Covid-19 in Brazil: Learning How to Walk in the Dark Without Leaving Anything Behind

Coronavirus; COVID-19; Pandemics; Catastrófica Illness/mortality; Hospitalization/economics; Equity in Access to Health Services; Diagnostic Tests/methds; Personal Protective Equipment; Masks; Ventilators Mechanical

“If you have ups and downs, celebrate it because it means you’re alive!”

The SARS-CoV-2 pandemic had its first case in Brazil on February 4, 2020 in the state of São Paulo. Like a real meteor, it practically paralyzed the planet without a more effective therapeutic approach to fight the virus, in addition to practices adopted worldwide only decades or even hundreds of years ago, with a very low level of robust modern evidence.¹ Brazil had the advantage of previewing the pandemic as it was a few weeks behind its Asian and European peers, so it was able to identify the hits and failures of these countries in their preparation to face the problem.

We have then found out by chance that we have the world’s third biggest number of ICU beds, second only to the USA and Germany,² but we are also faced with an enormous heterogeneity between states, limitations and bureaucracy for the purchase of personal protective equipment and testing kits. This has prevented some pandemic approaches taken by other countries from being adopted here, either due to the impossibility of a lockdown in a continental country, or due to the speed and costs to identify the cases of transmission. Besides, limitations on the number of tests to be done has left us without a precise guide as to the spread of the pandemic, making it difficult to plan the timely, orderly and efficient allocation of scarce resources.

The solution to that was to adopt our own strategies, which would allow us to walk in the dark while at least having a better idea of how far along the path we were. This information is essential for important decisions impacting not only the economic scenario of Brazil, but also the entire healthcare chain of a population that has been deprived of access to outpatient services and elective treatments due to the pandemic. Theoretical epidemiological mathematical models proved to be little capable of predicting our real figures, either because they overestimate the lethality of the disease that seems to stand closer to 0.2–0.5%,³ or because they are based on data from past pandemics with other transmission dynamics. Interestingly, the models that were most suitable for modeling at what point in the pandemic we are resulted from curve adjustment strategies or Baysean models using data from other countries or based on the preliminary data that we already had on our curves.⁴ , ⁵ These models appeared relatively spontaneously and outside traditional research groups, but proved to be more assertive in determining different moments of the pandemic.⁶

In a simpler way, but also using data on how the pandemic has previously progressed in other countries, we analyzed the charts of new cases per day from 30 countries with the highest number of covid-19 cases according to the following criteria: countries had reached a peak and presented at least 5 days of decline or stabilization of new daily cases.⁷ China was excluded as it concentrated thousands of previous cases on a later day; Brazil was excluded because it was subject to the application of the result. Five countries that did not have mandatory isolation were included. Of the 30 countries, 18 were considered to have completed the full pandemic cycle, with a number of new daily cases <70% of the peak of new cases per day. Using the date of the 1^st, 100^thand 200^thcase or 10^thdeath, the times between these dates and the peak of new cases/day were determined. From these results, it was found that 95% of the countries studied had their peaks 55±8 days from the 1^stcase, 31±5 days from the 100^thcase, 27±5 days from the 200^thcase and 19±4 days from the 10^thdeath. With this data, it would have been possible to establish the peaks of new cases, deaths and use of the hospital system in different states and cities in Brazil, even without being sure of the exact number of cases due to underreporting, based on the behavior of the pandemic in countries with different healthcare systems and mitigation measures. Death peak dates were established after 14 days of peak of cases and use of the hospital system after 26 days of peak of cases, considering incubation times, manifestation of symptoms, hospitalization and potential clinical worsening.⁸ With these criteria, we estimated the different peaks in each of the Brazilian states with the highest number of cases, as shown in Table 1 (limited only to the forecast from the 100^thcase to exemplify the model). What cannot be predicted with any of these models is the descending portion of the curve, faster in some places and much slower, showing a plateau in others, requiring hospital beds for longer times, a potential virulence effect that is different from the virus after many mutations.⁹

Table 1
– Prediction of state peaks of new cases/day, deaths and hospital use based on modeling from 30 countries (shown only with an estimate from the 100th case). Estimated data for research purposes, pending modification and verification

State	1stCase	100thCase	200thCase	10thDeath	Peak 100thCase	Lower 95%CI	Upper 95%CI	Death Peaks (100)	Hospital Peak (100)
SP (Metrop. Area)	Feb 4th	Mar 2nd	Mar 6th	Mar 6th	Apr 2nd	Mar 28th	Apr 6th	Apr 16th	Apr 28th
CE	Feb 14th	Mar 7th	Mar 10th	Mar 15th	Apr 7th	Apr 2nd	Apr 11th	Apr 21st	May 3rd
GO	Mar 2nd	Mar 8th	Mar 11th	Mar 26th	Apr 8th	Apr 3rd	Apr 12th	Apr 22nd	May 4th
SC	Feb 28th	Mar 14th	Mar 17th	Apr 4th	Apr 14th	Apr 9th	Apr 18th	Apr 28th	May 10th
RJ	Feb 27th	Mar 15th	Mar 18th	Mar 18th	Apr 15th	Apr 10th	Apr 19th	Apr 29th	May 11th
DF	Feb 26th	Mar 15th	Mar 18th	Apr 4th	Apr 15th	Apr 10th	Apr 19th	Apr 29th	May 11th
BA	Feb 26th	Mar 16th	Mar 19th		Apr 16th	Apr 11th	Apr 20th	Apr 30th	May 12th
RN	Mar 8th	Mar 18th	Mar 21st	Apr 7th	Apr 18th	Apr 13th	Apr 22nd	May 2nd	May 14th
RS	Mar 9th	Mar 21st	Mar 25th	Apr 8th	Apr 21st	Apr 16th	Apr 25th	May 5th	May 17th
MG	Mar 17th	Mar 23rd	Mar 27th	Apr 2nd	Apr 23rd	Apr 18th	Apr 27th	May 7th	May 19th
MT	Mar 19th	Mar 24th	Mar 27th	Apr 26th	Apr 24th	Apr 19th	Apr 28th	May 8th	May 20th
PR	Mar 12th	Mar 26th	Apr 1st	Apr 6th	Apr 26th	Apr 21st	Apr 30th	May 10th	May 22nd
AM	Mar 18th	Mar 28th	Apr 1st		Apr 28th	Apr 23rd	May 2nd	May 12th	May 24th
PE	Mar 12th	Apr 2nd	Apr 5th	Apr 1st	May 3rd	Apr 28th	May 7th	May 17th	May 29th
MA	Mar 20th	Apr 5th	Apr 7th	Apr 6th	May 6th	May 1st	May 10th	May 20th	Jun 1st
PA	Mar 18th	Apr 6th	Apr 10th	Apr 11th	May 7th	May 2nd	May 11th	May 21st	Jun 2nd
PB	Mar 19th	Apr 11th	Apr 17th	Apr 9th	May 12th	May 7th	May 16th	May 26th	Jun 7th
AL	Mar 10th	Apr 17th	Apr 21st	Apr 18th	May 18th	May 13th	May 22nd	Jun 1st	Jun 13th
PI	Mar 19th	Apr 17th	Apr 21st	Apr 18th	May 18th	May 13th	May 22nd	Jun 1st	Jun 13th

• CI: confidence interval.

But if we are used to being guided by more precise figures in cardiology, how do we know if the estimates are correct and if we have been drifting away from the actual figures? We have then looked for official sources that could allow us to infer and check these calculations. Unfortunately, at least until this date of the pandemic, there has been tremendous confusion regarding data reporting, which has caused the interpretation of the pandemic phase in Brazil to be greatly impaired. Due to delayed determination of SARS-CoV-2 infections, many cases have been reported days and even weeks late, causing the official authorities to report numbers of confirmed cases as numbers of actual daily cases, confusing press bulletins and often causing unnecessary fuss, especially when accumulated numbers of weekends and holidays were late announced on Tuesdays.¹⁰ In order to try to understand the numbers, it is necessary to check different sources with adjusted numbers and, above all, to look for more realistic statistics on the information on deaths occurring in the country, as this metric is much more robust from the point of view of reporting despite reflecting what happened 14 days prior. In this regard, it is worth mentioning the important contribution of the data reported on the Transparency Portal organized by the National Association of Registrars of Vital Statistics, which allows a more accurate monitoring of the number of deaths from Covid-19 or suspected deaths on the actual date of the occurrence, rather than on the reporting date.¹¹ Along with this data, information on monitoring hospitalizations for Serious Acute Respiratory Syndromes through the InfoGripe system also helped to continuously monitor trends and confirm or not the forecasts made.¹²

Despite all these tools at hand, the social isolation measures were taken in a very controversial way, often not considering the stage of the pandemic cycle we were at, with a late adoption, and sometimes following a course of action without presenting consistent data that would justify the measures taken. Given the great difference between the structure of resources found in the country and the phases of the pandemic in each state, the degrees of isolation should certainly be quite different since each measure individually or together has different effects on reducing viral transmission. In this sense, we must also remember the Pareto principle, where 20% of what we do reaches 80% of the result: the correct application of well-done social distancing, with 25% reduction from the original distancing, allows an effective transmission response to be maintained once R0 is initially reduced.¹³ Therefore, relatively simple measures of advising the population to wash their hands, keeping away from others, wearing masks, etc., as long as they are well applied, can often be better than attempts at taking drastic yet disorderly actions not well understood by the population concerned and without proper preparation.

The decision of what to do and at what point in the pandemic is crucial so that we do not turn the goal of saving lives into a merely appealing pitch with consolidated results that result in more deaths than lives saved. The second and third order effects that occur in any therapy can often be more harmful than the treatment itself, especially when it is carried out without proper planning. Very common in situations where centralization tries to simplify extremely complex processes that involve multiple chains (as we have in a famous example of pencil making),¹⁴ the final effect can be exactly the opposite of what we are pursuing. And here we have several situations where an unnecessary extension of confinement measures can lead to a greater number of deaths than from the disease itself. As of writing this, the number of deaths from Covid-19 in Brazil is about 10,000 patients and we have almost 45 days of isolation. Due to an expected 30% reduction in primary angioplasty surgeries, increased transfer times and rehospitalizations due to untreated acute coronary syndromes,¹⁵ an excess of cardiovascular deaths of over 3,000 is estimated in this period. Missing outpatient visits of various specialties increases the risk of death by up to 1.5 times, adding to this excess another 9,000 unnecessary deaths.¹⁸ The 1% increase in unemployment or the drop in Gross Domestic Product alone is associated with an increase in deaths of up to 1.63 times in the economically active population, adding another 3,500 excess deaths to this amount.¹⁹ Mortality calculations due to failure of therapies for neoplasms and misdiagnoses in Brazil do not yet exist, but in the USA and England, excess deaths were estimated at 34,000 and 6,000 respectively.²⁰ All of these deaths are associated with several causes, mainly due to the lack of timely access to overburdened healthcare systems solely focused on one cause of mortality. These will be the invisible deaths of the pandemic and isolation because they were not predicted as the side effects to single-point measures.

What will come in the next weeks and months is beyond the views expressed here and will depend, to a great extent, on how we will choose to defuse this health crisis, sooner or later. Expensive spending on chronic diseases is likely to increase dramatically in the coming months,²¹ overburdening a government already in debt. Unemployment and declining income will lead many Brazilians to migrate to a Public Healthcare System already burdened from the demand repressed over these months. At the same time, other solutions to increase the efficiency of the systems will be improved as they have been in this short period of time, such as digital prescriptions and telemedicine.²² This increase in medical productivity can partly alleviate these increases in demand and costs, making our healthcare system to demonstrate the resilience and effectiveness that has made us to witness a mortality per million up to 10 times less than other European countries at the same stage of the disease, despite all our difficulties. What we do know is that we will be ready for new challenges, as optimists have always surprised us as to how human inventiveness is capable of overcoming obstacles.

Referências

• ¹
  Nussbaumer-Streit B, Mayr V, Dobrescu AI, Chapman A, Persad E, Klerings I, et al. Quarantine alone or in combination with other public health measures to control COVID-19: a rapid review. Cochrane Database Syst Rev. 2020;4(4):CD013574.
• ²
  Wikipedia. List of countries by hospital beds. [Cited in 2020 Apr 20] Available from: https://en.wikipedia.org/wiki/List_of_countries_by_hospital_beds
  » https://en.wikipedia.org/wiki/List_of_countries_by_hospital_beds
• ³
  Eich B B. https://docs.google.com/spreadsheets/d/1zC3kW1sMu0sjnT_vP1sh4zL0tF6fIHbA6fcG5RQdqSc/htmlview#gid=0 2020.
  » https://docs.google.com/spreadsheets/d/1zC3kW1sMu0sjnT_vP1sh4zL0tF6fIHbA6fcG5RQdqSc/htmlview#gid=0
• ⁴
  Breban R, Vardavas R, Blower S. Theory versus data: how to calculate R0? PLoS One. 2007;2(3):e282.
• ⁵
  Murray CJ. Forecasting COVID-19 impact on hospital bed-days, ICU-days, ventilator-days and deaths by US state in the next 4 months. medRxiv. 2020 Mar. 30.
• ⁶
  Dana S, Simas AB, Filardi BA, Rodriguez RN, Valiengo LLC, Gallucci-Neto J. Brazilian modeling of COVID-19 (BRAM-COD): a Bayesian Monte Carlo approach for COVID-19 spread in a limited data set context. medRxiv. 2020 May 17.
• ⁷
  Worldometer. Coronavirus Cases; 2020. [citado 20 abr. 2020]. Disponível em: https://www.worldometers.info/coronavirus/coronavirus-cases/
  » https://www.worldometers.info/coronavirus/coronavirus-cases/
• ⁸
  Zhou F, Yu T, Du R, Fan G, Liu Y, Liu Z, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020;395(10229):1054-62.
• ⁹
  Yao H, Lu X, Chen Q, Xu K, Chen Y, Cheng L, et al. . Patient-derived mutations impact pathogenicity of SARS-CoV-2. medRxiv. 2020 Apr. 23.
• ¹⁰
  Ministério da Saude. Painel Coronavírus. Rio de Janeiro, DF: Ministério da Saúde; 2020. [citado 20 abr. 2020]. Disponível em: https://covid.saude.gov.br/
  » https://covid.saude.gov.br/
• ¹¹
  Brasil. Portal da Transparência [Internet].Especial COVID-19. Painel Registral; 2020. [citado 20 abr. 2020]. Disponível em: https://transparencia.registrocivil.org.br/especial-covid
  » https://transparencia.registrocivil.org.br/especial-covid
• ¹²
  Fiocruz. Info Gripe [Internet]. Rio de Janeiro: Fiocruz; 2020. [citado 20 abr. 2020]. Disponível em: http://info.gripe.fiocruz.br/
  » http://info.gripe.fiocruz.br/
• ¹³
  Lai S, Ruktanonchai NW, Zhou L, Prosper O, Luo W, Floyd JR, et al. Effect of non-pharmaceutical interventions to contain COVID-19 in China. Nature. 2020.[Epub ahead of print].
• ¹⁴
  Lopes B. Milton Friedman, a história de um lápis: A perspectiva e a lição que talvez passou despercebido; 2018. [citado 20 abr. 2020]. Disponível em: https://medium.com/@brunolopes_61254/milton-friedman-a-hist%C3%B3ria-de-um-l%C3%A1pis-a-perspectiva-e-a-li%C3%A7%C3%A3o-que-talvez-passou-despercebido-500de4c7a84a
  » https://medium.com/@brunolopes_61254/milton-friedman-a-hist%C3%B3ria-de-um-l%C3%A1pis-a-perspectiva-e-a-li%C3%A7%C3%A3o-que-talvez-passou-despercebido-500de4c7a84a
• ¹⁵
  Metzler B, Siostrzonek P, Binder RK, Bauer A, Reinstadler SJ. Decline of acute coronary syndrome admissions in Austria since the outbreak of COVID-19: the pandemic response causes cardiac collateral damage. Eur Heart J. 2020;41(19):1852-3.
• ¹⁶
  Balk M, Gomes HB, Quadros AS, Saffi MAL, Leiria TLL. Comparative analysis between transferred and self-referred STEMI patients undergoing primary angioplasty. Arq Bras Cardiol. 2019;112(4):402-7.
• ¹⁷
  Oliveira LMSM, Costa IMNBC, Silva DGD, Silva JRSS, Barreto-Filho JAS, Almeida-Santos MA, et al. Readmission of patients with acute coronary syndrome and determinants. Arq Bras Cardiol. 2019;113(1):42-9.
• ¹⁸
  McQueenie R, Ellis DA, McConnachie A, Wilson P, Williamson AE. Morbidity, mortality and missed appointments in healthcare: a national retrospective data linkage study. BMC Med. 2019;17(1):2.
• ¹⁹
  Roelfs DJ, Shor E, Davidson KW, Schwartz JE. Losing life and livelihood: a systematic review and meta-analysis of unemployment and all-cause mortality. Soc Sci Med. 2011;72(6):840-54.
• ²⁰
  Lai AG, Pasea L, Banerjee A, Denaxas S, Katsoulis M, Chang WH, et al. Estimating excess mortality in people with cancer and multimorbidity in the COVID-19 emergency. Preprint. 2020 Abr 20. Disponível em: https://www.researchgate.net/publication/340984562_Estimating_excess_mortality_in_people_with_cancer_and_multimorbidity_in_the_COVID-19_emergency?channel=doi&linkId=5ea8b957a6fdcc7050976a3e&showFulltext=true 2020.
  » https://www.researchgate.net/publication/340984562_Estimating_excess_mortality_in_people_with_cancer_and_multimorbidity_in_the_COVID-19_emergency?channel=doi&linkId=5ea8b957a6fdcc7050976a3e&showFulltext=true
• ²¹
  Stevens B, Pezzullo L, Verdian L, Tomlinson J, George A, Bacal F. The economic burden of heart conditions in Brazil. Arq Bras Cardiol. 2018;111(1):29-36.
• ²²
  Lopes MACQ, Oliveira GMM, Ribeiro ALP, Pinto FJ, Rey HCV, Zimerman LI, et al. Guideline of the Brazilian Society of Cardiology on Telemedicine in Cardiology - 2019. Arq Bras Cardiol. 2019;113(5):1006-56.

Publication Dates

• Publication in this collection
  03 July 2020
• Date of issue
  June 2020