<scp>CROWN</scp> initiative and preterm birth prevention: researchers and editors commit to implement core outcome sets

Abstract

The emphasis on research quality has taken great strides in recent years. Publication of clinical trial protocols (i.e. SPIRIT guideline), prospective registration (e.g. ClinicalTrials.gov, WHO registry), high standards in research conduct (i.e. ICH-Good Clinical Practice guidelines) and transparent reporting of the findings (i.e. CONSORT guideline) are all becoming features of clinical effectiveness research. However, many more gains are possible to further improve the efficiency of clinical research and therefore reduce waste in health research, a problem that is estimated to consume 85% of the global medical research budget each year.1 When addressing the issue of research waste, an area of concern is the use of different (incomparable) outcomes.1 In preterm birth research for example, 103 trials reported 72 different primary outcomes, while 33 Cochrane reviews reported 29 different outcomes.2 This lack of commonality in the outcomes has hampered meaningful data synthesis. If datasets cannot be combined, it is difficult to reach proper statistical power, subsequently limiting the impact research can have on clinical practice. An additional concern is that outcomes reported have not always been those that patients regard as relevant.3 Recognising that disease-specific international consensus on outcome measures is an important step towards addressing the above concerns, the CROWN (Core Outcomes in Women's and Newborn Health) initiative was launched in 2014.4 A unique, sizable consortium of 82 women's health journals in multiple languages, it promised to encourage researchers to develop disease-specific core outcome sets (i.e. the minimal sets of outcomes that relevant stakeholders consider critical for evaluation of interventions in specific health conditions and in different resource settings) using robust consensus methodology.4 Functioning as a stakeholder group, it contributed to and endorsed the first core outcome set for preterm birth studies published following peer-review in 2016, together with summary publications in several journals and languages.5 With this commentary, we wish to give an update on its dissemination. As a background, we can show (Figure 1) that the 13 outcomes incorporated in the core outcome set for preterm birth have never all been reported in any of the previous 103 randomised trials on this topic. Only four outcomes (maternal infection, harm to mother, gestational age at birth, and birth weight) have been frequently reported; the other 9 outcomes are reported in under one-third of the trials. Successful implementation of this core outcome set will result in a significant change in the frequency with which these outcomes are reported in preterm birth trials and therefore improve the quality and relevance of studies in this field of research. We do acknowledge that some trials might not be able to collect all items of the core outcome set. However, this core outcomes set should provide a strong incentive to researchers to argue for adequate funding to collect all outcomes. We also recommend that other outcomes outside the core set should be further explored. Furthermore, we do not insist that the sample size (power) calculations should be based on a core outcome. Finally, we accept this core outcome set guides what trials should be measuring but does not say how those measurements should be made. Some outcomes, as for example respiratory morbidity, require several measurements to capture respiratory distress syndrome and bronchopulmonary dysplasia. All these issues would need to be refined moving forward. What we do want to show is our commitment to addressing the problem of lack of harmony in outcome reporting in preterm birth research by adopting the core outcome set in our work. Table 1 provides a list of trials that have adopted or promised to adopt the core outcome set. Those running these trials have also reached consensus as to the definitions and measurement instruments for each of the outcomes in the core outcome set (Table 2). We offer our consensus as an exemplar of how proper coordination and information sharing between researchers developing core outcome sets can address methodological challenges and lead to a real prospect of progress. The need for progress in the development and implementation of core outcome sets is well recognised and has resulted in a significant change in the quality and relevance of research in, for example, rheumatoid arthritis trials in the nineties. Successful implementation in this field has enriched clinical practice by identifying new outcomes that are now routinely monitored by healthcare professionals around the world.5 The importance of patient involvement as an essential rather than optional feature in the process of design, development, and dissemination of a core outcome set goes without saying. This can be achieved by, for example, involving patient organizations/representatives and organizing patient focus groups. Patient involvement has led to improved quality of targeted research, wider dissemination of its findings, and better integration into healthcare policy and practice.7 Furthermore, we think that representation of different resource settings (high, middle, and low income countries) is crucial in achieving consensus for important conditions related to maternal and neonatal health. In summary, improving awareness and dissemination of the preterm birth core outcome set will increase the availability of comparable research data that are important to patients and health professionals that face the burden of preterm birth. This will reduce research waste, facilitate research synthesis, and promote evidence-based care, with the ultimate goal of improving the outcomes for these patients. We call upon all researchers, funders, ethics committees, and institutional review boards to facilitate the infrastructures needed to collect the preterm birth core outcomes and, in time, core outcomes for other subjects in our specialty. None declared. Completed disclosure of interests form available to view online as supporting information. All authors contributed sufficiently to be named as an author. Not applicable. No funding source. None. Please note: The publisher 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.