COVID-19 and outpatient care: a nationwide household survey

We aimed to assess the proportion of the population in 133 Brazilian municipalities who – from March to August 2020 – had a health problem but failed to seek care or failed to attend to a health service for routine appointment or examination. We conducted a household survey from August 24-27 in 133 Brazilian cities by asking the subjects if, since the beginning of the COVID-19 pandemic in March 2020, they had suffered from a health problem but did not seek care or failed to attend to a routine or screening examination. Poisson regression was used for the analyses. We interviewed 33,250 subjects and 11.8% (95%CI: 11.4-12.1) reported that, since March 2020, they failed to seek care despite being ill, 17.3% (95%CI: 16.9-17.7) failed to attend to a routine or screening examination and 23.9% (95%CI: 23.4-24.4) reported one or both outcomes. Health service closure and fear of the COVID-19 infection were the main reasons for not seeking care. Women and the poorest were more likely to not look for a health service, despite having a health problem or a scheduled routine appointment. On the other hand, those subjects who self-identified as white were less likely to not look for a health service. The COVID-19 pandemic is more critical for the indigenous people and the poorest, and these people are also more likely to not seek care for other health conditions during the


Introduction Methods
The data used were from the fourth round of a repeated population-based serological survey conducted from August 24-27, 2020 in 133 Brazilian sentinel cities.The cities included Brasília, 26 state capitals and the largest municipalities in each of the country's regions, as defined by the Brazilian Institute of Geography and Statistics (IBGE).Within each city, 25 urban census tracts were selected with probability proportionate to size, and 10 households were randomly sampled in each tract using households lists elaborated by the IBGE.In each household, all residents were enrolled using a smartphone app and one was randomly selected to be interviewed and tested for the presence of antibodies for SARS-CoV-2, using the WONDFO SARS-CoV-2 Antibody Test (Wondfo Biotech Co., https:// en.wondfo.com.cn/).If the selected resident refused to provide a finger prick blood sample, a second household member was randomly selected.If this person also refused, the interviewers moved on to the next household.The next household to the right was also selected in case of absent residents.Further details on the study methodology have been previously published 7 .
The interview included a short questionnaire for collecting sociodemographic information (gender, age, schooling, race/skin color, household size, and household assets) and compliance with social distancing measures.The Brazilian race/skin color classification recognizes five groups based on the question: "How do you classify yourself regarding skin-color or race?".The five response options are "white", "mixed-race" (pardo), "black", "yellow (Asian)", and "indigenous".Interviewers were instructed to check the "yellow" option when the subject mentioned being of Asian descent, and "indigenous" when any of the multiple indigenous nations were mentioned.The "mixed-race" category reflects mixed ancestry including European, African, and/or indigenous backgrounds.Socioeconomic position was assessed using a wealth index by analyzing household assets 8  subjects were asked about the presence of the following chronic conditions: hypertension, diabetes, asthma, cancer, chronic kidney disease, heart disease, or other chronic condition.
Considering health services, the subjects were asked if, since the onset of the pandemic in March 2020, they had suffered from a health problem but did not seek care or failed to attend to a routine or screening examination.For positive answers, the reason for failing to seek care was also recorded.
The Stata 15 software (https://www.stata.com)was used for the analyses.Numbers were compared using the chi-square test.For ordinal variables, the p-value for linear trend and for heterogeneity were estimated and the one with the lower value was presented.Poisson regression with robust variance was used to estimate prevalence ratios, and all analyses controlled for the cluster-sampling design using svy prefix.
Interviewers wore individual protection equipment (aprons, gloves, surgical face masks, and shoe and hair covers) that were discarded as hospital waste after each interview.They were tested for COVID-19 prior to the field work and every two days thereafter.The study was approved by the Brazilian National Research Ethics Committee (process n.CAAE 30721520.7.1001.5313),with written informed consent form from all participants; for children and adolescents, consent was provided by parents or guardians.

Results
We interviewed 33,250 subjects in the fourth round of the seroprevalence survey.In this round, the response rate was 59.1%, and non-response was mostly due to whole families being away from home.Compared with the overall Brazilian population, men and young individuals were less likely to be interviewed.Concerning race/skin color, indigenous individuals were overrepresented in the studied sample, whereas the proportion of subjects who reported being white was lower than the national estimates (Table 1).
Women were more likely than men to not look for a health service, despite having a health problem or skipped a scheduled appointment.Regarding age, the risk of not seeking health care slightly increased for individuals aged 40-59 years and then declined, whereas for routine examination the risk of missing the visit increased steadily according to age.Wealth was inversely associated with the risk of not seeking care or failing to attend to a routine examination.The number of subjects who did not seek health services was higher in the North and Northeast regions.Finally, both outcomes were positively related to the number of previous chronic conditions (Table 3).
Concerning the reasons for not seeking care, closure of health services was less frequently reported among the wealthiest, while fear of getting COVID-19 infection was positively associated with wealth.Subjects who self-reported being white were more likely to mention being afraid of getting COVID-19 infection as a reason for failing to seek care.The number of chronic conditions was not associated with fear of getting COVID-19 infection as a reason for failing to seek care, whereas closure of health services was more frequently reported among those with one or more preexisting chronic condition (Table 4).

Table 1
Distribution of the study sample according to socioeconomic and demographic characteristics.

Discussion
In the fourth round of a large population-based serological survey of prevalence of antibodies against SARS-CoV-2, we observed that about one out of four subjects either failed to seek care due to a health problem or missed a routine examination.Among these subjects, about half did not attend to a health service for fear of getting COVID-19 infection, whereas the closure of health services was reported by two out of ten subjects.The likelihood of not seeking care was higher among women, the poorest, those living in the North or Northeast regions and those who reported having preexisting chronic conditions.In the United States, Ziedan et al. 5 reported that state closure policies and suspension of nonelective medical procedures were not the major explanation for the decline in seeking health care, and suggested that patients' fear of COVID-19 infection should be considered as an alternative explanation.By asking about reasons for not seeking care, we could identify that fear of getting COVID-19 infection was the main factor for avoiding health care, but public policies related to the pandemic also affected the use of health services.Individuals who failed to attend to a health service may look for it in the future.Studies conducted in high-income countries have reported a later rescheduling, but it seems that the analysis is incomplete 2,3,4,5 .A study conducted in Italy 9 , evaluating cases of delayed access to health care, identified that fear of COVID-19 infection was the main reason for not seeking care.The delay in seeking health care may have current or future negative consequences, as shown by a modelling study based on data from the National Health Service (United Kingdom) on the impact of COVID-19 on cancer diagnosis delay and cancer survival 10 .Compared with pre-pandemic figures, the authors suggested that the mortality rate would increase from 7.9 to 9.6% for breast cancer, 15.3 to 16.6% for colorectal cancer, and 5.8 to 6% for esophageal cancer 10 .
We observed that the highest number of subjects who failed to seek health care were in the North and Northeast regions of Brazil.We expected this result because these regions showed the highest seroprevalence 11 and hospital case-fatality 12 at the time of the survey.Moreover, in Manaus, capital of the state of Amazonas, a large mortality rate due to respiratory diseases and other causes was observed in the first weeks of the pandemic 13 , which could indicate a collapse of the health care system.Our result could also be due to lower access and use of health services that were presented before the pandemic.Stopa et al. 14 , using data from the Brazilian National Health Survey, reported that people living in the South and Southeast regions had greater access to health services than residents in the other regions of Brazil.This difference can be explained by a greater offer of health services, both public and private, as well as the socioeconomic level of the population 15 .Medina et al. 16 suggest the need for a more territorialized, community-based, and home-based approach on primary health care to face any epidemic.This approach could accurately identify those people who are failing to seek health care during the pandemic and consequently failing to treat other diseases.Medina et al. also indicated the need to strengthen primary care to fight epidemics, guaranteeing services that have a more territorialized, community-based and home-based approach, aiming to achieve greater success in dealing with health crises and maintaining the continuity of care for other pre-existing conditions.
The COVID-19 pandemic is more critical for indigenous and the poorest 17,18 , and our study suggests that these individuals are also more likely to not seek care for other health conditions during the pandemic.This association continued after controlling for regions of the country as a confounding factor, since poor and indigenous Brazilians are concentrated in the North and Northeast.Lately, socioeconomic and health inequalities have diminished in Brazil, but not all ethnic groups have benefited from such advances, and wide inequalities in health and nutrition among Brazilian indigenous populations have been reported 19,20 .The negative consequences of delaying care or diagnosis of some conditions may increase social and health inequalities in the near future.
. The first component was divided into quintiles.The highest grade successfully completed was recorded as schooling level.The Cad. Saúde Pública 2022; 38(4):e00194121

Table 2
Proportion of subjects who failed to seek health care, or missed a routine or screening examination.

Table 3
Proportion and prevalence rate of subjects who failed to seek health care, or missed a routine or screening examination, according to socioeconomic, demographic, and number of preexisting chronic diseases.

March 2020, had a health problem and failed to seek health care Since March 2020, failed to attend to a health service for a routine or screening examination
* Test for heterogeneity; ** Test for linear trend.Cad.Saúde Pública 2022; 38(4):e00194121

Table 4
Proportion and prevalence rate of subjects who failed to seek health care because the service was closed or due to fear of getting COVID-19 infection according to socioeconomic, demographic, and number of previous chronic diseases.