Face Mask Use During the COVID-19 Outbreak: How Did Educated Brazilians Behave?

Abstract

Purpose:

To describe Brazilians’ behavior regarding face mask use and health literacy during the COVID-19 pandemic before and after the Ministry of Health of Brazil formal recommendation.

Design:

Cross-sectional surveys using a web-based questionnaire. Participants were recruited via snowball techniques.

Setting:

São Paulo state, the urban epicenter of the COVID-19 pandemic in Brazil at the time of the study.

Participants:

2.203 clicks to the survey link and 1.223 surveys completed (55.5% response rate). However, only 1118 surveys were considered after the exclusion criteria (>18 years-old and consent).

Measures:

Demographics, educational status, COVID-19-related symptoms (headache, cough, sore throat, rhinorrhea, fever, asthenia, diarrhea, dyspnea, nausea, vomiting, vertigo, anosmia, and ageusia), and face mask use.

Analysis:

Self-reports of COVID-19 symptoms were categorized as dichotomous variables (Cohen’s h = 0.94). Pearson Chi-square test evaluated differences between T1 and T2 and logistic multiple regression analyzed odds-ratio for the presence of symptoms and independent variables.

Results:

Face mask use increased from 43.60% in T1 to 90.52% in T2 (P < .0001) as the pandemic went on. Health literacy also changed within 2 weeks and people started to assume everybody should use face masks (62.93% in T1 vs 94.12% in T2; P < .0001; ES = 0.29) during outside activities (43.60% in T1 vs 90.52% in T2; P < .0001; ES = 0.39). Self-reports of face mask use were associated with fewer self-reports of COVID-19 symptoms (OR = 0.65, P = .01, 95% CI 0.48; 0.88).

Conclusion:

Face mask use was already high among educated Brazilians before the formal recommendation by the authorities. This may have contributed to fewer self-reports of COVID-19-related symptoms.

Keywords

masks COVID-19 COVID-19 pandemic SARS-CoV-2 infection pandemics health behavior

Purpose

Face masks, whether surgical or not, are usually recommended for preventing respiratory tract infections. Their use has never been so debated as now during the COVID-19 outbreak, both in journalism and scientific fields.^1
-3 Unlike the population of Eastern countries,⁴ Brazilians are not used to protecting themselves with face masks for respiratory diseases. The Brazilian population started to use face masks as soon as the first cases of severe pneumonia were reported in December 2019⁵ and was eager to engage in self-protection against an unknown virus. By then, face mask use had became mandatory in Eastern countries and others, such as the Czech Republic.⁶ However, mandatory use occurred much later in some Brazilian states.

By the end of March 2020, both the Ministry of Health of Brazil and the World Health Organization (WHO)^7,8 decided to support face mask use for all people and also recommended using homemade masks.⁷ Both were worried about avoiding a shortage of surgical masks for health services and professionals that were highly exposed to SARS-CoV-2 infected patients, which encouraged some authorities to postpone face mask use recommendation^7,8 as did Brazilian authorities.⁹

This research describes how Brazilians behaved regarding face mask use at 2 important moments at the beginning of the outbreak: 1) 15 days before, and 2) the day after the Ministry of Health of Brazil formally recommended countrywide face mask use. We also investigated the COVID-19-related symptoms the respondents reported.

Methods

This is a cross-sectional study developed at Universidade Federal de São Paulo (UNIFESP), Brazil. This protocol was approved by the Ethical Committee on Research of the UNIFESP (CAAE 30890420.7.0000.5505) and was developed according to the principles expressed in the Declaration of Helsinki. All participants provided their consent.

Sample

We designed 2 questionnaires at the Microsoft® Form platform. The first one was administered 3 weeks after Brazilian’s first notification of community SARS-CoV-2 infection and 15 days before the Ministry of Health of Brazil announced “masks for all” (T1) and the second questionnaire was released 2 weeks later, which was also a day after the Ministry of Health of Brazil announced “masks for all” (T2), in April 2020 (Figure 1). Those individuals who participated only at T2 also responded to the same questions presented in the first questionnaire before answering the new questions added to the second questionnaire. Questions and results of both questionnaires are shown in Table 1.

Figure 1.

Enrolment flowchart. Abbreviations: N, sample; G1, group 1; G2, group 2.

Table 1.

Demographical Data and Responses of the 1118 Participants in Times T1 and T2 During the COVID-19 Outbreak.

	T1	(n = 812)	T2	(n = 306)	P-value		ES
Age	37.68	(13.67)	41.72	(13.59)	< .001	^a	0.26	^c
Gender
Male	245	(30.17%)	64	(20.92%)	.22	^b	0.11	^d
Female	557	(68.60%)	242	(79.08%)
Not informed	10	(1.23%)	0	(0.00%)
Education
Illiterate	0	(0.00%)	0	(0.00%)	.114	^b	0.09	^d
Elementary school	7	(0.86%)	3	(0.98%)
High school	82	(10.10%)	20	(6.54%)
Undergraduate	383	(47.17%)	125	(40.85%)
Graduate	340	(41.87%)	158	(51.63%)
Health care worker/student	418	(51.48%)	148	(48.37%)	.52	^b	0.29	^d
Last 2 weeks social distancing	683	(84.11%)	269	(87.91%)	.119	^b	0.04	^d
Contact with suspected of confirmed COVID-19 cases
Yes	134	(16.50%)	56	(18.30%)	.515	^b	0.02	^d
No	444	(54.68%)	159	(51.96%)
I don’t know	234	(28.82%)	91	(29.74%)
Last month symptoms related to COVID-19	239	(29.43%)	66	(21.57%)	.01	^b	0.07	^d
Face mask use during outside activities	354	(43.60%)	277	(90.52%)	< .0001	^b	0.39	^d
When did you start using masks?
I still don’t use it	439	(54.06%)	29	(9.48%)	< .0001	^b	0.42	^d
A week ago	140	(17.24%)	61	(19.93%)
2 weeks ago	152	(18.72%)	93	(30.39%)
A month ago	71	(8.74%)	118	(38.56%)
In January/February 2020	10	(1.23%)	5	(1.63%)
In your opinion, face masks should be worn by:
Asymptomatic people	2	(0.25%)	0	(0.00%)	< .0001	^b	0.29	^d
Symptomatic people	299	(36.82%)	18	(5.88%)
Everybody	511	(62.93%)	288	(94.12%)
A face mask reduces the odds of a sick person to spread the virus	799	(98.40%)	305	(99.67%)	.201	^b	0.05	^d
A face mask reduces the odds of a healthy person to be infected	638	(78.57%)	285	(93.14%)	< .0001	^b	0.17	^d
A face mask does not protect against the virus	187	(23.03%)	59	(19.28%)	.255	^b	0.04	^d
Only health care workers/students must use the face mask	80	(9.85%)	8	(2.61%)	< .01	^b	0.12	^d
Anyone who needs to go outside should have been using a face mask	559	(68.84%)	290	(94.77%)	< .0001	^b	0.26	^d
A homemade face mask is not so efficient and, then, there is no need for using it	149	(18.35%)	17	(5.56%)	< .0001	^b	0.16	^d

Abbreviation: ES, effect-size.

^aMann-Whitney test.

^b Pearson’s chi-square.

^c Cohen’s d.

^d Pearson’s contingency coefficient.

Altogether, our sample was composed of subjects reached through the researcher’s personal social media (WhatsApp®, Facebook®, Instagram®). Those people were, in turn, encouraged to share the questionnaires with their personal friends and colleagues, constituting the snowball technique.

We excluded individuals who were younger than 18 years old or who did not fill the birthdate section properly.

Outcome and Analysis

Our primary outcome was self-report of ≥ 1 symptom of COVID-19 described on the survey, i.e. headache, cough, sore throat, rhinorrhea, fever, asthenia, diarrhea, dyspnea, nausea, vomiting, vertigo, anosmia, and/or ageusia.

Data were typed and organized in Microsoft Excel for Mac®, version 16.35, and analyzed in IBM® SPSS statistics, version 26. We used absolute and relative frequencies (nominal variables) and mean and standard-deviation (numeric variables) to describe our results. We calculated the sample power according to Cohen’s h method and used Pearson Chi-Square and Mann-Whitney tests to compare our results, when appropriate. We calculated the effect sizes through Cohen’s d (numeric variables) and contingency coefficient (categorical variables). Multiple logistic regression was performed to analyze self-reported COVID-19 symptoms as the dependent variable. In all tests, α = 0.05 and a 95% confidence interval were considered.

Results

By the end of data collection, we had 1223 participants. The first moment (T1) resulted in 840 responses (loss rate of 51.6% based on 1738 clicked survey links, ie, clicks) and the second moment (T2) resulted in 383 responses (loss rate of 17.6% based on 465 clicks). After applying the exclusion criteria, we analyzed 1187 responses (Figure 1).

Most of the participants were young adults (38.79 ± 13.76 years-old), women (71.47%), highly educated (45.44% undergraduate and 50.63% graduate), and highly adherent to social distancing (85.15%). Approximately half of the individuals were involved in health care (workers or students), and some participants had had contact with confirmed or suspected cases of COVID-19 (16.99%).

Table 1 shows a comparison of the questions asked in both T1 and T2. Face mask use increased from 43.60% in T1 to 90.52% in T2 (P < .0001; Effect Size [ES] = 0.39) and people started using masks 2 weeks before T2 (18.72% in T1 vs 30.39% in T2; P < .0001; ES = 0.42). Health literacy on face masks also changed within these 2 weeks and people started to assume everybody should use a face mask (62.93% in T1 vs 94.12% in T2; P < .0001; ES = 0.29) during outside activities (43.60% in T1 vs 90.52% in T2; P < .0001; ES = 0.42), predominantly because such use may reduce the risk of infection among healthy people (78.57% in T1 vs 93.14% in T2; P < .0001; ES = 0.17). Participants also started to think about the protective role of homemade masks (18.35% in T1 vs 5.56% in T2; P < .0001; ES = 0.16). Participants agreed that masks should be used by everyone, whether health professionals/students or not (9.85% in T1 vs 2.61% in T2; P < .01; ES = 0.12).

The T2 presents new information about maintenance of social distancing (88.53%), access to homemade or purchased masks (89.60%), number of people that live with the participant (2 or 3, 66.67%, and 4 or more people, 17.07%), working status (72.53% have a job), and the COVID-19-related symptoms that participants had experienced by that time (80.83%). Regarding those symptoms, the most frequent ones were headache (50.83%), cough (47.50%), sore throat (44.17%), rhinorrhea (45.00%), fever (32.50%), and asthenia (29.17%).

With a powerful sample (Cohen’s h = 0.94) we performed simple and multiple logistic regression analyses including reported face mask use, time of questionnaire application, age, gender, and education (Table 2), showing that face mask use is significantly associated with fewer reports of COVID-19 symptoms when some independent variables are controlled (OR = 0.65, P = .01, 95% CI 0.48; 0.88), as presented in model 4.

Table 2.

Simple and Multiple Logistic Regression (Dependent Variable: Self-Report of COVID-19-Related Symptoms).

	Simple			Model 1			Model 2			Model 3			Model 4
	P	OR	(95% CI)	P	OR	(95% CI)	P	OR	(95% CI)	P	OR	(95% CI)	P	OR	(95% CI)
Reported face masks use	.32	0.88	(0.67; 1.14)	.03	0.73	(0.54; 0.97)	<.01	0.62	(0.46; 0.84)	<.01	0.63	(0.47; 0.85)	.01	0.65	(0.48; 0.88)

Abbreviations: P, P-value; OR, odds-ratio; CI, confidence interval.

Model 1: adjusted for social distancing engagement and time of application.

Model 2: adjusted for social distancing engagement, time of application and age.

Model 3: adjusted for social distancing engagement, time of application, age and gender.

Model 4: adjusted for social distancing engagement, time of application, age, gender, and education.

Discussion

This study showed evidence that educated Brazilians behaved differently regarding face mask use as the outbreak went on. We found that face mask use was significantly associated with fewer reports of COVID-19-related symptoms.

During the first 3 to 4 months of the COVID-19 pandemic, the majority of health authorities in the West resisted recommending face masks for all, in order to avoid a shortage of personal protective equipment (PPE) at health facilities when the COVID-19 outbreak worsened. This advice prevented the general public from knowing that face masks could protect them against SARS-CoV-2 by providing a physical barrier against the virus. Cloth face masks over the mouth and nose of the speaker have been proven to reduce the spread of oral or nasal fluid droplets.¹⁰ Depending on the fabric used, cloth face masks also promote aerosol filtration³ and prevent respiratory infection^3,11,12,13 as suggested by fewer self-reports of COVID-19 symptoms in our study. Despite cloth face masks having been promoted because of the unavailability of surgical masks, they are an effective and cheap way of self-protection against SARS-CoV-2.

Considering that a “masks for all” recommendation would help to minimize personal and population risk of infection,^7,8 we proposed this research that described how the inhabitants of the most populous Brazilian state, São Paulo, behaved when the COVID-19 outbreak started. Information on the protective potential of homemade masks⁷ was nationally disclosed the day before T2 started and might have affected the T2 responses.

The responses we obtained in T1 and T2, only 2 weeks apart from each other, showed a significant change in attitude regarding the use of face mask (as confirmed through multiple logistic regression), which may have been influenced by the observation that the severity of COVID-19 pandemic was increasing. Health literacy on masks efficiency for protection of both symptomatic and asymptomatic people was significantly higher at T2. The change in personal behavior toward face mask use at T2 does not capture the effect of mask recommendation of the Ministry of Health. The difference we observed between T1 and T2 possibly shows that educated Brazilians were becoming more aware of the COVID-19 severity and its prevention strategies even before the recommendation by the Ministry of Health.

Our data regression analysis showed that reported face mask use was associated with less self-reported symptoms of COVID-19, even after controlling for confounders, such as age, gender, and education. In our models, we hypothesized that the time of application of the questionnaires could have interfered with the fewer self-reports of COVID-19 symptoms, so we included it as an independent variable.

Albeit social distancing showed to be a short-term highly effective strategy for stopping the spread of SARS-CoV-2, face mask use might be an important approach when it comes to a long-term policy.⁴ Even essential outside activities such as going to the supermarket can be considered a potentially risky situation, especially when COVID-19 cases are underreported⁴ as was happening in Brazil. It must be noted that most Western countries differ from East Asian ones, as the latter do not routinely use face masks for respiratory diseases. Nowadays, this cultural resistance to wearing face masks may end lots of lives.

Although 85% of the participants in our study limited their movements and stayed at home, population surveys have shown that adherence rates to social distancing have dropped sharply in São Paulo state since April 2020, going against WHO and State Government’s recommendations. Those who responded to our questionnaires may have remained more frequently isolated, and one of the reasons may have been that they had higher incomes.

We have demonstrated that the use of face masks was already high among educated Brazilians before the authorities formally recommended it, and it was associated with fewer self-reported COVID-19 symptoms.

Despite having a powerful sample size, we are aware of our study limitations. First, we had a predominance of highly educated participants in our sample which does not reflect the educational distribution of the general Brazilian population. Second, the use of self-reported symptoms in the present survey is another limitation. Third, our recruitment method was based on the researcher’s personal network and led to over-representation of individuals involved with the health field. Fourth, only individuals who had internet access participated in the study and this could have contributed to participation bias. Lastly, we have no data on the ethnicity of this sample, hence we could not analyze racial effects on face mask use.

So What?

What is already known on this topic?

It is known that SARS-CoV-2 is transmitted through droplets, and a face mask can minimize viral spreading.

What does this article add?

Our study shows that face mask use was strongly associated with fewer self-reported COVID-19 symptoms regardless of age, education, and gender. We also found that Brazilians started using face masks as the COVID-19 pandemic worsened.

What are the implications for health promotion practice or research?

As rates of social distancing drop, face mask use should be encouraged for both individual and collective protection. Using face masks may avoid preventable infections and should be promoted as a health policy in those countries which do not usually do so.

Footnotes

Authors’ Note

Danilo Euclides Fernandes: Questionnaires design, analysis, and interpretation of data, drafting the article, final approval of the version to be published.

Michelle Tiveron Passos: Questionnaires design, analysis, and interpretation of data, drafting the article, final approval of the version to be published.

Gianna Mastroianni Kirsztajn: Conception and design of the study, analysis, and interpretation of data, drafting the article, final approval of the version to be published.

Ethics statement

This is a unique center study developed at Universidade Federal de São Paulo (UNIFESP), Brazil. This protocol was approved by the Ethical Committee on Research of the UNIFESP (CAAE 30890420.7.0000.5505) and was developed according to the principles expressed in the Declaration of Helsinki. All participants provided their consent by the end of the questionnaires.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 2019/05266-5) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), finance code 001.

ORCID iD

Danilo Euclides Fernandes

References

Liu

Zhang

COVID-19: face masks and human-to-human transmission. Influenza Other Respir Viruses [Internet]. 2020;14(4):472–473. doi:10.1111/irv.12740

Cheng

Lam

Leung

. Wearing face masks in the community during the COVID-19 pandemic: altruism and solidarity. Lancet. 2020;S0140673620309181.

Konda

Prakash

Moss

Schmoldt

Grant

Guha

. Aerosol filtration efficiency of common fabrics used in respiratory cloth masks. ACS Nano. 2020;14(5):6339–6347.

Zhou

Yue

Zhang

. Mask is the possible key for self-isolation in COVID-19 pandemic. J Med Virol [Internet]. 2020;92(10):1745–1746. doi:10.1002/jmv.25846

Zhu

Zhang

Wang

, et al. A novel coronavirus from patients with pneumonia in China, 2019. N Engl J Med. 2020;382(8):727–733.

Czechs get to work making masks after government decree | World news. The Guardian [Internet]. 30 March 2020. Accessed April 25, 2020. https://www.theguardian.com/world/2020/mar/30/czechs-get-to-work-making-masks-after-government-decree-coronavirus

Brazilian Ministry of Health. Máscaras caseiras podem ajudar na prevenção contra o Coronavírus [Internet]. March 2020. Accessed April 25, 2020. https://www.saude.gov.br/noticias/agencia-saude/46645-mascaras-caseiras-podem-ajudar-na-prevencao-contra-o-coronavirus

World Health Organization. When and how to use masks [Internet]. March 2020. Accessed April 25, 2020. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public/when-and-how-to-use-masks

Lancet

COVID-19 in Brazil: “So what?”

Lancet. 2020;395(10235):1461. Elsevier.

10.

Anfinrud

Stadnytskyi

Bax

. Visualizing speech-generated oral fluid droplets with laser light scattering. N Engl J Med. 2020;382(21):2061–2063.

11.

Clase

Joseph

, et al. Cloth masks may prevent transmission of COVID-19: an evidence-based, risk-based approach. Ann Intern Med. 2020;173(6):489–491.

12.

Lima

Cavalcante

Macêdo

Galindo Neto

Caetano

JÁ

Barros

. Cloth face masks to prevent Covid-19 and other respiratory infections. Rev Lat Am Enfermagem. 2020;28:e3353.

13.

Cowling

Zhou

Leung

Aiello

. Face masks to prevent transmission of influenza virus: a systematic review. Epidemiol Infect. 2010;138(4):449–456.