responders reporting from the field. MSF, interpreting data based on the geographic dispersion of cases confirmed through methods other than laboratory confirmation and identification of family networks crossing Guinea, Liberia, and Sierra Leone, determined that cases were spreading in the latter well before May 26, when the first case was officially confirmed. 7 The result was data that justified both action and inaction by relevant stakeholders, with other political and economic pressures favoring the latter from March until July 2014. In Sierra Leone, the Ministry of Health and Sanitation shared data infrequently and sometimes not at all with its own National Ebola Response Centre (NERC) (which integrated UK DFID, UN, and other international stakeholders), but they would share it with WHO. NERC received summary data, but not detailed data relevant to its activities. WHO would publish its data according to its own criteria which affected the credibility of data issued by the NERC, which in turn had to request data from UK DFID and other aid or public health agencies. The delay had material, significant effect. According to one study, if resources committed in September and delivered in October had done so one month earlier, 12,500 cases could have been prevented. 62 8 Sharing of epidemiological and contact list data in Sierra Leone is contrasted with its sharing in Nigeria. There, Nigeria’s Port Health Services obtained records of an Ebola-infected patient’s travel, contact lists were compiled by public authorities, and 18,000 visits were made by local health workers to those contacts to ensure that the infection wasn’t spreading and that those who needed treatment received it. All data was collected at the direction of, and processed through, an emergency Ebola centre. ii. Operational Data Operational research requires the collection of accurate, harmonized, and routine data. Across Ebola treatment centres, even just within those run by a single organization like MSF, information was not collected, recorded, and shared through standardized methods. This discrepancy led to difficulties when trying to amalgamate and analyse patient data. Additionally, clinical interventions, such as the use of intravenous fluids, were not recorded systematically across Ebola treatment centres. This lack of records proved to be a lost opportunity because retrospectively assessing what effect interventions and others had on patient outcomes was not possible.76 iii. Clinical and Genetic Data The fractured state of health services provision in Guinea, Liberia, and Sierra Leone resulted in makeshift facilities those seeking treatment visited, limits on the control and use of diagnostics and blood sampling in those facilities, and the transfer of “thousands” of samples out of the countries (evidence suggests less control and therefore more transfer from Guinea and Sierra Leone), research upon which has gone largely unshared.24 Before more rapid Ebola diagnostics were developed, several methods for detecting infection and/or disease with Ebola virus had been developed that were amenable for use in clinical laboratory settings. 22 Those fell into three basic categories: (i) serologic tests that detect host antibodies generated against the virus, (ii) antigen tests that detect viral proteins, and (iii) molecular tests that detect viral RNA sequences.23 The latter method also allowed genetic data to be used to trace genetic mutations so as to track the virus’s spread, as well as to determine whether it was sustained by human-to-human transmission or by contact with bats or some other carrier.24 Genetic data also suggested new probable routes of infection and, importantly, revealed where and how fast mutations were occurring. In addition to identifying the source and spread of Ebola, this information is crucial to designing effective diagnostics, vaccines and antibody-based therapies. 25 Genome data sets are large and complex, and the best way to understand these complex sequences is to share data ”as widely and as quickly as possible.”26 Although researchers in some cases quickly uploaded genetic sequence analyses to platforms like GenBank, there was no standardized way to share or disseminate the data.25 Just as importantly, of the thousands of samples transferred for purposes of genetic sequencing, few and for many months zero, Ebola virus 9 sequences were made publicly available. 25 The result of uncontrolled collection and transfer of human and non-human biological samples has been the corresponding distrust fomented by the governments of Guinea and Sierra Leone especially. Access to biological samples is now significantly more regulated by those governments, which have imposed stronger limitations on access and sharing, and more elaborate notifications as to the subject and use of samples studied on or from Guinean or Sierra Leonean territory. Stakeholders from international charities, WHO, aid agencies, and biomedical researchers all confirmed that legal acquisition of samples for research through normal regulatory channels is now a major impediment to necessary research. b. Community-Level Barriers to Data Sharing The collection of epidemiological and surveillance data described above informed the emergency response which aimed in significant measure at medical outreach and behavior changes