Monday, 11 May 2015

The validity of Ebola vaccine trials in Sierra Leone

Associate Professor Manoj Gambhir has co-authored an article with numerous researchers from the USA and Canada, discussing the validity and feasibility of a randomised control trial and a stepped-wedge cluster trial design in vaccine development to determine vaccine efficacy in within Sierra Leone.

An Ebola Virus Disease outbreak in West Africa began in 2014 and is ongoing. Statistics from April 2015 indicate that the outbreak has caused 10, 400 deaths, with over 25, 100 reported cases (Chowell & Viboud, 2015). Several reports have discussed the extremely harsh symptoms and highlighted the lack of a vaccine to protect at-risk communities.

Measures adopted so far have been supportive, with a focus to keep patients comfortable and hydrated. However, this epidemic has spurred numerous pharmaceutical companies and public health organisations to evaluate the feasibility of vaccine options.

Whilst the incidence of Ebola has seen a recent decline, vaccine trials are crucial to help with future outbreaks. A randomised control trial began in Liberia in February 2015 and, following this, a ring vaccination trial commenced in Guinea in March. A stepped-wedge cluster trial was being considered in Sierra Leone; however, this has been modified to a phased-rollout randomised control trial which is expected to begin in the near future.

A stepped-wedge cluster trial involves all participants being vaccinated, potentially in a random order of different geographical clusters. A randomised control trial involves participants being allocated to either the control or the vaccine group randomly. Dr Gambhir and the authors in this study evaluate the benefits and possible ramifications of these study designs within Sierra Leone.

The researchers utilised Ebola virus case data from Sierra Leone to project the incidence for a six month period. With these data, both the randomised control trial and the stepped-wedge cluster trial were simulated in synthetic (i.e. representative computerised) populations, separated into different geographical risk groups. Logistical barriers, individual and cluster level variation risk were all considered. Using statistical modelling, the stepped-wedge cluster trial incidence trends generated more biases in the calculation of Vaccine Efficacy (VE). 

Randomised control trial analysis sustained statistical validity in all tested models. The authors estimate that the power to determine a 90% VE was between 49% and 89% through a randomised control trial. Using this same estimation, the power was between 6% and 26% for a stepped-wedge cluster trial. This however, depends on the incidence of Ebola within the trial community. They also note that a one month delay in the beginning of the trial would decrease the power of the randomised control trial by 20% and 49% for the stepped-wedge cluster trial.

The authors find that the spatiotemporal variation in the risk of Ebola does affect the statistical power of the stepped-wedge cluster trial. This also impacts the anticipated ethical benefits of the trial that the randomised control trial does not possess. This is due to a randomised control trial’s ability to prioritise the vaccine delivery to increased risk groups. They conclude that whilst a randomised control trial has the ability to determine efficacy of the vaccine, these trials must commence in the near future to account for the significant decline in the power of the trial due to the steady reduction of the outbreak.

You can read the paper in full here.

Chowell, G., & Viboud, C. (2015). Ebola vaccine trials: a race against the clock. Lancet Infectious Diseases, 1-2.

No comments:

Post a Comment


Related Posts Plugin for WordPress, Blogger...