A note on parameter estimation for variant Creutzfeldt-Jakob disease epidemic models

Area of Science:

• Epidemiology
• Biostatistics
• Neurodegenerative Diseases

Background:

• Recent studies questioned the estimation of key epidemiological parameters for variant Creutzfeldt-Jakob disease (vCJD).
• Specifically, it was argued that the epidemic scale for vCJD cannot be estimated and requires pre-fixing in analyses.
• Concerns were also raised regarding the selection of incubation period distributions in vCJD modeling.

Purpose of the Study:

• To address and refute issues concerning the estimation of variant Creutzfeldt-Jakob disease (vCJD) epidemiological parameters.
• To demonstrate that the scale of the vCJD epidemic is estimable.
• To clarify the appropriate use of likelihood methods and incubation period distributions in vCJD research.

Main Methods:

• Survival model fitting to case data.
• Utilizing the full likelihood function for parameter estimation.
• Critically evaluating the approximate likelihood methods previously employed.
• Analyzing the impact of incubation period distribution choices.

Main Results:

• The inability to estimate the vCJD epidemic scale stems from the use of approximate likelihood functions, not an inherent limitation.
• Estimation using the full likelihood resolves the issue, allowing for accurate assessment of the epidemic's scale.
• The choice of incubation period distribution is discussed in light of the corrected estimation methods.

Conclusions:

• The scale of the variant Creutzfeldt-Jakob disease (vCJD) epidemic can be reliably estimated.
• Previous concerns were an artifact of approximate statistical methods, not fundamental epidemiological challenges.
• This work validates the use of full likelihood in vCJD epidemiological modeling.