so whether or not the publishing journal endorsed the CONSORT statement at time of publication may be subject to misclassification bias in our study. Several journals did not reply to our request for information and it was not possible to determine the date on which they adopted the CONSORT guidelines from prior journal issues, and so nonresponse bias may also be present. Last, although these trials are recently published, for some, their design and conduct may have significantly pre-dated contemporary guidelines for reporting and may not be representative of the current conduct of all clinical trials.

Interpretation and implications in the context of the totality of current evidence

Outcome reporting bias has previously been investigated using cohorts of randomized controlled trials drawn from biomedical databases, specific journals, government-funded trials, trials approved by a specific ethics committee and a cohort of Cochrane systematic reviews [1, 11-15]. This work established that outcomes reported in trials are often inconsistent with those stated in trial protocols, and that complete data reporting is biased toward statistically significant outcomes [1, 5]. Among trials of immunosuppression in kidney transplantation, overall study quality and reporting quality of acute rejection, patient and graft survival outcomes have been assessed using the Jadad scale and CONSORT statement, respectively [7]. Variability in the definition of acute rejection, prevalence of use of specific kidney function end points and the power of studies to detect clinically meaningful differences in effect for reported outcomes have also been examined [6, 8, 10]. Our study focuses on the consistency and completeness of reporting of specific core trial outcomes using the ORBIT risk of bias tool in assessing outcome reporting for the first time in any single medical specialty. We provide the first assessment of the completeness of reporting of kidney function outcomes, despite their increasingly widespread use as surrogate end points in kidney transplant trials [6]. Our work also re-assesses the impact of CONSORT guidelines on outcome reporting in kidney transplantation for the first time in almost 10 years, the lack of improvement in reporting quality observed over that time period despite the widespread endorsement of these guidelines justifying our call for urgent action [7].

Assuming that investigators were concerned with asking broadly the same questions, how has inconsistent and incomplete outcome reporting occurred despite guidelines in the CONSORT statement [16]? First, it is possible that where clear efficacy of an immunosuppressive therapy has already been established in prior studies, the main study objective may be to analyze a variable unrelated to our proposed core outcomes. However, this is not the full explanation, interviews with trialists reporting that investigators are largely unaware of the implications of not reporting all outcomes fully [17]. Lack of consensus regarding choice of outcomes in particular clinical settings, journal space restrictions, the perceived importance of the result and its statistical significance are also reasons cited for inconsistent and incomplete reporting [12]. We observed that core outcomes that were reported less often were more likely to report a statistically significant difference between treatment arms than core outcomes reported more frequently, suggesting the presence of selection bias (selective reporting of outcomes based upon statistical significance). Quality of core outcome reporting also varied by statistical significance of the trial result: complete reports of core outcomes were less frequent than any report of core outcomes, but were more likely to report a statistically significant result. Our data also suggest that neither study duration nor the clinical phase of the trial explains differences in completeness of reporting.

Data omissions threaten the reliability of estimates of effect reported in trials. Readers may assume a “best case scenario” where intervention differences are overestimated; by assuming all eligible trial participants contributed measurements to the outcome “mean eGFR,” the precision of the difference improves, leading to likelihood of perceived statistical differences. Statistically significant outcomes are estimated to be more than twice as likely to be fully reported compared with nonsignificant outcomes regardless of funding source, sample size or number of study centers [11]. Such selective and incomplete data presentation also threatens the validity of the conclusions of meta-analyses with up to 25% of Cochrane systematic reviews thought to overestimate treatment effect by 20% or more [1].

Initiatives aiming to increase transparency and improve consistency of outcome reporting in specialty-specific trials have recently emerged. Core outcome measures in effectiveness trials (COMET) is an international initiative bringing together key researchers, patients and policy-makers to develop a consistent core set of outcomes that should be measured and reported, as a minimum, in all clinical trials of a specific condition or intervention [18]. Outcome measures in rheumatology (OMERACT) is a specialty-specific collaboration that has reduced the disparity and number of primary outcomes reported in trials of rheumatoid arthritis, demonstrating the potential advantages of such a collaborative process [19]. Such consensus among kidney transplant trialists would surely address any perceived lack of importance for core outcomes.

Our findings also have implications for journal editors and trialists. Journal CONSORT guideline endorsement has improved the overall quality of reporting of trials, though with equivocal effect on the complete presentation of estimates of effect and precision [20]. Industry sponsorship is associated with complete data reporting in trials in other specialties, but also with selective reporting of outcomes favorable to the sponsor and so alone cannot be assumed to predict low risk of reporting bias. Contrary to previous studies of kidney transplant trials, we observed more complete reporting in larger trials, in keeping with reports from other specialties [7, 21, 22].

Future research

This review aimed to provide a benchmark of current reporting practice from which to measure future improvement: akin to clinical governance, research needs quality control to enhance scientific endeavor. Trial registration has helped improve transparency, but may not be sufficient alone to overcome outcome reporting bias in clinical research [23]. A pragmatic way forward would be to emulate the OMERACT initiative. This would require international agreement of a standardized “core outcome set” with clear definitions and consensus regarding methods used to measure them for all published trials in kidney transplantation. Minimum magnitude of differences judged to be clinically important could also be standardized for key outcomes so as to further enhance between-trial comparisons. Furthermore, existing strategies—particularly the updated CONSORT statement—should be more widely publicized and strictly enforced by journal editors who have an ethical responsibility to improve the reporting quality of randomized trials and ensure publication of representative, undistorted data on which clinicians can base clinical decisions.

The authors are happy to provide full details of research methods, data sets and statistical analysis code for this study to any interested party.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Gail Higgins, who contributed to the design of the study and identified trial reports for inclusion. Matthew Bottomley and Dan Owers, who identified trial reports for inclusion and helped extract data. Partial preliminary reports of this analysis were presented at the 6th International Congress on Peer Review and Biomedical Publication, Vancouver 2009, the 27th and 30th Annual Scientific Meetings of The Transplantation Society of Australia and New Zealand, Canberra 2009 and 2013, and the American Transplant Congress, Seattle 2013.

Authors' contributions

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

PM identified trial reports for inclusion, extracted, prepared and analyzed the data, interpreted the results, and wrote and revised the manuscript. FAD identified trial reports for inclusion, extracted the data and revised the manuscript. LPR identified trial reports for inclusion, extracted the data, prepared and analyzed the data. PJK contributed to the analysis of the data and revision of the manuscript. AM contributed to the preparation and analysis of the data. JCC contributed to the design of the study, advised on interpretation and presentation of the results, and revision of the manuscript. ACW conceived and designed the study, identified trial reports for inclusion, extracted, prepared and analyzed the data, interpreted the results, wrote and revised the manuscript.

Disclosure

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

References

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information
  • 1
    Kirkham JJ, Dwan KM, Altman DG, et al. The impact of outcome reporting bias in randomised controlled trials on a cohort of systematic reviews. Br Med J 2010; 340: c365.
  • 2
    Mathieu S, Boutron I, Moher D, Altman DG, Ravaud P. Comparison of registered and published primary outcomes in randomized controlled trials. J Am Med Assoc 2009; 302: 977984.
  • 3
    Turner EH. A taxpayer-funded clinical trials registry and results database. PLoS Med 2004; 1: e60.
  • 4
    Webster A, Woodroffe RC, Taylor RS, Chapman JR, Craig JC. Tacrolimus versus cyclosporin as primary immunosuppression for kidney transplant recipients. Cochrane Database Syst Rev 2005; 4: CD003961.
  • 5
    Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007; 356: 24572471.
  • 6
    White CA, Siegal D, Akbari A, Knoll GA. Use of kidney function end points in kidney transplant trials: A systematic review. Am J Kidney Dis 2010; 56: 11401157.
  • 7
    Fritsche L, Einecke G, Fleiner F, Dragun D, Neumayer HH, Budde K. Reports of large immunosuppression trials in kidney transplantation: Room for improvement. Am J Transplant 2004; 4: 738743.
  • 8
    Fleiner F, Fritsche L, Glander P, Neumayer H-H, Budde K. Reporting of rejection after renal transplantation in large immunosuppressive trials: Biopsy-proven, clinical, presumed, or treated rejection? Transplantation 2006; 81: 655659.
  • 9
    Plint AC, Moher D, Morrison A, et al. Does the CONSORT checklist improve the quality of reports of randomised controlled trials? A systematic review. Med J Aust 2006; 185: 263267.
  • 10
    Ibrahim A, Garg AX, Knoll GA, Akbari A, White CA. Kidney function endpoints in kidney transplant trials: A struggle for power. Am J Transplant 2013; 13: 707713.
  • 11
    Chan A-W, Hróbjartsson A, Haahr MT, Gøtzsche PC, Altman DG. Empirical evidence for selective reporting of outcomes in randomized trials. J Am Med Assoc 2004; 291: 24572465.
  • 12
    Chan A-W, Altman DG. Identifying outcome reporting bias in randomised trials on PubMed: Review of publications and survey of authors. Br Med J 2005; 330: 753.
  • 13
    Chan A-W, Krleža-Jerić K, Schmid I, Altman DG. Outcome reporting bias in randomized trials funded by the Canadian Institutes of Health Research. Can Med Assoc J 2004; 171: 735740.
  • 14
    Al-Marzouki S, Roberts I, Evans S, Marshall T. Selective reporting in clinical trials: Analysis of trial protocols accepted by the Lancet. Lancet 2008; 372: 201.
  • 15
    Brooks RJ, Higgins GY, Webster AC. Systematic review of randomized controlled trial quality in pediatric kidney transplantation. Pediatr Nephrol 2010; 25: 23832392.
  • 16
    Hopewell S, Altman DG, Moher D, Schulz KF. Endorsement of the CONSORT Statement by high impact factor medical journals: A survey of journal editors and journal ‘Instructions to Authors’. Trials 2008; 9: 20.
  • 17
    Smyth R, Kirkham J, Jacoby A, Altman D, Gamble C, Williamson P. Frequency and reasons for outcome reporting bias in clinical trials: Interviews with trialists. Br Med J 2011; 342: c7153.
  • 18
    Williamson P, Altman D, Blazeby J, Clarke M, Gargon E. The COMET (core outcome measures in effectiveness trials) initiative. Trials 2011; 12: A70.
  • 19
    Tugwell P, Boers M, Brooks P, Simon L, Strand V, Idzerda L. OMERACT: An international initiative to improve outcome measurement in rheumatology. Trials 2007; 8: 38.
  • 20
    Turner L, Shamseer L, Altman DG, et al. Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials (RCTs) published in medical journals. Cochrane Database Syst Rev 2012; 11: MR000030.
  • 21
    Bath PM, Iddenden R, Bath FJ. Low-molecular-weight heparins and heparinoids in acute ischemic stroke: A meta-analysis of randomized controlled trials. Stroke 2000; 31: 17701778.
  • 22
    Farrokhyar F, Chu R, Whitlock R, Thabane L. A systematic review of the quality of publications reporting coronary artery bypass grafting trials. Can J Surg 2007; 50: 266277.
  • 23
    McGee RG, Su M, Kelly PJ, Higgins GY, Craig JC, Webster AC. Trial registration and declaration of registration by authors of randomized controlled trials. Transplantation 2011; 92: 10941100.

Supporting Information

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Authors' contributions
  9. Disclosure
  10. References
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ajt12444-sup-0001-SuppData.pdf522K

Table S1: Outcomes reported in included trial reports.

Table S2: Completeness of outcome data reporting of patient and graft function measures in included trial reports.

Table S3: Predictors of complete data reporting of patient and graft function outcomes.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.