Panorama of COVID-19, risk perception and vaccine confidence in São Paulo State population, Southeast Brazil.

In 2021, the COVID-19 pandemic remained uncontrolled in Brazil, with more than 17.6 million cases diagnosed and more than a half a million deaths in total. It is known that an accurate risk perception by populations and a consequent adherence to preventive measures can increase the effectiveness of policies to contain the transmission of infectious diseases. The objective of this study was to draw a panorama of COVID-19 in São Paulo State and investigate the population's risk perception about the new coronavirus, attitudes regarding preventive measures and vaccine confidence through an open web survey. A total of 1111 individuals participated (71.6% females). Of those, 15% declared themselves unemployed, almost 6% lost their jobs during the pandemic, and more than 30% suffered a decrease in their monthly income. As for the COVID-19, almost 32% of the participants were tested sometime in 2020, and 11.3% reported a confirmed diagnosis for a level of hospitalisation of 5%. As for the risk perception, it was observed that situations with crowds of people were more frequently considered as of high risk and cause of greater concern with contamination of family members than with themselves. Almost 23% of the participants questioned whether they would take the vaccine when available and do not know if the COVID-19 vaccine will be safe for the population. Almost 70% disagreed with the statement that the actions of the country's health authorities are effective in combating COVID-19. Considering the Cramer's V coefficient, these variables were not associated with gender or educational level.

What is known about this topic

The success of policies to slow down the rapid transmission of infectious diseases relies, primarily, on an accurate perception of the public about risk factors.

In 2021, Brazil remained without an efficient plan to control the pandemic, which leaves the population increasingly vulnerable both to the disease and to ideological influences, fraught with misinformation, lack of leadership and perspectives.

What the paper adds

The proximity to the disease and its outcomes may have generated a higher consciousness of self‐care and empathy with close people, increasing the risk perception.

Despite the high confidence in the COVID‐19 vaccine, more than 25% have doubts whether they will get vaccinated or doubt vaccine safety.

INTRODUCTION

In March 2020, the World Health Organisation warned of the alarming levels of spread and severity of the new coronavirus SARS‐CoV‐2, characterising the global epidemic of COVID‐19, while signalling the serious worldwide consequences if there were no quick and strategic actions for its containment (Cucinotta & Vanelli, 2020; World Health Organization, 2020).

The first confirmed case of COVID‐19 in Brazil took place in the city of São Paulo in February 2020 (Rodriguez‐Morales et al., 2020). To date (17 June 2021), 1 year and 4 months after the first case, the epidemic in Brazil remains uncontrolled, with more than 17.6 million cases diagnosed in total, most of which are concentrated in São Paulo State (3.51 million). With regard to deaths due to COVID‐19, Brazil has accumulated the sorrowful number of more than a half million deaths in total and continues to record nearly a thousand deaths per day (Castro et al., 2021). Despite Brazil's global recognition as an exemplary country in immunisation policies (Brazilian Ministry of Health, 2003), the vaccination as a strategy to control the COVID‐19 pandemic unfortunately started late. It is kept slowly in the country, and only a small portion of the population has been immunised so far (approximately 28% of the population received the first dose of the vaccine, and only about 11% received both doses).

As it is known in the scientific literature and according to the accumulated practical experiences from past epidemics, the success of policies to slow down the rapid transmission of a highly infectious disease relies primarily on the public to have accurate perceptions of personal and social risk factors (Abbas, 2021). In the midst of a political, sanitary and economic crisis, Brazil remains without a practical and efficient plan to control the pandemic, which leaves the population increasingly vulnerable both to the disease and to ideological influences, fraught with misinformation, lack of leadership and perspectives (Gramacho & Turgeon, 2021). Furthermore, denial movements persist despite the country facing the worst phase of the pandemic with catastrophic consequences reaching the apex, while there is a lack of control and policies to deal with the current crisis. A sea of fake news invaded the country, contradicting science and encouraging anti‐vaccine attitudes, disregarding basic preventive measures such as the use of masks, and casting doubt on the disease data (Gramacho et al., 2021; Lopes, 2021).

Risk perceptions have been strongly associated with intentions to act to keep oneself healthy, being generally assessed through two dimensions: personal susceptibility and perceptions of severity (Weinstein, 1993). Recent studies in the context of the COVID‐19 pandemic have shown evidence of associations of both dimensions with the likelihood of vaccinating oneself and those under their care. Individuals believing they were more vulnerable to COVID‐19 and that the effects would be serious were more inclined to get vaccinated (Khubchandani et al., 2021; Schneider et al., 2021; Viswanath et al., 2021).

Vaccine hesitancy and confidence are among the main challenges to ensuring global access to immunisation (Wouters et al., 2021,), and studies on risk perception and vaccine confidence are vital for observing changes in community behaviour. In addition, developing a panoramic view of the disease's dispersion can aid in the identification of population specificities, which can help with the challenge of vaccination. In this sense, the objective of the present study was to draw a panorama of COVID‐19 in São Paulo State and investigate the population's risk perception about the new coronavirus, attitudes concerning preventive measures and vaccine confidence.

METHODS

Study design and settings

This cross‐sectional study is characterised as an open web survey, with data collection carried out by electronic questionnaires. An electronic version of the data collection instrument was developed on the Research electronic data capture (REDCap) platform (Harris et al., 2009), which stores the database on a server at the Ribeirão Preto Medical School—University of São Paulo (USP). REDCap is a secure application for creating and managing online surveys and databases (http://project‐redcap.org).

The online questionnaire was administered to a sample of 15 individuals who are part of the study's target population as a pre‐test step. Both the understanding of the questions and the use of the online platform and its functionalities were evaluated. It should be emphasised that these individuals were not included in the study's final sample.

Invitations for participation in the study were posted on social networks (Facebook, Twitter, Whatsapp and others). The disclosure text posted together with the link to access the electronic questionnaire described the objectives of the survey and the target audience and provided information on the confidentiality of the responses. In addition, the participants were requested to share the link with their contacts after completing the survey in order to increase the participation levels and thus obtain a broader sample. Participation in this survey was voluntary, and no compensation was provided.

The first page of the electronic questionnaire contained the Informed Consent Form (ICF), and the participants could decide whether or not to participate in the study after reading it. When accepting participation, the first questions of the questionnaire referred to the inclusion and exclusion criteria. Age of 18 years or older, place of residence (some municipality in the State of São Paulo), and proficiency in Portuguese language were used as inclusion criteria. If any of these criteria were not met, the survey ended with a message of thanks. Those participants who did not answer all the questions in the questionnaire were not included in the final sample. Data collection was carried out from September 2020 to December 2020.

Variables

The online questionnaire was developed by the responsible researchers, including variables relevant to draw the panorama of COVID‐19 in the territory based on the literature review. Sociodemographic and behavioural variables were collected to characterise the sample. To investigate health conditions, variables such as self‐perception of health, chronic disease and comorbidities for COVID‐19, smoking and practice of physical exercises were considered. A panorama of SARS‐CoV2 infection and COVID‐19 was assessed through questions on previous test, type of test, diagnosis and type of infection (considering the symptoms of the disease), isolation level and presence of any sequel after the cure. To investigate the perception of the risk of contracting COVID‐19, some common everyday situations were presented to the participants, and they were asked to classify each of them, considering transmission risk as low, moderately low, moderate, moderately high or high. The last block of the questionnaire referred to attitudes regarding preventive measures, opinions about the efficacy of treatment and vaccination, and concerns about contracting COVID‐19.

Statistical analysis

Data from REDCap were exported to R software version 3.6.2 for descriptive analysis. The strength of the association between dependent and independent variables of interest was measured by using the Cramér's V coefficient (Everitt, 1992). Effect sizes were thus interpreted by using Rea and Parker (2014), in which references below 0.10 mean negligible association; between 0.10 and below 0.20 mean weak association; between 0.20 and below 0.40 mean moderate; and between 0.40 and 0.60 mean as a relatively strong association. p‐values for these associations were computed for a Monte Carlo test with 20.000 replicates, as proposed by Hope (1968). The “chisq.test” function of the R software with the argument “simulate.p.value = TRUE” was used to obtain these p‐values at a level of significance of 0.05.

Ethical issues

The study was approved by the Research Ethics Committee of Clinical Hospital of Ribeirão Preto Medical School (CAAE: 36947720.0.0000.5440). The informed consent form was presented on the first page of the online questionnaire and the participants were also informed that they could end their participation at any time. In addition, all the questions included the option “I prefer not to answer” between the response categories, i.e., participants were not required to answer any questions with which they were uncomfortable.

The research was carried out according to guidelines set by the Checklist for Reporting Results of Internet E‐Surveys (CHERRIES) (Eysenbach, 2004), and the applied methodology followed the guidelines for research procedures with any stage in a virtual environment made available by the National Research Ethics Commission (CONEP) in Circular Letter No. 1/2021‐CONEP/SECNS/MS.

RESULTS

The questionnaire was accessed by 1759 individuals during the data collection period. Of these, 227 individuals dropped out of the questionnaire before the end and were excluded from the total sample. Of the 1532 individuals who reached the end of the online questionnaire, 1111 were residents in the State of São Paulo (study inclusion criterion) and were part of the study sample. Of these, 796 were female (71.6%) and 313 were male (28.2%), and two individuals did not want to answer about their sex in the birth register (0.2%). Table 1 shows the characterisation of the sample according to sociodemographic variables.

TABLE 1

Distribution of the participants (n = 1111) according to sociodemographic variables. State of São Paulo, Brazil, 2020

Variables	Response categories	n (%)	Females	Males
			n (%)	n (%)
Age groups a	Up to 30 years	263 (24.1%)	196 (25.0%)	66 (21.5%)
	31–40 years	345 (31.6%)	261 (33.3%)	84 (27.4%)
	41–50 years	232 (21.2%)	146 (18.6%)	85 (27.7%)
	51–60 years	153 (14.0%)	108 (13.8%)	45 (14.7%)
	Over 60 years	100 (9.1%)	73 (9.3%)	27 (8.8%)
Marital status	Married/Live together	618 (55.6%)	430 (54%)	188 (60.1%)
	Single	397 (35.7%)	295 (37.1%)	101 (32.3%)
	Divorced	78 (7%)	58 (7.3%)	20 (6.4%)
	Widowed	15 (1.4%)	12 (1.5%)	3 (1%)
	I do not want to answer	3 (0.3%)	1 (0.1%)	1 (0.3%)
Number of people in the same residence	Live alone	137 (12.3%)	99 (8.9%)	38 (3.4%)
	Two people	327 (29.4%)	244 (30.7%)	82 (26.2%)
	Three people	313 (28.2%)	222 (27.9%)	91 (29.1%)
	Four people	229 (20.6%)	162 (20.4%)	66 (21.1%)
	Five people	79 (7.1%)	50 (6.3%)	29 (9.3%)
	Six to nine people	24 (2.2%)	17 (%)	7 (1.6%)
	Ten people or more	2 (0.2%)	2 (0.3%)	0 (0%)
Piped water in the residence	Yes	1108 (99.7%)	793 (99.6%)	313 (100%)
	No	3 (0.3%)	3 (0.4%)	0 (0%)
Live on an asphalt street	Yes	1106 (99.5%)	793 (99.6%)	311 (99.4%)
	No	5 (0.5%)	3 (0.4%)	2 (0.6%)
Educational level	Elementary school Incomplete	15 (1.3%)	11 (1.4%)	4 (1.3%)
	Elementary school complete	23 (2.1%)	16 (2.1%)	7 (2.2%)
	High school incomplete	24 (2.2%)	20 (2.5%)	4 (1.3%)
	High school complete	129 (11.6%)	103 (12.9%)	26 (8.3%)
	Higher education incomplete	161 (14.5%)	106 (13.3%)	55 (17.6%)
	Higher education complete	759 (68.3%)	540 (67.8%)	217 (69.3%)
Current working condition	Unemployed/no gainful activity	170 (15.3%)	138 (17.3%)	32 (10.2%)
	Registered or self‐employed work	777 (69.9%)	527 (66.2%)	248 (79.2%)
	Pensioner or retired	88 (7.9%)	73 (9.2%)	15 (4.8%)
	Informal work	58 (5.2%)	43 (5.4%)	15 (4.8%)
	I do not want to answer	18 (1.6%)	15 (1.9%)	3 (1%)
Frontline professional in the pandemic	Yes	375 (33.8%)	272 (34.2%)	102 (32.6%)
	No	723 (65.1%)	512 (64.3%)	210 (67.1%)
	I do not want to answer	13 (1.2%)	12 (1.5%)	1 (0.3%)
Type of work between March 2020 and currently	Working from home	495 (44.6%)	350 (44%)	144 (46%)
	Continued to work normally, but on a reduced workload	120 (10.8%)	86 (10.8%)	34 (10.9%)
	Keep normal activities at workplace as usual	218 (19.6%)	133 (16.7%)	84 (26.8%)
	Did not work during the pandemic	267 (24%)	216 (27.1%)	51 (16.3%)
	I do not want to answer	11 (1%)	11 (1.4%)	0 (0%)
Job loss during the pandemic	Yes	66 (5.9%)	49 (6.2%)	17 (5.4%)
	No	1034 (93.1%)	737 (92.6%)	295 (94.2%)
	I do not want to answer	11 (1%)	10 (1.3%)	1 (0.3%)
Percentage of income that was reduced during the pandemic	I did not have my income reduced	724 (65.2%)	516 (64.8%)	208 (66.5%)
	My income has been reduced by less than 30%	143 (12.9%)	96 (12.1%)	46 (14.7%)
	My income has been reduced by between 31% and 50%	92 (8.3%)	65 (8.2%)	26 (8.3%)
	My income has been reduced by more than 50%	69 (6.2%)	53 (6.7%)	16 (5.1%)
	I lost 100% of my monthly income during the Pandemic	55 (5%)	41 (5.2%)	14 (4.5%)
	I do not want to answer	28 (2.5%)	25 (3.1%)	3 (1%)
Current monthly income b	Has no monthly income	23 (2.1%)	20 (2.5%)	3 (1%)
	About 1 minimum wage (up to 1000 reais)	53 (4.8%)	41 (5.2%)	11 (3.5%)
	Between 2 and 5 minimum wages (between 2000 and 5000 reais)	440 (39.6%)	344 (43.2%)	95 (30.4%)
	Between 5 and 10 minimum wages (between 5000 and 10000 reais)	302 (27.2%)	199 (25%)	103 (32.9%)
	More than 10 minimum wages (more than 10000 reais)	230 (20.7%)	136 (17.1%)	94 (30%)
	I do not want to answer	63 (5.7%)	56 (7%)	7 (2.2%)

This variable contains missing data.

1 US dollar corresponds to approximately 4.7 Brazilian Reais.

Despite the higher frequency of female participants, it was observed that the participation of individuals of all age groups was similar. Most households consisted of up to four people, with 12.3% of the participants reporting living alone. The sample was primarily composed of people with complete higher education, and almost 70% of the participants declared themselves registered workers or self‐employed. Considering the type of work performed since March 2020, almost 45% worked from home, followed by nearly 20% who maintained normal activities at the workplace as usual. However, more than 15% of the participants declared to be unemployed or not be engaged in any gainful activity, almost 6% declared to have lost their jobs during the pandemic, and more than 30% declared that they suffered a decrease in their monthly income during the period.

Questions on health conditions and health behaviours are shown in Table 2. Most participants classified their health as good. However, almost 40% considered themselves part of the risk group for COVID‐90, which is a similar frequency of participants who declare having a chronic disease. The frequency of self‐reported obesity in the sample was approximately 18% (similar among males and females) and almost 45% declared they did not practice physical exercise. The prevalence of smoking in the total sample was approximately 9%, being higher among males (11.5%).

TABLE 2

Health conditions and behaviours according to sex. State of São Paulo, Brazil, 2020

Variables	Response categories	n (%)	Females	Males
			n (%)	n (%)
	Good	903 (81.3%)	647 (81.3%)	254 (81.2%)
Self‐perception of health Consider yourself part of the risk group in relation to COVID‐19	Regular	192 (17.3%)	137 (17.2%)	55 (17.6%)
	Poor	16 (1.4%)	12 (1.5%)	4 (1.3%)
	Yes	422 (38%)	299 (37.6%)	123 (39.3%)
	No	647 (58.2%)	463 (58.2%)	182 (58.1%)
	I do not know	42 (3.8%)	34 (4.3%)	8 (2.6%)
Chronic disease	Yes	418 (37.6%)	282 (35.4%)	135 (43.1%)
	No	693 (62.4%)	514 (64.6%)	178 (56.9%)
Obesity	Yes	200 (18.0%)	140 (17.6%)	59 (18.8%)
	No	911 (82.0%)	656 (82.4%)	254 (81.2%)
Physical exercises	Do not practice	497 (44.7%)	363 (45.6%)	133 (42.5%)
	Practice 1–2 times a week	308 (27.7%)	211 (26.5%)	97 (31%)
	Practice 3–5 times a week	258 (23.2%)	189 (23.7%)	68 (21.7%)
	Practice 6–7 times a week	45 (4.1%)	30 (3.8%)	15 (4.8%)
	Do not want to answer	3 (0.3%)	3 (0.4%)	0 (0%)
Smoker	Yes	97 (8.7%)	61 (7.7%)	36 (11.5%)
	No	1013 (91.2%)	734 (92.2%)	277 (88.5%)
	Do not want to answer	1 (0.1%)	1 (0.1%)	0 (0%)
Number of cigarrettes (among smokers)	Up to 5 cigarrettes	24 (24.7%)	15 (24.6%)	9 (25%)
	Between 6–10 cigarrettes	26 (26.8%)	17 (27.9%)	9 (25%)
	Between 11–20 cigarrettes	32 (33%)	17 (27.9%)	15 (41.7%)
	More than 20 cigarrettes	12 (12.4%)	10 (16.4%)	2 (5.6%)
	Do not want to answer	3 (3.1%)	2 (3.3%)	1 (2.8%)
Passive smoker a	Yes	184 (18.2%)	144 (19.6%)	38 (13.7%)
	No	829 (81.8%)	590 (80.4%)	239 (86.3%)
Ex smoker a	Yes	153 (15.1%)	102 (13.9%)	51 (18.4%)
	No	860 (84.9%)	632 (86.1%)	226 (81.6%)
Time after quitting smoking (among ex smokers)	Less than 1 year	23 (15%)	16 (15.7%)	7 (13.7%)
	Between 1 and 5 years	28 (18.3%)	16 (15.7%)	12 (23.5%)
	Between 6 and 10 years	20 (13.1%)	15 (14.7%)	5 (9.8%)
	More than 10 years	82 (53.6%)	55 (53.9%)	27 (52.9%)

This question contains missing data.

Table 3 shows a panorama of SARS‐CoV2 infection and COVID‐19 disease in the São Paulo State population. In the investigation of the SARS‐Cov2 infection and COVID‐19 diagnosis, only almost 32% of the participants were tested sometime in 2020. A total of 125 (11.3%) participants reported a confirmed diagnosis, most frequently in July, August, September 2020 or after. Of the confirmed cases, most participants classified the type of infection as mild symptomatic and reported some sequel, whereas almost 5% declared the need for hospitalisation. As for the isolation level after the confirmed diagnosis, most participants continued to live only with people in the same residence for 14 days, whereas approximately 5% declared that they did not stay isolated.

TABLE 3

Panorama of SARS‐Cov2 infection and COVID‐19 according to sex. São Paulo state, Brazil, 2020

Variables	Response categories	n (%)	Females	Males
			n (%)	n (%)
Did any test for SARS‐CoV2 /COVID‐19	Yes	353 (31.8%)	253 (31.8%)	99 (31.6%)
	No	758 (68.2%)	543 (68.2%)	214 (68.4%)
Type of test (among those who were tested) a	Serology (blood test)	171 (15.4%)	123 (15.5%)	47 (15.0%)
	Serology (finger stick – fast test)	94 (8.5%)	69 (8.7%)	25 (8.0%)
	PCR	187 (16.8%)	137 (17.2%)	50 (16.0%)
Confirmed diagnosis of COVID‐19 / positive SARCOV2 infection test	Yes	125 (11.3%)	93 (11.7%)	32 (10.2%)
	No	981 (88.3%)	701 (88.1%)	278 (88.8%)
	Do not want to answer	5 (0.5%)	2 (0.3%)	3 (1.0%)
Type of infection (among those with confirmed diagnosis)	Asymptomatic	10 (8%)	6 (6.5%)	4 (12.5%)
	Mild symptomatic	67 (53.6%)	53 (57%)	14 (43.8%)
	Strong symptomatic	42 (33.6%)	29 (31.2%)	13 (40.6%)
	Severe symptomatic ‐ hospitalised	6 (4.8%)	5 (5.4%)	1 (3.1%)
Isolation level after diagnosis (among those with confirmed diagnosis)	100% isolated (alone) for 14 days	33 (26.4%)	26 (28%)	7 (21.9%)
	100% isolated (alone) for less than 14 days	9 (7.2%)	6 (6.5%)	3 (9.4%)
	Continued to live only with people in the residence for 14 days	70 (56%)	55 (59.1%)	15 (46.9%)
	Continued to live only with people in the residence for less than 14 days	7 (5.6%)	3 (3.2%)	4 (12.5%)
	Not isolated	6 (4.8%)	3 (3.2%)	3 (9.4%)
Sequel of COVID‐19	Yes	68 (54.4%)	50 (53.8%)	18 (56.2%)
	No	57 (45.6%)	43 (46.2%)	14 (43.8%)
Number of people you know who had a confirmed diagnosis of covid‐19	None	211 (19%)	147 (18.5%)	64 (20.4%)
	Between 1 and 5 people	600 (54%)	425 (53.4%)	174 (55.6%)
	Between 6 and 10 people	210 (18.9%)	156 (19.6%)	54 (17.3%)
	Between 11 and 20 people	54 (4.9%)	45 (5.7%)	8 (2.6%)
	More than 20 people	30 (2.7%)	19 (2.4%)	11 (3.5%)
	I do not know	6 (0.5%)	4 (0.5%)	2 (0.6%)
Lost a family member or close to COVID‐19?	Yes	164 (14.8%)	109 (13.7%)	55 (17.6%)
	No	945 (85.1%)	686 (86.2%)	257 (82.1%)
	Do not want to answer	2 (0.2%)	1 (0.1%)	1 (0.3%)
The actions of health authorities in my country are being effective in combating COVID‐19	I agree	112 (10.1%)	74 (9.3%)	38 (12.1%)
	Maybe/Not sure	249 (22.4%)	179 (22.5%)	70 (22.4%)
	I disagree	750 (67.5%)	543 (68.2%)	205 (65.5%)

It is considered that a participant may have taken more than one test in the period.

Figure 1 presents the participants' risk perception in common everyday situations, considering the transmission of the virus. For each situation, the participants were asked to classify the transmission risk of contracting the SARS‐CoV2 as low, moderately low, moderate, moderately high or high. Situations in which there are crowds of people, such as attending bars, cinemas, barbecues, events, concerts, religious meetings and buffet parties, were considered those with the highest frequencies of “high risk” responses (approximately 70% or more). Sending children to school, camps or daycare centres was also among the activities considered of high risk by 74% of the participants. The situations less frequently perceived as of “high risk” were fueling the car at a gas station, outdoor walking, running or cycling and receiving a correspondence or a delivery (frequency of classification as “high risk” less than 6%). However, participants were quite divided regarding rating these situations as low risk, moderate‐low risk, moderate risk or moderate‐high risk.

FIGURE 1

Participants' risk perception in common everyday situations, considering the transmission of the SARS‐CoV‐2 virus. São Paulo State, Brazil, 2020.

Table 4 presents the participants' opinions about the efficacy of treatment and vaccination, and concerns about contracting COVID‐19 according to sex and educational level. With regard to the efficacy of the treatment available for COVID‐19, more than 60% of the participants considered it to have a medium efficacy, and more than 70% considered that the chance of a big outbreak of COVID‐19 in the city is high where they live. As for the concern with the possibility of contracting COVID‐19, it was observed a higher concern with contamination of family members than of themselves. Almost 23% of the participants doubted whether they would take the vaccine if available and do not know if the COVID‐19 vaccine will be safe for the population. These variables were not associated with sex or educational level. The study of associations according to other independent variables were presented as supplementary material.

TABLE 4

Opinions about efficacy of treatment, vaccination and concerns about contracting COVID‐19 according to sex and educational level. São Paulo state, Brazil, 2020

			Sex			Educational level
Variables	Response categories	n (%)	Females n (%)	Males n (%)	Cramer's V a	Until High School n (%)	Higher education incomplete n (%)	Higher education complete n (%)	Cramer's V a
Opinion about efficacy of treatments for COVID‐19	High	118 (10.6%)	82 (10.3%)	36 (11.5%)	0.07	21 (11%)	20 (12.4%)	77 (10.1%)	0.02
	Medium	711 (64%)	526 (66.1%)	184 (58.8%)		121 (63.4%)	103 (64%)	487 (64.2%)
	Low	282 (25.4%)	188 (23.6%)	93 (29.7%)		49 (25.7%)	38 (23.6%)	195 (25.7%)
Chance of a major outbreak of COVID‐19 in the city where live	High	799 (71.9%)	599 (75.3%)	198 (63.3%)	0.16	149 (78%)	108 (67.1%)	542 (71.4%)	0.06
	Medium	268 (24.1%)	179 (22.5%)	89 (28.4%)		37 (19.4%)	48 (29.8%)	183 (24.1%)
	Low	44 (4%)	18 (2.3%)	26 (8.3%)		5 (2.6%)	5 (3.1%)	34 (4.5%)
Concern with the possibility of contracting COVID‐19 a	Much concern	575 (58.6%)	431 (61.5%)	144 (51.8%)	0.15	96 (64.4%)	81 (59.1%)	398 (57.3%)	0.04
	Moderate concern	370 (37.7%)	256 (36.5%)	113 (40.6%)		49 (32.9%)	49 (35.8%)	272 (39.1%)
	No concern	36 (3.7%)	14 (2%)	21 (7.6%)		4 (2.7%)	7 (5.1%)	25 (3.6%)
Concern about family members contracting COVID‐19	Much concern	950 (85.5%)	697 (87.6%)	252 (80.5%)	0.14	168 (88%)	141 (87.6%)	641 (84.5%)	0.03
	Moderate concern	149 (13.4%)	97 (12.2%)	51 (16.3%)		22 (11.5%)	18 (11.2%)	109 (14.4%)
	No concern	12 (1.1%)	2 (0.3%)	10 (3.2%)		1 (0.5%)	2 (1.2%)	9 (1.2%)
Took all vaccines in the adult immunisation schedule	Yes	915 (82.4%)	661 (83%)	252 (80.5%)	0.03	167 (87.4%)	125 (77.6%)	623 (82.1%)	0.07
	No	116 (10.4%)	81 (10.2%)	35 (11.2%)		14 (7.3%)	27 (16.8%)	75 (9.9%)
	Do not know	80 (7.2%)	54 (6.8%)	26 (8.3%)		10 (5.2%)	9 (5.6%)	61 (8%)
Would take the COVID‐19 vaccine if available
	No	37 (3.3%)	26 (3.3%)	11 (3.5%)	0.04	7 (3.7%)	7 (4.3%)	23 (3%)	0.07
	I doubt if I would take	252 (22.7%)	189 (23.7%)	62 (19.8%)		57 (29.8%)	27 (16.8%)	168 (22.1%)
I would take for sure	822 (74%)	581 (73%)	240 (76.7%)	127 (66.5%)		127 (78.9%)	568 (74.8%)
COVID‐19 vaccine will be safe for the population	Yes	808 (72.7%)	564 (70.9%)	243 (77.6%)	0.07	123 (64.4%)	115 (71.4%)	570 (75.1%)	0.08
	No	53 (4.8%)	38 (4.8%)	15 (4.8%)		16 (8.4%)	12 (7.5%)	25 (3.3%)
	Do not know	250 (22.5%)	194 (24.4%)	55 (17.6%)		52 (27.2%)	34 (21.1%)	164 (21.6%)

Results of Cramer's V with asterisks correspond to p < 0.05.

DISCUSSION

In the present study, we tried to explore some demographic and behavioural variables regarding the COVID‐19 pandemic in population if the State of São Paulo, which can help to explain the current situation in Brazil and draw panorama of the disease in this state (Brazilian Ministry of Health, 2021; SAEDE, 2021).

In addition to the sanitary crisis, the country is currently experiencing severe consequences of the pandemic, namely: an increase in unemployment to rampant levels and, consequently, an increase in poverty and social vulnerability (Freitas et al., 2020; Neves et al., 2021; Tavares & Betti, 2021). In our sample, more than 15% of the participants declared to be unemployed or not engaged in any gainful activity, almost 6% declared to have lost their jobs during the pandemic, and more than 30% declared that they suffered a decrease in their monthly income during the period. Considering data from the Brazilian Institute of Geography and Statistics (IBGE, 2020), in the third trimester of 2020, there were 13.7 million unemployed, and in the first 5 months of 2020, there was an increase of 12.4% in the requests for unemployment insurance compared to 2019. According to studies by Costa (2020) and Souza et al. (2020), the economic crisis in the context of the pandemic affected the population who lives in informality and resides in deprived areas of Brazil with greater intensity. Despite that, our results show that even populations with a higher level of education (almost 70% of our sample completed higher education) and better standards of living were somehow economically affected by the pandemic.

It was observed in our study that almost 32% of the participants reported being tested for COVID‐19 sometime in 2020. This percentage is higher than the national proportion of testing, that is, approximately 14.3%, according to the National Household Sample Survey—PNAD COVID19 (IBGE, 2020; Kameda et al., 2021). Since the pronouncement of the World Health Organisation declaring the COVID‐19 pandemic in March 2020, both health authorities and scientists around the world have pointed to the mass testing of the population as a crucial strategy for the control of the pandemic, especially in suspected cases when general testing is not feasible (Lopes‐Junior et al., 2020). However, massively testing the population can be difficult to implement in practice, especially in low‐middle‐income countries (Kameda et al., 2021). More than a year later and at a time when epidemiologists are warning of the risk of a new collapse in winter (June to September in Brazil), massive testing against COVID‐19 in Brazil remains low and disorganised. Brazil performs 7857 tests per 100,000 inhabitants, according to the Ministry of Health's epidemiological bulletin, ranking behind countries such as India and its neighbours in South America (Brazilian Ministry of Health, 2021).

The proportion of diagnosed cases of COVID‐19 in the sample was 11.3%, lower than data from PNAD COVID‐19, which indicated that approximately 23% of the tested individuals received a positive diagnosis until November 2020 (IBGE, 2021). In addition, interesting data on the disease sequels and persistent symptoms were identified in our study as nearly 55% of the participants who had COVID‐19 reported them. Several studies have been conducted to investigate the consequences of COVID‐19 to the human body. However, information on symptoms persisting after recovery is lacking (Peramo‐Álvarez et al., 2021). A study performed by Carfì et al. (2020) found that in Italian patients who had recovered from acute COVID‐19, 87.4% reported persistence of at least one symptom, particularly fatigue and dyspnea. In Brazil, a few studies on this issue have been published, which does not allow any the comparison of our findings with those from clinical and controlled studies. A systematic review conducted by Nasserie et al. (2021) discussed that despite the need for studies with longer follow‐up periods to better understand the multifactorial nature of COVID‐19 manifestations and reliably quantify the risks, there is already consistent evidence that the symptoms of the disease commonly persist beyond the acute phase of infection, with implications for health‐associated functioning and quality of life.

With regard to the risk perception, the participants reported that high‐risk situations are those in which there are crowds of people, such as attending bars, cinemas, barbecues, events, concerts, religious meetings and buffet parties, in addition to sending children to school, camps or daycare centres. Thus, we can observe a coherent perception of what is known about the transmissibility of the SARSCov‐2 virus in the sample, which evidenced by the risk classifications for the situations described in Figure 1. In a study conducted by Dryhurst et al. (2020), who presents comparative evidence of how people perceive the risk of COVID‐19 around the world, the authors point out that risk perception can be modulated by people's experiences, values and trust in institutions, which are higher among those with direct personal experience of the virus and in those who hold naïve worldviews. This draws attention that in our sample, 54% of the respondents reported knowing between one and five people who had a confirmed diagnosis of COVID‐19, and more than 26% know six people or more, whereas almost 15% reported that they lost a family member or someone close due to the disease. In this sense, we consider that the proximity to the disease and its outcomes may have generated a higher consciousness of self‐care and empathy with close people due to the concern with the infection and its possible harmful consequences, increasing the risk perception. In addition, almost 70% of the participants disagreed with the statement that the health policies of the country's authorities were effective in combating COVID‐19, and more than 70% rated the chance of a major outbreak of COVID‐19 in the city where they live as “high”.

Despite the high confidence in the COVID‐19 vaccine and the high proportion of participants who responded that they would definitely take the vaccine, there is concern that more than 25% think otherwise or have doubts about whether to take the vaccine or whether the vaccine will be safe. The vaccination for COVID‐19 started in Brazil in February 2020 and has been progressing at a slow pace. So far, no strong vaccination campaign has been conducted to inform, reassure and raise awareness about the importance of mass vaccination in Brazil. In addition to the low proportion of immunised individuals in the country (about 12%), Brazil is currently experiencing a delay in the search for the second dose. In May 2020, according to official data (Brazilian Ministry of Health, 2021), about 1 million people did not show up to take the second dose of the vaccine, which is extremely worrying data given the current situation in the country during the pandemic.

The present study is not free from limitations. Firstly, the sectional nature of the study does not allow cause‐and‐effect interpretations. In addition, the methodology of data collection inherent to the open web survey design can address participants who do not necessarily represent the studied population, thus hindering generalisations, as it is difficult to obtain an accurate sampling frame or an accurate estimate of the population characteristics. Another point is that self‐reported bias can also affect the results. Individuals can easily hide, omit or lie about any information they report in the online survey, so there is no guarantee of accuracy. Furthermore, according to the most recent PNAD, 81.7% of the residents in São Paulo State have access to the internet, which is also a limitation of web surveys, which may not reach some layers of the population.

AUTHORS CONTRIBUTION

All authors contributed to the study conception, design and data collection. The analyses were performed by Edson Zangiacomi Martinez. The first draft was written by Miriane Lucindo Zucoloto and all authors revised, contributed and approved the final version of the manuscript.

CONFLICT OF INTEREST

All of the authors declare that there are no conflicts of interest in connection with this paper.

ETHICAL ISSUES

The study was approved by the Research Ethics Committee of Clinical Hospital of Ribeirão Preto Medical School (CAAE: 36947720.0.0000.5440). The research was carried out according to the checklist for verification of the electronic research guidelines carried out on the Internet (CHERRIES), and the applied methodology followed the guidelines for research procedures at any stage in a virtual environment made available by the National Research Ethics Commission (CONEP) in Circular Letter No. 1/2021‐CONEP/SECNS/MS.

Supinfo S1

Click here for additional data file.