Clinical trials in settings with intermittent or non-existent internet and power connectivity, for example during humanitarian emergencies, present challenges in the synchronisation of data across different sites, in addition to accessing a centralised database in real-time. To overcome these, we designed a novel hybrid analogue/digital data management system which was deployed during the rapid implementation of a Phase III evaluation of a two-dose preventative vaccine for Ebola virus disease in Goma, Democratic Republic of the Congo, from 2019 to 2022. We provided study participants with an Enhanced Participant Record Card (EPRC) that served as eligibility for, and confirmation of, vaccination and was used in combination with Open Data Kit (ODK) electronic case report forms to create an off-grid study participant management system. To understand the utility of the EPRC, we analysed data from 15,327 study participants who received both vaccines and various types of prompts or reminders to return for dose 2, including home visits, telephone calls, or short messaging service (SMS). A total of 53% participants referred to the date on the EPRC as a prompt to return for dose 2 and 36.1% mentioned this as the only prompt. A multivariable generalised linear mixed-effects model showed that those who were not working, those aged 45–64 years or who had a chronic medical condition identified prior to receiving dose 2 were more likely to use the date on the EPRC as a prompt. Our findings demonstrate the utility of this system in the facilitation of decentralised data collection in off-grid locations that may be useful for future trials in complex humanitarian settings.

BACKRGOUND

The recombinant vesicular stomatitis virus-Zaire Ebola virus (rVSV-ZEBOV) vaccine is the only WHO prequalified vaccine recommended for use to respond to outbreaks of Ebola virus (species Zaire ebolavirus) by WHO's Strategic Advisory Group of Experts on Immunization. Despite the vaccine's widespread use during several outbreaks, no real-world effectiveness estimates are currently available in the literature.

METHODS

We conducted a retrospective test-negative analysis to estimate effectiveness of rVSV-ZEBOV vaccination against Ebola virus disease during the 2018-20 epidemic in the Democratic Republic of the Congo, using data on suspected Ebola virus disease cases collected from Ebola treatment centres. Those eligible for inclusion had an available Ebola virus RT-PCR result, available key data, were eligible for vaccination during the outbreak, and had symptom onset aligning with the period in which a ring-vaccination protocol was in use. After imputing missing data, each individual confirmed by RT-PCR to be Ebola virus disease-positive (defined as a case) was matched to one individual negative for Ebola virus disease (control) by sex, age, health zone, and month of symptom onset. Effectiveness was estimated from the odds ratio of being vaccinated (≥10 days before symptom onset) versus being unvaccinated among cases and controls, after adjusting for the matching factors. The imputation, matching and effectiveness estimation, was repeated 500 times.

FINDINGS

1273 (4·8%) of 26 438 eligible individuals were positive for Ebola virus disease (cases) and 25 165 (95·2%) were negative (controls). 40 (3·1%) cases and 1271 (5·1%) controls were reported as being vaccinated at least 10 days before symptom onset. After selecting individuals who reported exposure to an individual with Ebola virus disease within the 21 days before symptom onset and matching, the analysis datasets comprised a median of 309 cases and 309 controls. 10 days or more after vaccination, the effectiveness of rVSV-ZEBOV against Ebola virus disease was estimated to be 84% (95% credible interval 70-92).

INTERPRETATION

This analysis is the first to provide estimates of the real-world effectiveness of the rVSV-ZEBOV vaccine against Ebola virus disease, amid the widespread use of the vaccine during a large Ebola virus disease outbreak. Our findings confirm that rVSV-ZEBOV is highly protective against Ebola virus disease and support its use during outbreaks, even in challenging contexts such as in the eastern Democratic Republic of the Congo.

During the 2018–2020 Ebola virus disease outbreak in Democratic Republic of the Congo, a phase 3 trial of the Ad26.ZEBOV, MVA-BN-Filo Ebola vaccine (DRC-EB-001) commenced in Goma, with participants being offered the two-dose regimen given 56 days apart. Suspension of trial activities in 2020 due to the COVID-19 pandemic led to some participants receiving a late dose 2 outside the planned interval. Blood samples were collected from adults, adolescents, and children prior to their delayed dose 2 vaccination and 21 days after, and tested for IgG binding antibodies against Ebola virus glycoprotein using the Filovirus Animal Nonclinical Group (FANG) ELISA. Results from 133 participants showed a median two-dose interval of 9.3 months. The pre-dose 2 antibody geometric mean concentration (GMC) was 217 ELISA Units (EU)/mL (95% CI 157; 301) in adults, 378 EU/mL (281; 510) in adolescents, and 558 EU/mL (471; 661) in children. At 21 days post-dose 2, the GMC increased to 22,194 EU/mL (16,726; 29,449) in adults, 37,896 EU/mL (29,985; 47,893) in adolescents, and 34,652 EU/mL (27,906; 43,028) in children. Participants receiving a delayed dose 2 had a higher GMC at 21 days post-dose 2 than those who received a standard 56-day regimen in other African trials, but similar to those who received the regimen with an extended interval.

During the 2018–2020 Ebola virus disease (EVD) outbreak, residents in Goma, Democratic Republic of the Congo, were offered a two-dose prophylactic EVD vaccine. This was the first study to evaluate the safety of this vaccine in pregnant women. Adults, including pregnant women, and children aged ≥1 year old were offered the Ad26.ZEBOV (day 0; dose 1), MVA-BN-Filo (day 56; dose 2) EVD vaccine through an open-label clinical trial. In total, 20,408 participants, including 6635 (32.5%) children, received dose 1. Fewer than 1% of non-pregnant participants experienced a serious adverse event (SAE) following dose 1; one SAE was possibly related to the Ad26.ZEBOV vaccine. Of the 1221 pregnant women, 371 (30.4%) experienced an SAE, with caesarean section being the most common event. No SAEs in pregnant women were considered related to vaccination. Of 1169 pregnancies with a known outcome, 55 (4.7%) ended in a miscarriage, and 30 (2.6%) in a stillbirth. Eleven (1.0%) live births ended in early neonatal death, and five (0.4%) had a congenital abnormality. Overall, 188/891 (21.1%) were preterm births and 79/1032 (7.6%) had low birth weight. The uptake of the two-dose regimen was high: 15,328/20,408 (75.1%). The vaccine regimen was well-tolerated among the study participants, including pregnant women, although further data, ideally from controlled trials, are needed in this crucial group.