Harmonisation of research outcomes for meaningful translation to practice: The role of Core Outcome Sets and the CROWN Initiative.

Abstract

In 2014, the CoRe Outcomes in Women's and Newborn health (CROWN) initiative was launched to support the development and use of research outcomes that are meaningful for researchers, clinicians and health consumers. For most areas of clinical research, among the myriad of outcomes, there has been little guidance for selecting those that are most important and feasible.1 Consequently, there is significant heterogeneity in the outcomes used and reported, as well as how they are defined and measured. This heterogeneity results in significant wastage of research resources with lost time, opportunity, study participants, and money, consuming up to 85% of the global medical research budget.2 Such waste delays advances in identifying and implementing effective interventions for optimising women's health outcomes. This wastage is clearly evident in the women's health literature. A recent systematic review of outcomes used in pre-eclampsia intervention trials identified 79 randomised controlled trials (RCTs) that reported 119 different maternal and newborn outcomes with related mortality only reported in 14% and 20% of trials, respectively, concluding that there is no consensus as to which outcomes are most important to measure, which definition or instrument should be used, and whether outcome measures remain valid, regardless of the time point of measurement.3 Similarly, a systematic review of 66 RCTs investigating diet and activity-related interventions in pregnancy found 142 outcomes were used with only 60% reporting ‘critically important’ outcomes of gestational diabetes mellitus (GDM), preterm birth and caesarean section, with half of outcomes reported in only a single trial. The reporting quality was also limited with only half clearly stating a primary outcome in a reproducible way.4 Given the large number of trials studying similar cohorts, significant progress could be made by combining studies to reach sample sizes that are sufficiently powered to answer clinical questions. However, the failure to include important outcomes, and differences in defining and measuring outcomes, mean that there are insufficient comparable outcomes to enable meta-analyses to answer these questions, with the evidence base muddied even further by the selective reporting of outcomes. To address this problem the Core Outcomes Measures in Effectiveness Trials (COMET) initiative was launched in 2010 following a meeting of 110 people representing health consumers, trialists, systematic reviewers, journal editors, trial funders, policy makers, trial registries and regulators. Being no less an issue in women's and newborn health research, journal editors joined to form the CROWN initiative. The role of both initiatives is to support the systematic development, dissemination and implementation of Core Outcome Sets (COS). A COS is a minimum data set of important outcomes with standardised definitions and reporting prioritised by stakeholders, including healthcare professionals, researchers and patients. The COS harmonisation is a result of: (i) systematically reviewing the literature to produce a potential outcomes inventory; (ii) identifying representative stakeholder groups, especially patient participants,5 and (iii) subjecting the inventory to a consensus-building process, for example a modified delphi process, to agree on the minimum core outcomes, how they are defined and measured. A similar process is used to achieve a consensus on instruments that are truthful, discriminatory and feasible,6 for measuring the outcomes, or to determine that a new instrument needs to be developed. Once developed, trialists are encouraged to include these outcomes; however, they need not be the only outcomes measured nor does a core outcome have to be the primary outcome. Since its launch, 39 COS protocols have been registered with CROWN, with six published defining Core Outcomes in:7 While women's health COS are at an early stage of development, they are now being incorporated into clinical trials2 for example prevention of preterm birth,8 in the same way other disciplines such as rheumatology, have done using the Outcome Measures in Rheumatology (OMERACT) COS. The OMERACT initiative has resulted in improvements in RCT conduct, with 70% of trials reporting all COS measures, importantly including patient-reported outcomes.3 To achieve a similar level of success in women's health we need to support COS development and implementation, for which there are a number of resources including the COMET and CROWN websites as well as a detailed handbook produced by COMET.9 To support Australasian researchers, an Australasian CROWN centre will be launched in March 2018 as a collaboration between RCOG, RANZCOG, and Monash University. ANZJOG also invites submissions for COS protocols and completed sets, and can provide guidance on their use by submitting authors. ANZJOG is a proud supporter of the CROWN initiative and looks forward to the benefits COS will offer researchers with trial design and translation of research findings to improve outcomes for women's and newborn health.