Estimation of the generation interval using pairwise relative transmission probabilities

Area of Science:

• Epidemiology
• Infectious Disease Dynamics
• Mathematical Modeling

Background:

• Understanding infectious disease spread relies on transmission intervals, like the generation interval and serial interval.
• Estimating these intervals is challenging due to difficulties in identifying transmission chains.
• Existing methods often require extensive contact tracing or genomic data, which are not always available.

Purpose of the Study:

• To develop a novel method for estimating transmission intervals using readily available surveillance or outbreak data.
• To overcome limitations of existing methods that require contact tracing or pathogen whole genome sequence data.
• To accurately estimate the generation interval distribution for various infectious diseases.

Main Methods:

• Developed an expectation maximization algorithm incorporating relative transmission probabilities and noise reduction.
• Utilized simulations to validate the method's accuracy across different disease parameters (reproductive numbers, mutation rates).
• Applied the method to routinely collected tuberculosis surveillance data from Massachusetts (2010-2016).

Main Results:

• Simulations demonstrated the method's ability to accurately estimate generation interval distributions.
• The method successfully estimated the serial interval for tuberculosis in Massachusetts using surveillance data.
• The novel approach provides a viable alternative for transmission interval estimation when detailed data is scarce.

Conclusions:

• The developed method offers a robust and accessible approach for estimating transmission intervals.
• This technique enhances understanding of infectious disease dynamics, particularly for diseases with limited data.
• Accurate serial interval estimation aids in public health interventions and disease control strategies.