This study’s data were derived from Qingbo Big Data’s platform, which includes all COVID-19 information released by the People’s Daily official WeChat account and the dissemination index of each piece of information. The research cycle covered a total of 105 days, exporting all COVID-19 information from January 19, 2020, which is when the People’s Daily first released COVID-19 information, to May 2, 2020, which is when China’s 34 provinces (ie, autonomous regions and municipalities) completely lifted the first-level response to public health emergencies [26]. This period was chosen because it covers the entire duration of the COVID-19 pandemic, from the mainstream media’s first report to the subsequent development and normalization of the COVID-19 pandemic. The public experienced information scarcity, caused by few media reports being released at first, which was followed by an information explosion due to the media’s collective focus. Finally, a public opinion vortex was caused by the surplus of COVID-19 information [27]. This period comprehensively covers the whole process of COVID-19’s information dissemination, including each stage of communication [28].

This study included information that was related to COVID-19 and published within the study’s established period. Any information that did not meet these criteria was excluded. All the information about COVID-19 that was released by the official WeChat account of the People’s Daily was included in this study. Specifically, out of the 1984 pieces of information that the account released, 1621 items were related to COVID-19.

This study used SPSS, version 21.0 (IBM Corp), to conduct its statistical analyses. Descriptive analysis was used to describe the basic characteristics of the COVID-19 information that was released by the People’s Daily, and it describes the information content with a higher dissemination power, according to the dissemination trend chart in the Results section. A multiple stepwise regression analysis was used to analyze the factors that affected COVID-19 information dissemination. Dummy variables were manually established for the multi-categorical disordering of variables, where the dummy variables were grouped under the same factor in the same block, then “ENTER” was selected as part of the inclusion method to ensure that these dummy variables entered and exited at the same time. Other continuous variables and binary variables were grouped into another block, and the inclusion method was “STEPWISE.” Then, a multiple linear regression was performed after each setting. The model’s goodness of fit was checked as well as whether there was a collinearity problem.

As per the protocol at Weifang Medical University, China, the Institutional Review Board does not review studies that do not involve human subjects. As this study did not include human subjects, ethics approval was not required.

The data sets used and analyzed during this study are available from the corresponding author on reasonable request.

From January 19 to May 2, 2020, the People’s Daily released 1621 pieces of information related to COVID-19 (Table 1). In terms of media salience, this included 1129 headline news items (69.65%) and 492 nonheadline items (30.35%), which refer to prominence. This can also be divided into 1023 emotional information items (63.11%) and 598 neutral information items (36.89%), which refer to valence. The number of words for each information item was primarily between zero and 500. In terms of content, the main theme concerned the development of COVID-19 (410/1621, 25.29%). Meanwhile, the sources were primarily government agencies (1478/1621, 91.18%), and most of the information was original content from the People’s Daily, including 923 items (56.94%). In terms of the information’s format, the release time was primarily between 6:01 AM and 6 PM, and it was mostly presented as text + pictures (957/1621, 59.04%). Furthermore, the number of fonts and colors ranged between 1 and 6.

During the COVID-19 epidemic period, the People’s Daily released several pieces of information related to COVID-19 every day, and the dissemination index for each piece of information represents its dissemination effect. Based on the time series analysis of the average dissemination index for all of the information released daily, we created a chart depicting the information dissemination trends. According to this chart, there were seven peaks of information dissemination from January 19, 2020, to 12 AM on May 2, 2020 (Figure 1). The analysis of these peaks and the content related to key events that occurred during these peaks revealed numerous insights.

The first peak in the dissemination of COVID-19 information occurred around January 21 (mean dissemination index of 91,101.43). The People’s Daily and other media began to report on COVID-19, which aroused public attention; in addition, the National Health Commission released the first announcement in 2020, which categorized COVID-19 as a Class B infectious disease, while adopting management measures as per the guidelines for more serious Class A infectious diseases. Specifically, when Zhong Nanshan confirmed that COVID-19 could be transmitted from person to person, the first peak of information dissemination was triggered. The second peak occurred around January 26, when the news reported that many new coronaviruses had been found through an epidemiological retrospective analysis [29]. The third peak occurred on February 2, when the news reported that the construction of Huoshenshan Hospital had been completed and the hospital was taken over by the Chinese People’s Liberation Army. The dissemination index on that day was 92,980.18. The fourth peak of information dissemination occurred on February 7, when Li Wenliang passed away, arousing strong public concern. The People’s Daily released five news items to commemorate him, and the State Supervision Commission appointed a team to investigate his death. The fifth peak happened on February 10, when the medical team’s support for Wuhan reached its peak, and nearly 6000 medical workers arrived in Wuhan, the largest number of medical staff to arrive in Wuhan since the start of the COVID-19 epidemic. The sixth peak occurred on March 21, when no new cases had been reported in either Wuhan or the Hubei Province for the first time. In fact, there were no increases in cases for the first time in 31 provinces or in Xinjiang Production and Construction Corps. The news that the epidemic situation had improved aroused public concern. The seventh peak occurred around April 3, which was a period of national mourning. The People’s Daily released some messages in memory of the dead, which triggered an emotional resonance among the public and led to a peak of information dissemination (mean dissemination index of 90,338.12).

Since then, the information dissemination index gradually decreased. Experts on infectious diseases and other public health emergencies found that the public have a strong curiosity and demand for information when they have a low awareness of diseases [30]. The distribution of peaks of information dissemination during this time reflects the public’s psychological needs during an epidemic, illustrating that the public will seek the truth about an event at its early stage. At the middle stage, they will question some of the events that occurred, and at the epidemic’s late stage, due to its normalization, the public’s concern will gradually decline. The whole process presents a regular development trend [31].

The dimensions of media salience, information content, and format contained 10 factors. A multiple stepwise regression analysis was conducted with these 10 factors serving as independent variables and the dissemination indices as dependent variables. We found that eight variables entered the equation: prominence, valence, attention, theme, source, vividness, originality, and the number of fonts (Table 2). As the tolerance of the equation was greater than 0.1 and the variance inflation factor was less than 10, we could determine that no collinearity existed between independent variables (ie, the credibility of the multiple linear regression was high) [32]. As shown in Table 2, in terms of media salience, the dissemination of headline news items was better than that of nonheadline items, while the dissemination of emotional information was better than that of neutral information. Further, the fewer words that a piece had, the better its dissemination. In terms of informational content, the following variables affected the dissemination of COVID-19 information (P<.05): theme, source, and originality. In terms of the information’s format, its vividness affected its dissemination; specifically, this pertains to whether information was presented by text, picture, or video. The number of fonts utilized also affected dissemination, as more font variety equated to a superior dissemination effect. However, the release time and number of colors had no significant impact on the dissemination of COVID-19 information.

Based on results of the WCI, this study examined the COVID-19 information that was first spread by the People’s Daily. This was conducted by analyzing the information’s media salience, content, and format. Thereafter, we summarized the COVID-19 information that was popular during the pandemic period. Seven information dissemination peaks were discerned. We further analyzed the influencing factors and key elements of COVID-19 information dissemination. We found that among the three dimensions of COVID-19 information—media salience, content, and format—eight factors affected the spread of COVID-19 information.

During the COVID-19 pandemic period, most of the information released by the People’s Daily focused on the pandemic’s development, with daily announcements about the number of confirmed cases, which kept the public abreast of related developments and helped dispel rumors. The public WeChat account of the People’s Daily mostly took the form of headlines upon releasing COVID-19 information, and most items had positive emotional tendencies. It seems that extensive positive reports played a role in stabilizing people’s emotions [33]. Most of the information came from the national and local health commissions, which ensured that the People’s Daily portrayed authority and fairness [34]. The People’s Daily is the most authoritative official media channel in China and, as such, the WCI of its official WeChat account ranked the highest during the epidemic. It was the responsibility of official media outlets to release relevant information about COVID-19 in a timely manner to spread relevant knowledge. This helped the public to have an awareness of the pandemic and improved the situation in China [35]. This shows the network media’s power when it comes to disseminating COVID-19 information [36].

By analyzing the trend regarding the average daily dissemination index of information, as well as content and key events, we found that there were seven peaks of information transmission during the COVID-19 epidemic period, which began on the following dates in 2020: January 21, January 26, February 2, February 7, February 10, March 21, and April 3. These peaks were primarily concentrated on the following issues: the media’s first reported coverage about COVID-19, epidemiological studies that confirmed the existence of human-to-human disease transmission, and government measures to manage the epidemic. These findings are consistent with research conducted by Liu Lanlan, who found that the dissemination peaks corresponded to January 26 and other dates similar to those found in this study, and information mainly concentrated on Zhong Nanshan’s confirmation of COVID-19’s transmission via person-to-person contact and Li Wenliang’s commemoration, among other key events [31]. The WCI’s peak value shows where the public’s attention was focused and their psychological need for epidemic information. Thus, the mainstream media should release COVID-19 information in a targeted manner and disseminate the type of information that the public values the most [37]. It should also improve the timeliness of information release, educate people on preventive measures more effectively, and improve people’s health literacy so they can better manage COVID-19 [38].

Our research shows that media salience, information content, and format affect COVID-19 information dissemination, specifically in terms of eight variables: prominence, valence, attention, theme, source, vividness, originality, and number of fonts. However, the time that the information was released and its number of colors had no effect on dissemination. These results suggest that different types of epidemic information disseminate differently; therefore, we must consider the factors affecting information dissemination and choose the best dissemination practices during the process of COVID-19 prevention and control to ensure that information is spread effectively and in a responsible manner. This would thereby prevent the spread of rumors and false information, improve the information’s scientific and authoritative reputation [39], increase the ability to manage information and public opinion during epidemics, and contribute to a comprehensive public health system [40].

This study has several strengths regarding its design and analysis. First, the selected research objects were relatively representative. The WCI of the People’s Daily official WeChat account showed that it was top-ranked among all WeChat official accounts, and this account is an important channel for the Chinese people to obtain official COVID-19 information. Second, based on big data technology, the WCI obtains data from all WeChat users (more than one billion) [41], resulting in a large sample size and high accuracy; therefore, the research findings are representative. Finally, the findings have practical significance; for example, this study found that headline news item dissemination has more impact than nonheadline items. These findings could inform the accurate dissemination of COVID-19 information and knowledge in the future. However, several limitations should be considered when interpreting the results. Because of the large number of potential factors and limited related research, we selected the three most commonly used dimensions—constituting 10 variables—from existing research, which may limit the research results. In addition, this study mainly analyzed the most popular social media platform in China. Although this study serves as an example for other countries to study COVID-19 information dissemination, it may not be completely applicable to other countries. This limits the inferences one may make from these results. Future studies can use more comprehensive potential factors to verify or enrich the conclusions of this study.

Currently, only a few indicators have been explored to quantify the effect of COVID-19 information dissemination. It is difficult to measure the spread of information when using quantitative data. It is even more complex as it occurs in conjunction with the unexpected and uncertain nature of COVID-19. This leads to only a few studies in the literature that have attempted to understand the effect and measurement of COVID-19 information dissemination. Thus, to fill this gap, we measured the dissemination of COVID-19–related information using WCI data, integrating the areas of public health and communication science (ie, media salience), to determine the factors affecting information dissemination. This study showed that the different types of pandemic-related information had varying dissemination power. It also revealed the factors that affected this dissemination: media salience, information content, and information format. These findings suggest that for effective dissemination of COVID-19 information, we should pay attention to the factors that affect this dissemination, which include the theme, source, and originality of COVID-19 information. Furthermore, we should disseminate the type of information that the public is most concerned about, choose those dissemination practices that effectively spread information in a responsible manner to curtail misinformation, regulate fear among the public, and improve information governance. Further research is needed to confirm the findings of this study and to validate the impact of these findings on the effect of COVID-19 information dissemination in other countries.