Table 1.

Suggested Elements for a Core Minimal Dataset of Observation-based Data for Designing Clinical Trials for High-priority Pathogens

Nature of Information	Value to the Conception and Design of Clinical Trials
Case counts for previous outbreaks	Serves as rudimentary estimate of the feasibility of sample-size requirements. Clinical trial groups should prioritize the most efficient trial designs when a low number of cases is expected.
Temporal and geographical profile of previous outbreaks	This is required for logistical planning, to ensure that local teams are sufficiently trained in research practices (such as good clinical practice) and trial-specific equipment is available.
An agreed-upon case definition	Clinical characteristics of the disease are used to define enrollment criteria.
Analysis of strength of evidence for factors associated with increased disease severity or fatality	Stratification (or other statistical adjustment) on the basis of severity is often required when interpreting the clinical trial outcome.
Best available descriptions of the type and rate of clinical outcomes	Clinical outcomes will function as a trial outcome measures. Understanding the natural course of illness will also help differentiate disease course from adverse events from treatment.
Assessment of confidence in estimates of clinical outcomes	Heterogeneity in patient outcomes between or within outbreaks creates uncertainty for power calculations and will affect selection of a statistical design for a trial. Spurious heterogeneity may occur due to random error in small cohorts, or represent ascertainment, lead-time, measurement, or follow-up bias. Real heterogeneity can occur due to improvements in care over an outbreak, pathogen evolution, or changes in host susceptibility and vulnerability but should be adjusted for.
Analysis of known or suspected covariates of outcome	Highlights possible confounders that will alter outcome independently of treatment and that will require adjustment if unequally distributed between treatment and control arms.
The mean time from onset of symptoms to outcome	Allows for an estimation of the feasibility and logistics of medical intervention.
Agreed-upon standards of care for patient treatment	Determines if there is standardized supportive therapy to be adopted in all arms of a trial. This is especially important for multicenter research
The performance characteristics of the favored diagnostic method	Determines whether a trial will be performed on an ITT basis or following laboratory confirmation.
Mean time for laboratory diagnosis	Determines whether a trial will be performed on an ITT basis or following laboratory confirmation.
Community priorities and expectations for trials	Determines the priorities of affected communities in terms of access to trials, acceptable methodology, and acceptability of treatments or vaccines.

Abbreviation: ITT, intention-to-treat.