“In the context of public health emergencies... study teams should be encouraged to collaborate on existing, ongoing protocols rather than starting new, independent trials.”
In the U.S. only a few COVID-19 randomized clinical trials (RCTs) have been centrally organized, e.g. by NIAID, PCORI and individual PIs. Over 400 such trials have been registered on clinicaltrials.gov with dozens being added each day. Many of them are designed to answer similar questions and combining data or aggregating evidence could dramatically increase their efficiency and precision, getting answers to doctors faster and more reliably. Many of the studies getting off the ground right now are being run outside of the research environment by hospitals in need of decision-making information for their doctors. Like these studies, our primary goal is to get high-quality evidence to doctors quickly.
Furthermore, local outbreaks may taper off before institutions are able to enroll their target sample size. Notably, this has happened in China, where many trials are effectively suspended with incomplete enrollment and inconclusive results. If the US response to COVID-19 proves successful, hospitals will be lucky to find themselves in a similar position with a sudden drop off in COVID cases that could affect enrollment. If protocols are public and open for collaboration, an RCT can be picked up in different regions as the outbreak moves across the country. Unfortunately, no platform currently exists for such collaboration on RCTs.
CovidCP fills this gap by publicizing protocols whose PIs are open to various levels of collaboration: joining forces with other research teams to create a core protocol; admitting new sites under the existing PI and IRB; sharing anonymized interim and/or final data through our partner Vivli with other sites that choose to independently operate a trial under a similar protocol.
Protocols can be in draft form, enrolling, or completed. To ensure that the authors receive credit, anyone using a protocol will be required to include a citation in any resulting research products. Submissions will be vetted for legitimate scientific value and well-defined interventions and outcomes. Submit a protocol here.
To request to collaborate on a protocol, click here. In doing so you agree to appropriately cite the authors of the original protocol in any resulting research products. Requests will be vetted for legitimate scientific value.
We strongly encourage collaboration. If a protocol is not yet enrolling subjects at the time of submission, we encourage the submitting study team to accept input from other interested research groups in order to create a collaborative multi-site protocol. The Trial Innovation Network and SMART IRB will prioritize and expedite requests to initiate multi-site studies that come to them through this platform. If a protocol is currently enrolling, it may be possible for interested research groups to join the existing study as a second or additional site under the submitting PI and IRB of record. If that is not possible we encourage interested research groups to match eligibility criteria, intervention definitions, and a subset of the six core outcomes (presumed COVID-19 diagnosis, hospitalization, ICU admission, supplemental oxygen use, mechanical ventilation, death) as closely as possible. In all cases, we encourage all researchers to make their data available for aggregated analyses and to make every attempt to include or reference an analysis that combines evidence from all collaborating studies in any resulting reports.
Cite protocols as [Authors] (2020). Protocol for [Study Name]. Retrieved from covidcp.org.
Statisticians from top (bio)statistics departments will volunteer time to run or advise on aggregated analyses, including working with study-specific Data Monitoring Committees to ensure that any interim analyses are grounded in best statistical practice, ethical, and maximally informative. Resources to anonymize and house interim and final data provided free of charge by Vivli.
For help, email us.
None of the individuals or organizations affiliated with or contributing material to this Site, nor any of their units, programs, employees, agents, or individual trustees, shall be liable for any improper or incorrect use of information obtained through the use of this Site.