Table 2 Considerations for design of chemoprophylaxis trials against emerging epidemic disease

Study power

Trials should be powered based on adequate scepticism of expected effect size (recognising that modest effect size can have non-trivial utility in prophylaxis at scale against emerging infection) and attack rates (using early epidemic data and/or similar diseases/circumstances).

For large studies premised on modest effect sizes, inclusion of interim analysis with suitable stopping rules can enable early termination if a prophylactic agent has a true high efficacy.

Population studied

Study population choices may be driven by the primary form of chemoprophylaxis efficacy under evaluation (see “Objective” below) and trial efficiency grounds. Demonstrating efficacy in a clinically relevant population should also drive study population e.g. selecting those at highest risk of severe outcomes if this is the primary study objective.

Objective: prevention of (symptomatic/asymptomatic) infection, symptomatic disease or severe disease and death

Infection is a prerequisite for disease and for onward transmission. If symptomatic disease is not substantially different from asymptomatic infection in health utility or transmission, then the higher frequency of any-infection over symptomatic infection may warrant consideration, including role in transmission interruption — a potential public health goal — and endpoint accrual rate.

Conversely, severe outcomes such as hospitalisation, death and long-term sequelae of infection, while rare events represent clinically important outcomes.

Covering different severities (infection, disease and severe disease between primary and secondary objectives of the study is likely to be necessary).

Endpoints: ever infected — serological markers of infection

Regular repeat or end of study serology may be an endpoint.

Baseline serology may also be valuable in PEP and PREP studies: infection status at baseline may be used as e.g. a stratifying factor for pre-specified analysis.

Serological sampling and frozen storage should be considered even if relevant serological assays are not available for a newly emerging disease epidemic at the start of the trial. Subsequent development or expansion of access could enable future analysis of stored samples.

Endpoints: current infection — samples for pathogen detection by (RT-)PCR or other methods

Testing for infection during a symptomatic episode is required if disease or severe disease is an endpoint.

Frequent testing enables detection of asymptomatic or pre-symptomatic infection and conversion to disease.

Sampling and frozen storage should be considered even if relevant pathogen detection assays are not available for a newly emerging disease epidemic at the start of the trial. Subsequent development or expansion of access could enable future analysis of stored samples [36].

Platform (adaptive) designs

Platform trial designs enable managed entry and exit of candidate chemoprophylactic agents as information accrues.

Relatedly, randomisation need not be simple 1:1 against placebo but can be multi-arm and factorial.

Multi-centre/multi-site studies.

Geographical hotspots of transmission may move. Multi-centre/multi-site studies, including international collaborations, may facilitate endpoint accrual within a study. Alternatively, international collaborations may enable the use of equivalently designed chemoprophylaxis trials that support integrated analysis.

For chemoprophylaxis trials, “multi-centre/multi-site” does not specifically refer to hospitals as the only setting for recruiting to studies.

Inter-epidemic research infrastructure preparedness: trials available early in epidemics

Research infrastructure for epidemic disease chemoprophylaxis enables rapid deployment when transmission is identified, supporting endpoint collation.

For newly emergent diseases, catch-all registrations and approvals are unlikely to be feasible. Chemoprophylaxis research for outbreaks of known epidemic disease could be facilitated by specific prior trial registrations additional to the presence of general research infrastructure.

Pre- vs post-exposure chemoprophylaxis designs

Post-exposure trials would be expected to have higher attack rates in participants than pre-exposure trials, supporting trial efficiency, particularly in the absence of an uncontrolled generalised epidemic. An important trade-off in trial design is that time from disease (more specifically infectiousness) onset and diagnosis in the index case to initiation of PEP candidates in enrolled participants may enable infection to arise where the same agent may demonstrate efficacy if used pre-exposure. This biases towards the null. The same effect arises if index case and household contact have a common source of infection, or the “contact” is the index case infection source.

PEP studies that have severe disease/death as an endpoint may also be testing early treatment effects of candidate drugs.

Therapeutics and anti-infective chemoprophylactics

Drugs for chemoprophylaxis do not need to be the same agents used therapeutically against severe endpoints in patients with the disease of interest; evidence from one scenario does not necessarily inform the other. For example, a drug with effects against viral replication and infection may have limited scope against inflammatory or autoimmune disorders in managing disease where an immunomodulatory drug may be most appropriate and vice-versa.