Public health interventions are everyday actions that the public can take to stop the spread of an infectious disease. As indicated by the word in the parentheses, taking medication or vaccination is not considered a public health intervention.

In view of the fast spread of the Delta variant, China has initiated stricter border control.

Read more in South China Morning Post: https://www.scmp.com/news/china/politics/article/3143989/coronavirus-china-steps-border-controls-race-contain-fast

Tracing and quarantining suspected and infected cases as well as imposing restrictions on high-risk social settings have been shown necessary.

Read more in Nature Medicine: https://www.nature.com/articles/s41591-020-1092-0

Wesolowski and the team used mobile phone data to map human mobility in Pakistan in 2013, the year this country saw large dengue outbreaks. They found that human mobility within the country can predict the spread and the timing of epidemics.

The authors analyzed 335 emerging infectious disease incidents between 1940 and 2004 and found that most had originated from wildlife and low-altitude regions. They also revealed an alarming lack of reporting effort in areas that are "hotspots" of these diseases.

Brockmann and Helbing proposed effective distance, as a replacement for conventional geographic distance, for simplifying the transmission patterns of global epidemics.

The proposed approach was successfully applied to data on the worldwide 2009 H1N1 influenza pandemic and 2003 SARS epidemic.

This comment article summarized the transmission control measures that had been taken in China, suggesting their potential effectiveness. Space is reserved for further quantitative research, though.

In the previous sections, the authors have studied the effectiveness of two types of interventions (the Wuhan lockdown and the transmission control measures) individually. In this section, they move on to simulate the transmission patterns if no interventions had been taken, only one of the two types of intervention had been taken, and under the synergetic effect of the two.

The replacement rate of 1 means that one case in a completely susceptible population can lead to one secondary case. Assuming that the initial case can be cured and reach full recovery — which is what all of us hope — a rate equal to 1 means that one is infected when another gets well, leading to an unchanged number of cases.

The authors used an SEIR model for the fitting (with more mathematical details presented in the Supplementary Materials). In this model, the entire population is divided into four groups: susceptible (S), exposed (E), infectious (I), and recovered (R). Susceptible people are those who can be infected. People who are infected but have yet to show any symptoms or become infectious themselves are exposed. After the incubation period, the time taken to develop symptoms, people become infectious. Lastly, having beaten the disease and become immune to it, people are recovered.

To confirm the conclusion drawn from the regression, the authors tried a different way of computing the regression coefficient, confidence interval, and P value. The consistency between results derived using two methods indicates that the results present here are robust to the choice of the regression model.

With the effectiveness of early interventions in slowing the spread of COVID-19 first demonstrated by the Mann-Whitney U test, the authors further conducted a regression analysis to reveal the association between several factors and the number of laboratory-confirmed cases. You can find more details of the regression in the Supplementary Materials.

The Mann-Whitney U statistic can be seen as the cumulative result of comparing the randomly drawn values from two populations. z is the standardized value and is dependent on the sizes of the two populations. P is the probability for a null hypothesis to be true.

The Mann-Whitney U test is a test for a null hypothesis that it is equally possible for a value drawn from a population to be greater or smaller than another from a different population. This test is useful when the distributions of values are unspecified. In this study, one population comprises cities that took action earlier, and the other is for cities that responded later.

The result shows that early intervention can decelerate the spread of COVID-19.

To calculate the percentage used here, the authors first calculated the difference in the average numbers of laboratory-confirmed cases for two groups of cities. The difference was then divided by the number of cases for cities that introduced controls later.

A P value below 0.01 means less than a 1% chance that the null hypothesis is true. Typically, a P of 0.05 or below suffices to conclude a significant correlation between two variables.

This probability is derived through a Student's T-test. This test requires the correlation coefficient (r) and the number of samples (n) as inputs.

P is the probability that the null hypothesis is true. The null hypothesis, in this case, is that the total number of cases reported from each province shows no significant linear correlation with the total number of travelers from Wuhan.

r is the correlation coefficient for an association between two factors. It can take values between -1 and 1. A correlation coefficient equal to -1 indicates a perfectly inverse linear correlation, meaning that one variable decreases in its value in response to the increase of the other linearly. In contrast, an r equal to 1 indicates a perfect linear positive correlation. An r of zero signifies the nonexistence of an association.

COVID-19 spread fast. To show this, the authors compared the outbreak data for COVID-19 with those for H1N1, another contagious disease, in China.

By referring to the cumulative number of cities that reported COVID-19 cases and comparing the spreading trend to that of the H1N1 pandemic, the authors concluded that the spread of COVID-19 was rapid.

To investigate the impact of the Wuhan travel ban on the spread of COVID-19, the authors first evaluated the effectiveness of this ban on stopping the movement of people.

The evaluation was realized by comparing the travel data from the year 2020 with those from the previous two years.

Note that the start and end of the time period of records are fixed with reference to the date of Spring Festival, which, however, varies from year to year.

For the quantitative analysis—linear regression in this work— the input data include official reports of the health commission, mobile phone data, and travel information recorded in online databases.

For details regarding the time span of the data sets and the means of data acquisition, please refer to the first two sections of the Materials and Methods described in the Supplementary Materials.

The authors focused on the first 50 days since the detection of the first COVID-19 case in Wuhan. They summarized the timeline for the implementation of key control measures over this period in Fig. 1.

Measures taken by China have been proven effective in a list of countries that carried out lockdowns.

Read more in Politico: https://www.politico.eu/article/coronavirus-europe-lockdown-effectiveness-graphics/

The activation of Level-I alert in Tibet marked the nationalwide activation of this highest-level alert.

Read more in China Daily: https://www.chinadaily.com.cn/a/202001/29/WS5e318a36a3101282172739c1.html

A cordon sanitaire is a movement restriction of people into or out of a specific region. This measure is taken to stop the rapid spread of an infectious disease.

A pathogen is any tiny organism that causes disease.

The first number in the parentheses is the mean for the number of reported cases, and the range after the semicolon is the 95%CI.

The Wuhan travel ban has also been found to slow the spread of COVID-19 worldwide.

Read more in Science: https://science.sciencemag.org/content/368/6489/395

Agents are the causes of diseases and injuries, but they are not the sole determinant for the occurrence of a disease. The other two factors are the host (the human who can get the disease) and the environment that brings the agents and the host together.

Viboud and colleagues found a strong correlation between people's traveling rate between their workplace and home and the spread of infection. A gravity model can describe the workflow.

The authors used mobile phone data to depict the travel patterns of people in Kenya over a year. They found the exportation of malaria from specific regions to others. The finding suggested human mobility as an essential factor in the spread of malaria.

The authors conducted a correlational study using mobility data from Wuhan and case data from different regions of China.

The spatial distribution of confirmed cases of COVID-19 before the implementation of travel restrictions was found related to the human mobility data. Such correlation was diminished after the implementation, and the growth of COVID-19 in many areas turned to a decline.

Wang and co-workers synthesized the up-to-date understandings of COVID-19 by then. They summarized the time of the early-stage outbreak, the characteristics of infected patients, and control measures that had been taken by the time.

Lu and the team collected samples from nine patients presenting symptoms of viral pneumonia.

They then used the samples for DNA sequencing. The full-length sequence reported in this study confirmed the distinction of the newly reported coronavirus from the previously reported ones.

Zhu and colleagues isolated the virus that had caused the outbreak of an epidemic in Wuhan, China.

They captured images of the virus and performed DNA sequencing. Based on the results, they reported the identification of a new type of coronavirus.

The authors summarized all major findings in this paragraph.

The simulated result agrees with the actual data, revealing that simultaneously taking both measures may be the reason for the reduction in the total case number.

The simulated patterns of transmission with single or no intervention taken are different from the actual scenario shown in Fig. 4A. The discrepancy means that neither of these interventions could overturn the spreading and growth of COVID-19 in China when implemented individually.

The significant reductions in the case reproduction number due to the regional (stage 1) and massive (stage 2) implementation of transmission control measures point to the importance of intervention in stopping the epidemic.

The basic reproduction number is the average number of secondary cases that can be produced by one case in a completely susceptible population. \(R_0\) higher than 1 indicates that an epidemic will continue, and \(R_0\) lower than 1 is a sign for the epidemic to end.

The transmission control measures were also found effective in reducing the total number of cases.

Reaching the peak of the number of confirmed cases means that an epidemic stops growing and turns to an end. Thus, reaching the peak sooner is desired in this case.

The coefficients in Table 2 are statistically significant. This means that each coefficient is significantly different from the value under the null hypothesis of no correlation, indicated by the probability for the null hypothesis to be true, P, smaller than 0.01.

These results indicate correlations between the factors shown in the table and the number of laboratory-confirmed cases.

This is the sample size. In this case, n is the number of cities that fall into a category.

The authors summarized measures taken by different cities and the timing of responses.

The differences in the types of measures and the timings of their implementations offer directions for the correlational studies present in the later sections of this paper.

The Wuhan lockdown was found to effectively delay the spread of COVID-19 to other cities in China.

Based on the statistically significant correlation shown in Figure 2C, the authors concluded that the outflow of people from Wuhan before the shutdown is positively associated with the spread of COVID-19.

This finding lays a foundation for the later examination of the effectiveness of the Wuhan shutdown in slowing the dispersal of COVID-19.

We need to take r values with caution, though. Calculating r values is still possible even for plots where two variables are not linearly associated. It is always wise to look at the scatter plot an r corresponds to.

In the present study, the scatter plot in Figure 2C indeed shows a clear linear association.

A sharp drop in the number of movements from Wuhan confirmed the effectiveness of the travel ban.

Epidemiology investigates the distribution and determinants of health-related states or events concerning specific populations. The study guides the control of health problems.

By reviewing the most updated understanding of COVID-19 at the time, the authors identified a research gap concerned with the transmission patterns with and without interventions taken.

A better understanding of this aspect is necessary for evaluating the effectiveness of public health measures.

The present work addressed this question by studying how the spread of COVID-19 was impacted by the travel bans and the closure of entertainment venues. The travel bans encompass the one for Wuhan city and the suspension of inter- and intra-city public transport.

Three of these studies investigating the spreading patterns agree that human mobility is a critical factor in determining the spreading pattern of these infectious diseases, e.g., malaria, H5N1 influenza, etc.

These two articles summarized the measures introduced in the early stage of the outbreak.

Kraemer and the team showed that the mass control over transportation effectively restricted the spread of COVID-19.

Based on the RNA sequencing results, both works suggested that the novel coronavirus is closely related to a previously identified bat virus.

Etiology is the medical study of the causes of disease. An etiological agent refers to the origin identified.

The two studies presented images and genome sequencing results of the novel coronavirus. The results agree on the identification of a new type of human-infecting betacoronavirus.

The confidence interval is a range that is likely to contain the true population parameter with a confidence level specified by the percentage.

In this case, there is a 95% probability that the confidence interval of 2.54-3.29 contains the average delay of arrival of COVID-19.

Coronaviruses are a family of viruses. This name comes from the solar corona-like characteristic appearance of these viruses under the electron microscope.

Outbreaks are the occurrence of a more-than-expected number of cases.

An epidemic is the appearance of a disease in a large number of people within a short period of time.

Geocoding is a process to transform commonly text-based descriptions of locations (e.g., addresses) to coordinates on the Earth's surface. Geocoding allows further mapping and spatial analysis using various software. In a geocoded repository, all data are attributes to spatial coordinates.