Simultaneous Confidence Band Approach for Comparison of COVID-19 Case Counts

Area of Science:

• Epidemiology
• Biostatistics
• Time Series Analysis

Background:

• The COVID-19 pandemic caused global disruption, raising questions about demographic differences in case counts.
• Analyzing daily COVID-19 cases requires methods that handle correlated data and non-linear trends due to events like variants and vaccinations.

Purpose of the Study:

• To determine if there is a significant difference in daily COVID-19 case counts between males and females in seniors.
• To address the limitations of traditional statistical tests for analyzing time-series epidemiological data.

Main Methods:

• A simultaneous confidence band approach was employed for time-series analysis.
• B-spline estimation was used to construct confidence bands for the trend of an autoregressive moving-average model.
• The method was applied to daily COVID-19 case data for Ohio seniors (≥60 years) from April 2020 to March 2022.

Main Results:

• A statistically significant difference was found in daily COVID-19 case counts between male and female seniors.
• The difference was significant at the 95% confidence level, after adjusting for population size.

Conclusions:

• The study demonstrates a significant gender disparity in COVID-19 case counts among older adults in Ohio.
• The applied statistical methodology effectively analyzes complex epidemiological time-series data with changing trends.