A monitoring and surveillance system of microcephaly was implemented by the Brazilian Ministry of Health (MoH) in November 2015, in response to an outbreak of Zika virus infections.1 Weekly reports demonstrate an increase in reported cases of microcephaly, as well as the number of fetal and neonatal deaths related to it.2 More recently, the association between Zika virus infection in pregnancy and microcephaly in newborns was confirmed.3
The MoH initially adopted a broad definition of suspected cases of microcephaly, as a head circumference (HC) ≤33 cm in term neonates (both genders) and a HC less than −2 standard deviation (SD) for preterm babies on the Fenton curves.4 Because a high number of false positives was observed adjustments in the criteria have been made. In December 2015, the cutoff point was decreased to ≤32 cm in term babies (both genders).5,6 Since March 2016, microcephaly was redefined as a HC less than −2 SD for age and sex according to the world health organization tables for term neonates (≤31.9 cm for boys and ≤31.5 cm for girls) and less than −2 SD according to the Intergrowth study for preterm infants.7
The weekly reports consolidate data collected from health facilities by the Secretaries of Health in municipalities across the entire country. Reported cases of microcephaly are categorized as suspected cases. They remain under investigation by neuroimaging and/or specific laboratory tests [polymerase chain reaction (PCR) and/or serology specific for Zika virus] until they are either confirmed or excluded as being a case of “microcephaly and neurologic abnormalities suggestive of congenital infection.” The suspected cases that progressed to fetal or neonatal death are categorized in the same way.2,4 As the availability of reverse transcription-PCR and/or serology tests (IgM/IgG) for Zika virus is limited, most confirmed cases rely on the presence of neuroimaging abnormalities suggestive of congenital infection. For Public Health purposes, the Brazilian MoH considers the presence of neuroimaging abnormalities as probably caused by Zika virus, as no other plausible cause such as syphilis, toxoplasmosis, rubella, cytomegalo virus and herpes virus infection seem to explain the sudden increase of cases during the outbreak.7
To October 15, 2016, there were 9862 reported cases: 6827 completed laboratory and neuroimaging investigations, resulting in 4764 excluded and 2063 confirmed cases of microcephaly and neurologic abnormalities, suggestive of congenital infection. Presently, 3035 remain under investigation.2 In this same period, there were 490 reported deaths: 268 completed the investigation, 171 were confirmed cases of microcephaly and/or neurologic abnormalities suggestive of congenital infection and 97 were excluded, leaving 222 deaths still under investigation.2
Because of limited resources, there has been a delay in the confirmation of cases of microcephaly and neurologic abnormalities suggestive of congenital. Accordingly, the true case fatality rate (CFR) for this condition is unknown. We estimated this CFR considering different alternatives for the potential number of deaths (numerator) and affected population (denominator) in the current epidemic in Brazil.
MATERIALS AND METHODS
In this analysis, we used data extracted from the epidemiologic reports published by the Secretary of Health Surveillance of the Brazilian MoH that monitors cases of microcephaly associated with congenital infection in Brazil, from November 10, 2015, to October 15, 2016.2 In these reports, the notified cases were classified as confirmed, excluded, under investigation or Zika virus-associated. For the present analysis, we used (1) under investigation cases, which are cases that have been reported but not yet completely investigated and (2) confirmed cases, which are those considered to be true microcephaly and neurologic abnormalities suggestive of congenital infection by neuroimaging and/or laboratory tests (PCR and/or serology specific for Zika virus). We calculated the CFRs considering a combination of possibilities for the numerator (number of deaths) and denominator (affected population). The numerators used had 3 possibilities, as of the date of the report: (1) confirmed fetal or neonatal deaths from microcephaly; (2) confirmed deaths + number of deaths if all under investigation cases are true microcephaly deaths and (3) confirmed deaths + number of deaths if the rate of confirmed deaths remains the same in the future. The rate of confirmed deaths was calculated as the percentage of deaths confirmed as true cases of microcephaly in relation to all cases of deaths that have completed investigation. A similar strategy was used for the denominator, which are cases of microcephaly and neurologic abnormalities suggestive of congenital infection. The denominators used had three possibilities, as of the date of the report: (1) confirmed cases of microcephaly; (2) confirmed cases + number of cases if all under investigation cases are true microcephaly and (3) confirmed cases + number of cases if the rate of confirmed cases remains the same in the future. The rate of confirmed cases was calculated as the percentage of cases confirmed as true cases of microcephaly in relation to all cases that have completed investigation.
CFRs were calculated dividing the numerator by the denominator and presented as percentages; 95% confidence intervals (CIs) were calculated for each CFR. χ2 was used to compare CFRs, with a significance established as 0.05, when appropriate. As we used freely available public data, there was no need for an ethics committee review or approval.
The possible CFRs are presented in Table 1. They varied from 3.4% (95% CI: 2.9–3.9) to 19.0% (95% CI: 17.4–20.8), depending on the calculation method. The current estimate of CFR, considering all confirmed cases of microcephaly and all associated fetal or neonatal deaths was 8.3% (95% CI: 7.2–9.6). The current rate of confirmation of cases is 30% (2063 confirmed cases/5098 confirmed + excluded cases). The confirmation rate of deaths is 64% (171/268). If cases of microcephaly and associated deaths follow the same rate of confirmation observed to date, the estimated CFR will be 10.5% (95% CI: 9.5–11.7) when the investigations of all the remaining suspected cases and deaths have been completed, which is significantly higher than the current estimate CFR of 8.3% (P = 0.018).
We found that the CFR for the fetal and neonatal deaths associated with microcephaly and neurologic abnormalities suggestive of congenital infection in Brazil varied from 3.4% to 19.0%. If the rate of confirmation of cases and deaths remains the same in the future, the CFR could be significantly higher than currently estimated.
The CFR of microcephaly associated with congenital Zika virus infection is unknown. A retrospective review of 19 identified cases of congenital brain lesions in fetuses and newborns in French Polynesia following an epidemic of Zika virus included 8 cases with major brain lesions and severe microcephaly, 6 cases with severe cerebral lesions without microcephaly and 5 with brainstem dysfunction. Medical termination of pregnancy was performed in 11 cases, which precludes the evaluation of the CFR. Two cases with brainstem dysfunction died in the first months of life.8
Our analysis suggests that the situation is worse than the current data suggest. The current confirmation rate of deaths is twice the confirmation rate of cases. If these rates remain the same in Brazil, the CFR could be as high as 10.5%, when all the diagnostic evaluation of existing cases and deaths remaining under investigation has been completed.
However, our analysis has limitations. First, inaccuracy in measurement of HC, as well as changes in case definition could have resulted in under/over ascertainment of suspected cases. However, the estimates of CFR were calculated in relation to cases and deaths that were confirmed by neuroimaging or specific laboratory tests. Second, large number of cases remaining under investigation hampers the quantification of the true number of microcephaly cases. Third, the limited availability of specific tests for Zika virus impairs the evaluation of the precise impact of the virus as a cause of the neurologic abnormalities. Fourth, the estimates are restricted to calculations applied to the confirmed cases and deaths as of July 2016, but the dynamics of the Zika epidemic in Brazil is currently unknown. It is not known if the infection gives short- or long-term immunity. Some authors suggest that there could be a second burst of infection if the virus, similar to herpes viruses, can remain dormant in the host cells. Furthermore, the association between gestational age at infection and the degree of microcephaly severity is also unknown. We recently performed an analysis using the MoH reports indicating that the number of new cases of microcephaly is declining in the country (submitted). If the outbreak is waning, the projected estimate of the CFR based on the confirmation rate of cases and deaths may be close to the true one. We believe our analysis contributes new data to the growing body of knowledge about the relation between microcephaly and the Zika virus in Brazil.
3. Rasmussen SA, Jamieson DJ, Honein MA, et al. Zika virus
and birth defects–reviewing the evidence for causality. N Engl J Med. 2016;374:1981–1987.
4. Fenton TR, Kim JH. A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants. BMC Pediatr. 2013;13:59.
8. Besnard M, Eyrolle-Guignot D, Guillemette-Artur P, et al. Congenital cerebral malformations and dysfunction in fetuses and newborns following the 2013 to 2014 Zika virus
epidemic in French Polynesia. Euro Surveill. 2016;21:1–9.