COVID-19: Nothing is Normal in this Pandemic

Area of Science:

• Epidemiology
• Mathematical Biology
• Statistical Modeling

Background:

• The COVID-19 pandemic highlighted widespread inaccuracies in epidemiological concepts.
• Misapplication of terms like "normal epidemic curve" and "log-normal distribution" has been prevalent in scientific literature and social media.
• Established textbooks and academic courses have perpetuated these misleading concepts for years.

Purpose of the Study:

• To address and correct inaccurate epidemiological concepts disseminated during the COVID-19 pandemic.
• To highlight the misuse of statistical distributions, such as Gaussian and log-normal, when describing epidemic curves.
• To advocate for updated epidemiological education and modeling approaches.

Main Methods:

• Analysis of common representations of epidemic curves in scientific literature and social media.
• Critique of the application of statistical distributions (e.g., Gaussian, log-normal) to epidemic data.
• Review of existing epidemiological textbooks and educational materials.

Main Results:

• Epidemic curves are frequently misrepresented as normal or log-normal distributions.
• The temporal nature of epidemic data is often ignored when fitting to distributions like Gaussian.
• Current epidemiological teaching and literature contain significant conceptual errors regarding epidemic curve characteristics.

Conclusions:

• Epidemic curves, while sometimes resembling Gaussian functions, are not true normal or log-normal distributions.
• Pandemic data is complex and requires sophisticated statistical and mathematical modeling beyond "one-size-fits-all" solutions.
• Updating epidemiological textbooks and practices is crucial for providing reliable information for public health policy.