Uncertainty and Sensitivity Analysis of the Basic Reproductive Rate: Tuberculosis as an Example

• Published in: American journal of epidemiology Vol. 145; no. 12; pp. 1127 - 1137
• Main Authors: Sanchez, Melissa A., Blower, Sally M.
• Format: Journal Article
• Language: English
• Published: Cary, NC Oxford University Press 15-06-1997
• Subjects: Adult; Aged; basic reproductive rate; Biological and medical sciences; Disease Outbreaks - statistics & numerical data; Disease Transmission, Infectious; epidemic; Epidemiology; General aspects; Humans; Incidence; Life Expectancy; Medical sciences; Methodology; Middle Aged; Models, Statistical; Mycobacterium tuberculosis; Population Dynamics; Public health. Hygiene; Public health. Hygiene-occupational medicine; Reproduction; Sample Size; sensitivity analysis; Survival Rate; tuberculosis; Tuberculosis - epidemiology; Tuberculosis - transmission; uncertainty analysis; Infection; Epidemic; Human; Uncertainty; Sensitivity analysis; Tuberculosis; Methodology; Bacteriosis; Mycobacterial infection; Transmission from man to man; Epidemiology; Public health
• ISSN: 0002-9262; 1476-6256
• DOI: 10.1093/oxfordjournals.aje.a009076

The basic reproductive rate (Ro) is a measure of the severity of an epidemic. On the basis of replicated Latin hypercube sampling, the authors performed an uncertainty and sensitivity analysis of the basic reproductive rate of tuberculosis (TB). The uncertainty analysis allowed for the derivation of a frequency distribution for Ro and the assessment of the relative contribution each of the three components of Ro made when TB epidemics first arose centuries ago. (The three components of Ro are associated with fast, slow, and relapse TB.) Ro estimates indicated the existence of fairly severe epidemics when TB epidemics first arose. The Ro for the susceptible persons who developed TB slowly (RoSlow) contributed the most to the Ro estimates; however, the relative RoSlow contribution decreased as the severity of TB epidemics increased. The sensitivity of the magnitude of Ro to the uncertainty in estimating values of each of the input parameters was assessed. These results indicated that five of the nine input parameters, because of their estimation uncertainty, were influential in determining the magnitude of Ro. This uncertainty and sensitivity methodology provides results that can aid investigators in understanding the historical epidemiology of TB by quantifying the effect of the transmission processes involved. Additionally, this method can be applied to the Ro of any other infectious disease to estimate the probability of an epidemic outbreak. Am J Epidemiol 1997; 145: 1127–37.