Background
Clinical trials, when appropriately designed, conducted, and reported, are the gold standard study design for generating evidence about treatment efficacy, safety, effectiveness, and efficiency. To be able to critically evaluate and use the results of a trial, however, readers require complete, clear, and transparent information with respect to what was planned, what was done, and what was found [
1]. Inadequate reporting of clinical trials is well documented in the medical literature, even with respect to basic methodological details such as the definition of the primary outcome, specification of who was blinded, and explanation of how trial sample size was calculated [
2‐
4]. Such incomplete reporting contributes to significant and avoidable waste of health research investment and impedes reproducibility [
5,
6].
In an effort to improve trial reporting quality, reporting guidelines have been developed to standardize the reporting of clinical trial reports and their corresponding protocols. These guidelines include Consolidated Standards of Reporting Trials (CONSORT) [
7], first developed in 1996 and updated in 2010, for trial reports published in academic journals, and Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT), developed in 2013, for trial protocols [
8]. Numerous extensions have since been developed for CONSORT and SPIRIT, refining their applications to specific populations, study designs, interventions, and contexts [
9]. There is evidence that completeness of trial reporting has improved since the development of CONSORT, as shown by increased reporting of CONSORT checklist items, particularly in journals that have endorsed the guideline [
10,
11]. CONSORT is now endorsed by more than half of core medical journals, including
The Lancet,
BMJ,
JAMA, and the
New England Journal of Medicine as well as field-specific journals [
12]. However, endorsement policies are not always clear; thus increased effort by journals with respect to enforcement of completion and evaluation of guideline adherence may help further improve the current state of suboptimal trial reporting [
10].
Despite the availability and implementation of these well-established reporting guidelines for trials, significant concerns regarding the quality of the reporting of trial outcomes remain [
13‐
18]. In the context of a clinical trial, an outcome refers to what is being measured on trial participants to examine the effect of exposure to a health intervention [
19]. SPIRIT and CONSORT provide some general guidance on how to report outcomes [
7,
8], including pre-defined primary and secondary outcomes, method of outcome assessment, and timing of outcome assessment. However, outcome reporting remains insufficient across disciplines and academic journals; key information about the selection process, definition, measurement, and analysis of primary outcomes is often missing or poorly reported [
2,
3,
13,
20‐
24]. It has been estimated that up to 60% of trials change, introduce, or omit a primary outcome between protocol and publication of the trial report [
20,
25‐
27], when a protocol is even available for comparison. Less is known about secondary outcomes, which may be even more prone to bias and inadequate reporting [
17]. As clinical trials are “only as credible as their outcomes” [
28], this lack of transparency and completeness reduces or prevents reproducibility, critical appraisal, knowledge synthesis, and uptake of trial results into clinical practice. Moreover, it enables the introduction of bias into the medical literature by facilitating selective reporting and outcome switching.
Although calls for improved reporting of trial outcomes have been made [
13,
14,
16,
29], to date it is unknown what actually constitutes useful, complete reporting of trial outcomes to knowledge users. No evidence- and/or consensus-based detailed guidance currently exists for authors to follow to ensure that their reporting is complete, transparent, and replicable. SPIRIT requires more information on trial outcomes to be reported than CONSORT, but neither reflects, for example, the increasingly widespread attempts to incorporate the patient voice into clinical research. There is no requirement to report why an outcome was selected, to provide a rationale for the way the outcome was defined, or to describe the acceptability to patients of measuring the chosen outcome. The advent of SPIRIT and CONSORT extensions for patient-reported outcomes (PROs) [
30,
31] as well as a CONSORT extension for harms [
32] represent important steps in improving the reporting of trial outcomes. Recently published guidelines for the content of statistical analysis plans (SAPs) are also now available [
33]. However, more comprehensive and generic guidance that is applicable to all outcome types, disease areas, and populations is still needed for trial protocols and published reports.
This protocol outlines the development process for an internationally harmonized outcome reporting standard for clinical trial protocols and reports, called the Instrument for reporting Planned Endpoints in Clinical Trials (InsPECT). Through an evidence-based and consensus-based process, InsPECT will evaluate what constitutes complete reporting of trial outcomes, with respect to outcome selection, rationale, definition, measurement, outcome analysis and its presentation, interpretation, and transparent reporting of any outcome modifications between trial report and protocol. InsPECT will ultimately form two evidence-based reporting extensions, one specific to trial protocols (SPIRIT extension) and one specific to trial reports (CONSORT extension). The InsPECT extensions will be complementary to the work of the core outcome set developers and the Core Outcome Measures in Effectiveness Trials (COMET) Initiative, which provides information and guidance on which outcomes to measure and report for particular health conditions [
34].
Discussion
InsPECT will provide guidance on how to completely report any type of outcome in clinical trial reports and protocols. The development and implementation of the InsPECT extensions to CONSORT and SPIRIT have the potential to help harmonize and standardize outcome reporting across published trials reports and protocols, which will help facilitate trial reproducibility, transparency, and critical appraisal. While the emphasis of InsPECT is on clinical trials, the resultant guidance is expected to be generally applicable to all evaluative study designs, including observational studies and other study designs; future studies may develop specific InsPECT extensions for other study designs. The adoption and implementation of the InsPECT extensions will facilitate systematic reviews and meta-analyses by helping to standardize outcome reporting in the primary studies and promises to help reduce, or at least better identify, selective reporting between protocols and trial reports. The InsPECT extensions will also help increase the value of core outcome sets, which are increasingly being developed and implemented [65], by helping to enhance the clear and reproducible reporting of the core outcomes across trials and trial documents.
We expect that our implementation of a consensus-based and iKT approach will lead to improved clarity and acceptance and use of the InsPECT extensions among the broader research community. One potential challenge in the development of InsPECT is maintaining an optimal balance between usability and comprehensiveness. Different types of outcomes will require different types of information to be reported to enable reproducibility and transparency, and may also vary depending on the trial context including the intervention and population. For example, reporting on outcome assessor training may be quite relevant for some clinician-reported outcomes, but less so for biological markers such as cholesterol levels measured using standard laboratory processes. InsPECT will identify the minimal level of information required to be reported. Involving a large, diverse, and international group of stakeholders in the development of InsPECT may increase usability among the broader research community. As of September 2018, an initial list of InsPECT candidate items has been generated and the scoping review is ongoing. The Delphi study will be completed prior to the 2019 consensus meeting.
Trial status
Not applicable.