4.1. Version 1

In the first layout (Figure 1), the format is square and not completely suitable for mobile phone screens, which are mostly rectangular. In other words, it does not use all the space on these devices’ screens. According to ^{Lupton and Phillips (1983/2008}), the grid is a type of internal organisation of the layout, dividing the space between vertical and horizontal lines to help in the construction and alignment of the elements against each other, depending on the project and the designer’s choice, but it can also be circular, angular and even irregular. The grid is a way to “arrange images and organise information” (^{Lupton & Phillips, 1983/2008}, p. 175) by occupying empty spaces and making the whole set have rhythm and scale, helping to compose the image as a whole all. In the case of the analysed infographics, it is possible to notice that horizontal and vertical lines were used, demarcating the space as a whole.

Source. From Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 30.03.2020, 15h40 [Photograph], by ^{Fundação Municipal de Saúde [@fmsrioclaro], 2020a}, Instagram. (https://www.instagram.com/p/B-XtkJ0H0P5/)

Figure 1 First published infographic (March 30, 2020) 

The infographic has many straight lines and two distinct main layers. The background glimpses parts of a stylised coronavirus as a watermark in lighter blue. With a larger font and white at the top, the second layer spells the bulletin’s name (“Situação de Rio Claro/SP”; Situation of Rio Claro/SP). The virus watermark is the only illustration in the infographic, which opts for geometric figures, especially rectangles, to break up the information, almost as if they were modules. For example, the module on suspects and reported cases; the module on hospitalisations; finally, the module on reported cases per age group, creating a relationship between them, the first module centred and above and the other two below, establish a reading flow. The colours of these information blocks, mostly blue, ultimately highlight only one module, the red one, referring to the item “deaths under investigation”. This highlight is mainly due to the use of primary colours. According to the colour circle theory (^{Heller, 2000/2013}), the use of primary colours and their opposite colours - in this case, the predominance of blue with a red point -, generates the feeling of contrast. The use of primary colours (blue, red and yellow) and secondary colours (green) and their shade variations, especially blue, will be maintained throughout the infographics updates.

On April 9, there is a small change: the first death in the city is reported in a red column, and deaths under investigation are in black (Figure 2). After that, the confirmed cases column also becomes red, creating a new highlight in the image.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em *09.04.2020, 16h30* [Photograph], by ^{Fundação Municipal de Saúde [@fmsrioclaro], 2020b}, Instagram. (https://www.instagram.com/p/B-xl_L0nxbw/)

Figure 2 First change in the infographic (April 9, 2020) 

We believe that the choice for red in this first pandemic wave is related to the notion of danger to which this colour refers, built from the physical concept of optical dispersion - the red colour has the longest wavelength and is one of those that do not disperse much, remaining in a range visible to humans (^{Halliday et al., 2003/2012}). From the standpoint of physics, this feature is especially important in sending warning and danger signals when the “visibility” in the surroundings is low and hazy - by analogy, the case of COVID-19 in the first months of the pandemic.

On the other hand, neuroscientists ^{Kuniecki et al. (2015}) claim that red can engage the attention of humans in emotional contexts, increasing the capacity and speed of their motor response to an unexpected event. According to the study, this colour seems to guide attention, specifically in emotionally-valenced circumstances, with red having both negative (e.g., blood, fire, danger) and positive (sex, food) connotations. In a nutshell, on seeing the red colour, the human mind is immediately activated or alerted, effectively drawing attention, especially in conditions that demand quick responses.

On April 11, the column of “confirmed” cases changes to “positive”, and a new column (“positive by rapid test”) is introduced, in yellow, with a note indicating tests taken in a private laboratory, information that remained until May 27. Although the tone chosen is not intense - it is not in the primary colour but has a variation, a tertiary colour - the information stands out in contrast to the predominant shades of blue, white and black. Its place in the layout (Figure 3), right in the centre of the infographic, also works as a focus point.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 11.04.2020, 17h. *Os casos POSITIVOS VIA TESTE RÁPIDO (Laboratório Particular) não são considerados resultados finais de diagnóstico* [Photograph], by ^{Fundação Municipal de Saúde [@fmsrioclaro], 2020c}, Instagram. (https://www.instagram.com/p/B-2rG8unplh/)

Figure 3 Second change in the infographic (April 11, 2020) 

On April 14, another small change is introduced: the column which previously indicated reported cases by age is now called “positive by age group”. In this change, red is used in a third column to show the confirmed cases by age group. Thus, in this version, all information related to case confirmations and deaths is red (positive, confirmed deaths, and positive by age group; Figure 4). Considering the fear that SARS-CoV-2 infection raised, the infographic’s intention seems to be to warn of the risk of contracting the disease and falling victim to a fatal outcome.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 14.04.2020, 17h [Photograph], by ^{Fundação Municipal de Saúde [@ fmsrioclaro], 2020d}, Instagram. (https://www.instagram.com/p/B--dA03HAIV/)

Figure 4 Third change in the infographic (April 14, 2020) 

It is important to note that between March 30 and the end of April, the publication of more than one bulletin per day was common, with up to three daily issues on March 30 and 31 and on April 4, 5, 6, 7, 12, and 29. Analysing these additional issues, we see an effort to update new cases or deaths based on the municipal health surveillance notification system. From May on, the bulletin is released once a day, usually in the late afternoon, between 4 and 6 pm. Throughout the analysed period, the bulletin was not published only on May 3 and 9.

The health crisis caused by COVID-19 was unprecedented, and much was learned as the pandemic evolved. The changes included in this first version suggest that some contents that had not been initially foreseen were added, indicating the team’s flexibility to add new information that showed the management actions and what was happening in the municipality at each moment. Including the time of day in which the bulletin was published seems to demonstrate that the numbers should be read considering the hours and minutes since the situation could change quickly.

4.2. Version 2

On May 7, the infographic’s layout changes significantly. The square format remains, but the tone is brighter and lighter (Figure 5). A lighter, brighter blue replaces the dark blue of the background. A new layer was added, framing blocks of initial information, although the original layout remains similar. The information is now displayed in a single block with a white background. This new layer is also highlighted by a combination of shadow and movement, adding depth to the image. The relationship between the areas of the infographic, although now in other tones and with greater prominence (the white background), follows the same previous flow - a larger module centred on top and two smaller blocks below.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 07.05.2020, 16h20. Deste total de internados, 05 estão em UTI [Photograph], by ^{Fundação Municipal de Saúde [@fmsrioclaro], 2020e}, Instagram. (https://www.instagram.com/p/B_5o6rknLwj/)

Figure 5 New infographic template (May 7, 2020) 

In this version, there is an attempt to make the infographic cleaner and more objective. The only direct reference to the coronavirus in the previous version, the sentence “protect yourself against respiratory viruses, including COVID-19”, is removed. That may suggest that the people who prepared the infographic believed the disease was already naturalised and the public knew the numbers referred to COVID-19.

New colours are added to the block: green and two shades of grey. The colour green has gained prominence, often displayed to present information quickly. In the case under analysis, the column in green presents the number of recovered cases, a controversial piece of data from the epidemiological point of view since, in practical terms, it does not broaden the reading on the evolution of an epidemic event. However, since the beginning of the health crisis, it has been used by different institutions, including the academicscientific. The use of green for this information aimed to speed up the process of assimilation of an “encouraging” message amid pandemic chaos, in other words, that the dreaded disease had many more cured than dead. The grey is used as a background for the interfaces, thus minimising the contrast between the darker and lighter colours of the image, reducing visual fatigue when moving from one to the other (^{Farina et al., 2006}).

On May 27, 2020, the column “positive rapid test” is excluded, and the positives are now counted indistinctly, still in red. In this version, without this previously mentioned item, yellow is removed from the colour palette, which now remains in shades of blue, red, green and grey, besides white and the font in black and white (Figure 6).

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 27.05.2030, 17h [Photograph], by ^{Fundação Municipal de Saúde [@ fmsrioclaro], 2020f}, Instagram. (https://www.instagram.com/p/CAtJXRuHaP3/)

Figure 6 First change in the infographic’s second version (May 27, 2020) 

On June 22, the information includes the total number of people admitted to wards and intensive care units (ICU), divided into public and private institutions. Here, the layout of the elements, now with added information, is more “squeezed”, with less space between the margins of each line. In this new division, different intensities of the same colour are again used to highlight information in the same “block”. The “total inpatients” item is shown in medium blue, followed by the “public” and “private” items in a lighter shade of blue. The “ICU” column follows these in a darker blue, and the items “public” and “private” in a lighter shade of blue than the previous one but more saturated than those of the inpatient item (Figure 7). Making a visual column in a gradient - first, the lightest block (admissions) followed by the darkest block (ICU) - trying to highlight also the seriousness of beds occupation, since an ICU admission bed demands more care than a clinical admission bed, thus having a certain gradation of intensity and severity of admissions.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 22.06.20, 17h [Photograph], by ^{Fundação Municipal de Saúde [@ fmsrioclaro], 2020g}, Instagram. (https://www.instagram.com/p/CBwK6jmH6o4/)

Figure 7 Second change in the infographic’s second version (June 22, 2020) 

The analysis of the content included in the infographics up to that point shows they were still changing as the pandemic evolved in the municipality. Hence, information that no longer made sense in that context was excluded. For example, given the expansion of rapid tests in the public system⁴, displaying the number of positive cases obtained through this method in private laboratories no longer made sense.

The block’s title “number of suspects hospitalised” may confuse the reader since it reads the number of people hospitalised, including in the ICU. That is, they are not “simple” suspected cases. Perhaps they did not have their diagnosis laboratory confirmed (the context of that time was the challenge of laboratory diagnosis, although all people with symptoms were considered to have COVID-19). However, this number refers to admission, that is, to people with aggravated disease symptoms and requiring hospital care (clinical or intensive care). Therefore, considering them as “suspects” does not explain what is being reported.

Within this block, information on the occupation of public and private beds was included because, as the number of patients increased (as shown by the number of inpatients), it became necessary to inform the situation of the health network as a whole.

4.3. Version 3

On July 2, the infographic’s layout has a third major change. The format is now rectangular and more suitable for mobile phone screens. This new version is different from the previous layout. It changes the scheme of internal organisation and the relationship between the different areas of the infographic. Instead of a block condensing all the information with more defined internal lines, the new design introduces a new division in the internal patterns (even though it is divided into blocks). The new arrangement is a larger, centred block covering two-thirds, a highlighted horizontal strip at the bottom corner of this block, and a smaller, centred module at the bottom of the infographic.

The use of the same colours recalls the previous template, but the organisation of the internal elements changes to a larger upper block, subdivided into two, the one on the left showing a very objective and direct flowchart, in which the number of positive cases covers a large column in red, unfolding into a flowchart: blue for cases receiving treatment; green for recovered cases; and black for deaths.

The right side comprises blocks of information referring to hospitalisation or home isolation. However, this side has much information, arranged and organised, perhaps seeking a “ranking” of intensity: green for cases in home isolation and different shades of blue for hospitalised cases. However, this “order” is broken because the ward is in a darker shade than the ICU, and this gradation seems broken. Also, within each module (ward and ICU), there is another colour gradation, though here to highlight and differentiate the public beds from the private ones. Maybe an alternative here would be grading the ward beds in the lightest colour, without any difference in colours, but keeping the module and the ICU block in the darkest colour, which could convey the idea of intensity more clearly, as in Version 2 (Figure 7).

Although still not making explicit textual reference to COVID-19, in this layout, it is possible to observe the comeback of a three-dimensional image of the virus as a watermark in the lower right corner and several viruses in the background of the title (“Situação de Rio Claro/SP), in the upper margin of the infographic. This template (Figure 8) remained in use until July 31, when we concluded the research.

Source. From Warning Situação de Rio Claro atualizada via informação da Vigilância Epidemiológica em 02.07.20, 17h [Photograph], by ^{Fundação Municipal de Saúde [@ fmsrioclaro], 2020h}, Instagram. (https://www.instagram.com/p/CCJ3GibHiQd/)

Figure 8 Infographic published from July 2, 2020 

There is a shift in the use of colour tones; the design resumes the “flatter” image of Version 1, without any combination of light and shade, using only superimposition - especially the grey band with the items “deaths under investigation”, “suspects under investigation” and “discarded”. New colours are now used compared to the previous versions. Yellow returns, but still in a tertiary tone, representing the item “suspects under investigation”. Besides the light green (highlighted), there is another shade of green, a bluish green or blue-green (tertiary tone), making the predominant shades remain blue.

Interestingly, this version, published 65 days after the first infographic, is more objective and synthetic than previous ones. A good example is the exclusion of the designation “confirmed deaths”, which at that time already totalled 33 and is now presented only as “deaths”. The same occurs with the number of admissions: they are presented as more organised, and the title informs that it is about overall hospitalisations, including suspected cases.

As we showed throughout the analysis, the layout and content of the infographics changed according to the need to exclude or include new information. The team responsible for this dissemination was apparently testing the models over time, according to the demands imposed by the succession of events and the uncertainties typical of extreme health events, such as the pandemic of COVID-19. This last design, launched on July 2, seems to have been the most consistent, not having undergone any changes until the end of our analysis on July 31.

One aspect that drew our attention was that in all versions, the infographics sought to display the number of positive and/or reported cases by age group. It seems there was an interest in alerting the population to the spread of the disease in different population groups. Furthermore, it implicitly pointed to the need for specific prevention measures for these groups.

As for the use of colours, the infographics seem to have followed basic design aspects. For ^{Kress and Van Leeuwen (2006}), colours are a materiality that has several uses and produces many meanings and effects. They do not have “fixed functions”, and their meanings change according to sociability and culture. It was possible to observe in the infographics that the solid colours highlighted the information according to the nature of the different and concurrent events arising from the epidemic: red to highlight the positive cases and confirmed deaths; yellow, first, for positive cases via rapid test and, then, suspected cases; and green, for recovered cases. This choice seems to allude to the traffic lights’ colours, danger/stop, attention and go ahead. It is worth noting that blue, in its various shades, remained in the three designs. Widely used in official communication of the municipality, the colour refers to the slogan of Rio Claro, “blue city”.

The texts play a key role in using graphic resources for data visualisation since they complement each other (^{Unwin, 2020}). For the infographic used to convey to the population information about COVID-19 in the municipality of Rio Claro, some aspects seem critical to us and need to be discussed, as they hinder the understanding and/or confuse the reader. Namely the title, the use of some words such as “suspects” and “investigation”, which appear in all versions, sometimes more than once, and the number of hospitalisations.

The choice of the title “Situação de Rio Claro”, which despite all the changes in the layout remained over time, always without explicit reference to COVID-19, does not meet the function of informing clearly and objectively what the infographics are about. The word “situation” is used in the dictionary sense as: “a condition or combination of conditions that exist at a particular time; conjuncture” (^{Houaiss, 2001}, p. 2587). The definition alone indicates, therefore, that it would be necessary to inform to which event the word was referring. One possibility that could explain the absence of a title is that in some bulletins, the background watermark of Sars-CoV-2 (an image widely used in the material of the Ministry of Health and made popular by the mainstream media in Brazil) would be used to inform on its own that it was the COVID-19 pandemic. However, this assumption does not hold because it was not kept in all layouts and was lost in the overlapping colours and information. Thus, despite making it explicit that they were official data, since all infographics displayed the logos of the Unified Health System, the Municipal Health Foundation and the city of Rio Claro, to broaden the understanding and scope of the image, the title should explain that the infographic was about the situation of COVID-19 in the city.

Although the infographics were designed to inform the general population, the language used remained too attached to epidemiology and health jargon. Clearly, in the information on reported cases, they excluded the word “reported” in the first update of Version

2. That is an important change because “reported” is not a familiar repertoire for most of the population. However, they did not make this effort with other terms such as “suspects” and “investigation”. These two terms are directly associated with crime and the notion that something is wrong and hidden, requiring detective work to uncover. Thus, having been used in infographics that sought to inform the general population, ^{Finset et al. (2020}), discussing effective health communication during the COVID-19 pandemic, state: “it is important to provide information in lay language that is clear, specific, unambiguous and consistent” (p. 874). Furthermore, it is important to consider the discussion presented by ^{Oxman et al. (2022}) about the communication of public agencies, in which they point out that it needs to be transparent and trustworthy, which also avoids “victim-blaming” by using terms or threats that contribute to the stigmatisation of people with certain pathologies.

Using absolute numbers seems helpful for most aspects the infographics wanted to inform. However, in the case of admissions, the total numbers or divided into people admitted to the ICU and in wards and public and private beds was information of little practical use since it did not give the number of beds available in the municipality, making it impossible to identify the actual bed occupancy rate.