Confidence limits based on the first occurrence of an event

Area of Science:

• Biostatistics
• Medical Informatics
• Epidemiology

Background:

• Estimating population prevalence and incidence from medical databases is crucial for public health.
• Traditional methods often rely on binomial and Poisson distributions, which may yield wider confidence intervals.
• Accurate confidence limits are essential for reliable interpretation of epidemiological data.

Purpose of the Study:

• To introduce and evaluate confidence intervals based on geometric and exponential distributions for prevalence and incidence data, respectively.
• To compare the precision of these new intervals with those derived from binomial and Poisson distributions.
• To demonstrate the practical applications of these improved statistical methods.

Main Methods:

• Derivation of confidence limits using geometric distribution for prevalence (first occurrence).
• Derivation of confidence limits using exponential distribution for incidence (time to first occurrence).
• Comparative analysis of interval widths and coverage probabilities against binomial and Poisson methods.

Main Results:

• Lower confidence limits derived from geometric/exponential distributions are identical to those from binomial/Poisson distributions.
• Upper confidence limits are consistently smaller when using geometric/exponential distributions.
• The resulting confidence intervals are shorter, indicating increased precision.

Conclusions:

• Geometric and exponential distributions provide more accurate and precise confidence intervals for prevalence and incidence in medical databases.
• These refined methods enhance the reliability of statistical inferences in epidemiological studies.
• The application of these intervals can lead to better-informed public health decisions.