Simulating and Forecasting the Cumulative Confirmed Cases of SARS-CoV-2 in China by Boltzmann Function-based Regression Analyses

Xinmiao Fu, Qi Ying, Tieyong Zeng, Tao Long, Yan WangmedRxiv 2020 · DOI

Abstract

An ongoing outbreak of atypical pneumonia caused by the 2019 novel coronavirus (SARS-CoV-2) is hitting Wuhan City and has spread to other provinces/cities of China and overseas. It is very urgent to forecast the future course of the outbreak. Here, we provide an estimate of the potential total number of confirmed cases in mainland China by applying Boltzmann-function based regression analyses. We found that the cumulative number of confirmed cases from Jan 21 to Feb 14, 2020 for mainland China, Hubei Province, Wuhan City and other provinces were all well fitted with the Boltzmann function ($R^2$ being close to 0.999). The potential total number of confirmed cases in the above geographic regions were estimated at 95% confidence interval (CI) as 79589 (71576, 93855), 64817 (58223, 77895), 46562 (40812, 57678) and 13956 (12748, 16092), respectively. Notably, our results suggest that the number of daily new confirmed cases of SARS-CoV-2 in mainland China (including Hubei Province) will become minimal between Feb 28 and Mar 10, 2020, with 95% CI. In addition, we found that the data of cumulative confirmed cases of 2003 SARS-CoV in China and Worldwide were also well fitted to the Boltzmann function. To our knowledge this is the first study revealing that the Boltzmann function is suitable to simulate epidemics. The estimated potential total number of confirmed cases and key dates for the SARS-CoV-2 outbreak may provide certain guidance for governments, organizations and citizens to optimize preparedness and response efforts.

Introduction

In the early stages of an epidemic, forecasting the "ceiling" (total cases) and the "end date" is notoriously difficult. Traditional epidemiological models, such as the SEIR (Susceptible-Exposed-Infectious-Recovered) model, are highly effective but require complex parameters—like transmission rates and incubation periods—that are often unknown or fluctuate wildly during a new outbreak. This paper, published in February 2020, explores a simplified mathematical approach to forecasting the spread of SARS-CoV-2 by applying a function typically found in thermodynamics to epidemiological data.

Methods

The core of the study is the application of the Boltzmann function, a type of logistic or "S-curve" (sigmoid) function. In physics, this function describes the distribution of particles across energy states, but here it is used to model the growth of an epidemic toward a saturation point.

  • The Equation: The model uses a four-parameter regression: $C(x) = A_2 + \frac{A_1 - A_2}{1 + e^{(x-x_0)/dx}}$.
  • Key Parameter ($A_2$): This represents the estimated potential total number of confirmed cases at the end of the outbreak.
  • Uncertainty Analysis: To account for potential under-reporting or data noise, the authors used a Monte Carlo technique, running 1,000 non-linear regressions with a 10% standard deviation to generate 95% confidence intervals (CI).
  • Key Date Definition: The authors defined the "end" of the outbreak as the date when daily new cases dropped below 0.1% of the total estimated cases.

Key Results and Forecasting

The researchers found that the cumulative growth of the epidemic followed a predictable sigmoidal path when strict intervention measures were in place. The Boltzmann function showed an extremely high correlation ($R^2 \approx 0.999$) across different regions, suggesting it is a robust tool for simulating coronavirus epidemics.

Table 1: Estimated potential total number of confirmed cases and key dates for SARS-CoV-2 in various regions of China.

As detailed in the table above, the model fit the data for mainland China, Hubei, and Wuhan with high precision. For mainland China, the potential total number of cases was estimated at 79,589 (95% CI: 71,576–93,855). The model predicted that daily new cases in mainland China would become minimal between February 28 and March 10, 2020. These findings were further reinforced by historical validation; 2003 SARS data fit the model with $R^2$ values of 0.999 (China) and 0.998 (Worldwide).

Limitations and Future Work

While the Boltzmann function provides a powerful "quick-response" forecasting tool, it has notable limitations:

  • Assumption of Stability: The model assumes that overall conditions (like social distancing measures) remain constant. Any premature relaxation of these measures could cause the curve to deviate from the prediction.
  • Data Sensitivity: The model is highly dependent on the quality of reported data. For instance, a massive spike on February 12 due to changed diagnostic criteria in Hubei required manual data adjustment for a proper fit.
  • Single Wave Focus: This regression models a single "wave" and does not account for secondary outbreaks or long-term cyclical patterns.

Despite these limitations, the study demonstrates that simple, data-driven regression tools can provide reliable real-time forecasts to help health organizations optimize their response efforts during emerging infectious disease outbreaks.

Summary updated

Abstract

Abstract An ongoing outbreak of atypical pneumonia caused by the 2019 novel coronavirus (SARS-CoV-2) is hitting Wuhan City and has spread to other provinces/cities of China and overseas. It very urgent to forecast the future course of the outbreak. Here, we provide an estimate of the potential total number of confirmed cases in mainland China by applying Boltzmann-function based regression analyses. We found that the cumulative number of confirmed cases from Jan 21 to Feb 14, 2020 for mainland China, Hubei Province, Wuhan City and other provinces were all well fitted with the Boltzmann function ( R 2 being close to 0.999). The potential total number of confirmed cases in the above geographic regions were estimated at 95% confidence interval (CI) as 79589 (71576, 93855), 64817 (58223, 77895), 46562 (40812, 57678) and 13956 (12748, 16092), respectively. Notably, our results suggest that the number of daily new confirmed cases of SARS-CoV-2 in mainland China (including Hubei Province) will become minimal between Feb 28 and Mar 10, 2020, with 95% CI. In addition, we found that the data of cumulative confirmed cases of 2003 SARS-CoV in China and Worldwide were also well fitted to the Boltzmann function. To our knowledge this is the first study revealing that the Boltzmann function is suitable to simulate epidemics. The estimated potential total number of confirmed cases and key dates for the SARS-CoV-2 outbreak may provide certain guidance for governments, organizations and citizens to optimize preparedness and response efforts.
BibTeX
@misc{fu2020simulating,
  author = {Xinmiao Fu and Qi Ying and Tieyong Zeng and Tao Long and Yan Wang},
  title = {{Simulating and Forecasting the Cumulative Confirmed Cases of SARS-CoV-2 in China by Boltzmann Function-based Regression Analyses}},
  year = {2020},
  doi = {10.1101/2020.02.16.20023564},
  howpublished = {medRxiv},
}
Figure