from an open collaboration between investigators and institutions in Africa, Europe, and North America. These teams, coordinated by the WHO, were able to generate and exchange critical data for the development of urgently needed novel vaccines along faster timelines than have ever before been achieved.” But a story published by ScienceInsider, a news website of the journal Science, argued that the theory that NewLinks, the licensee of the most promising vaccine candidate, “dragg[ed] its feet because it was "worried about losing control over the development of the vaccine."47 NewLink received at least $50 million and future post-licensing royalties to collaborate with Merck to further develop and commercialize their vaccine.47 Protection of data and trade secrets is necessary for intellectual property and related commercial claims that might be later asserted. 47 Researchers on data sharing between firms especially noted that there were insufficient incentives and infrastructure to do so.48 13 This problem was particularly applicable for “negative” and inconclusive results. For example, Tekmira Pharmaceuticals developed a promising potential treatment for Ebola, TKM-EbolaGuinea. 50 When the treatment did not show efficacy in Sierra Leone, Tekmira halted enrollment in clinical trials. The company stated that it was “not likely to demonstrate an overall therapeutic benefit.”50 Tekmira stated that “the data were being analysed and would be made available later.”50 Similarly, Chimerix of Durham, North Carolina, would not publicly reveal why it withdrew support for a trial of brincidofovir in late January after four patients had been treated. 18 The experimental drug ZMapp was given to a handful of patients before supplies ran out in August 2014. Detailed information on the patients’ reactions to the drug was not released, owing to fears that this would prevent researchers from publishing on the cases.18 At least two interviewees attributed the failure to share negative or inconclusive data about therapeutics to the potential market for those drugs – because they were developed with several disease targets in mind, not even primarily Ebola – negative results about Ebola trials could adversely affect investor interest or regulatory approvals for other diseases for which they might be effective. Although there have been ethnographic criticisms of data gathered through mobile phone applications, it is largely understood to represent a promising way to communication information, contact trace, and coordinate response. Cell phone data including GPS locations are not part of an official record, but could communicate routes, patterns, and other relevant data. The ownership of cell phone data was ambiguous and a subject of significant dispute: was that data owned by the government, by the cell phone companies, or users; how do we share data that is not specifically obtained in response to a form generated as solicited messages to users? West African telecommunications companies—Orange, Safaricom, Digicel, and others—had in the past shared call record data for public health programs, but telecommunications firms, the United Nations and the Sierra Leonean government could not agree on terms by which that data would be shared.68 j. The publishing imperative One of the fundamental dilemmas for researchers with primarily academic affiliations is in the incentive structure for tenure, promotion, pay, and status. Release of preliminary data may not only subject the researcher to later criticism if the data is erroneous or flawed, but also may jeopardize his or her opportunity to publish the data in peer review journals that satisfy tenure and promotion criteria. 21 Delays and barriers to data sharing of relevant Ebola information included the time taken by authors to prepare, write and submit their papers; desire to first submit results to high profile journals; time taken by journals to review and make decisions about publication; and time taken to complete the publication process. These delays applied to all data related to the research response including epidemiological, surveillance, emergency response, health facility data, pathogen genome data, research data including surveys, observational studies, clinical trials of diagnostics, therapeutics and preventives, quality controlled interim results, final research results, and inconclusive results. Measurement of this delay has been far more robust for clinical trial data. As of November 14 14, 2017, ClinicalTrials.gov had listed 35 completed Ebola trials. None had posted a summary of its results onto the database, even though 30 trials had passed the one-year disclosure deadline set by the WHO. For clinical trials related to Ebola, the median publication lag-time (from the end of the study) was 338 days (range 157–621), the median submission lag-time (from study end to submission to the journal where it was eventually published) was 297 days (116–450), and the median review lag-time (from submission to publication) was 178 days (137–193).64 Currently, there are few mainstream efforts to address these barriers during public health emergencies. Publishing research takes time, and the peer review process is often not expedited during public health emergencies. The earliest published genomic analyses of the Ebola outbreak, crucial for determining where it originated and how it