Microcephaly in Brazil: how to interpret reported numbers?

Publication History

Brazil is facing its first outbreak of Zika virus, particularly in the northeast region. Most cases of Zika virus infection are self-limited and without sequelae, but there have been clusters of cases of microcephaly in some areas of known Zika virus transmission. Although strongly suspected, the causal relation between in-utero exposure to Zika and microcephaly is yet to be established. The increased number of microcephaly cases in Brazil has led to a high level of concern among pregnant women throughout the country. On Feb 1, 2016, WHO’s International Health Regulations Emergency Committee advised that the clusters of microcephaly and other neurological disorders and their possible association with Zika virus constitutes a Public Health Emergency of International Concern.1

Measurement of newborn head circumference is useful as a screening tool for detecting microcephaly, independently of its cause. Before 2015, the annual numbers of reported cases of microcephaly in Brazil were consistently below 200.2 Between mid-2015 and Jan 30, 2016, 4783 suspected cases of microcephaly were reported, including newborn and fetal losses.3 Of these, 1103 cases have completed clinical, laboratory, and imaging examinations, and 404 (36·2%) were classified as confirmed cases of microcephaly. Among the confirmed cases, brain abnormalities were detected by imaging in 387 babies and Zika virus was detected in 17 babies, including in two fetal losses.3 The remaining 709 cases were discarded and 3670 suspected cases of microcephaly remain under investigation.3 Although 36·2% seems to be a high rate of true positives, it has to be interpreted with caution because in the present situation newborn babies with visible cranial deformities are likely to be fast-tracked for in-depth examination. This temporal increase in suspected cases of microcephaly could also be distorted given both raised awareness, with more children than usual being measured and reported, and changing definitions of microcephaly over time. The possibility of over-reporting and misdiagnosis was recently raised by the Latin American Network of Congenital Malformations,4 and their report led to speculation in the international scientific press on the magnitude of the increase in microcephaly cases.5

To help interpret these numbers, it is instructive to assess how head circumference criteria for defining suspected cases of microcephaly have evolved (table). Before Dec 8, 2015, Brazil’s Ministry of Health6 recommended a cutoff for head circumference of less than or equal to 33 cm for term newborn babies (both sexes and all gestational ages); for preterm babies, the cutoff was the 3rd centile of the Fenton7 curves of head circumference by gestational age and sex. On Dec 8, 2015, the Ministry of Health in Brazil revised the case definition for suspected microcephaly in newborn babies and reduced the head circumference criterion in term newborn babies to less than or equal to 32 cm.8 On Jan 21, 2016, the Pan American Health Organization (PAHO)9 proposed the use of fixed cutoffs of 32·0 cm and 31·6 cm for term boys and girls, on the basis of the 3rd percentile for term newborn babies of any gestational age according to the WHO Growth Standards.10 For preterm babies, PAHO9 recommended use of the 3rd centile of either the Fenton7 or InterGrowth11 curves. Use of fixed cutoff head circumference measurements for all term infants is inappropriate since it does not account for the fact that 68·1% of term newborn babies in Brazil are below 40 weeks’ gestational age, partly owing to the fact the country has the highest caesarean section rate in the world.12


Preliminary estimates of the specificity, sensitivity, and number of suspected cases of microcephaly in Brazil according to different screening criteria
Table Thumbnail. Opens Table in new tab.
*Based on applying the InterGrowth standards to the distribution of livebirths by gestational age in Brazil.
Preliminary results based on a case series of 31 newborn babies with radiological evidence of brain abnormalities.
Calculated on the basis of sensitivity and the gestational age distribution of Brazilian newborn babies.

The reasons for choosing the Fenton7 reference are unclear. The Fenton chart is based on a meta-analysis of six pre-existing studies from high-income countries with non-standardised methods, as is the case for most neonatal anthropometric charts.13 Instead, we applied the sex-specific and gestational-age specific InterGrowth standards11 to the gestational age distribution of Brazilian infants in 201214 to estimate the specificity of evolving definitions of suspected cases of microcephaly (table). InterGrowth11 is a prescriptive standard for fetal and newborn growth based on healthy gestations from eight countries. In this prospective, multicentre study, women had a reliable ultrasound estimate of gestational age using crown–rump length before 14 weeks of gestation or biparietal diameter if antenatal care started between 14 weeks and 24 weeks or less of gestation.11 Newborn anthropometric measures were obtained by identically trained anthropometric teams using the same equipment at all sites, which included Brazil. The InterGrowth study was designed to be fully consistent with the WHO Growth Standards,15 which are used throughout the world, thus providing a comparable standard for fetal growth and newborn size. Both standards are perfectly matched for term newborn babies. Because microcephaly cases were excluded from the InterGrowth samples, the distribution of head circumferences in the standard is appropriate for estimating specificity of a given cutoff.

The use of a cutoff of −3 SD below the mean value for newborn head circumference was proposed by a 2013 systematic review of microcephaly, on the basis of the finding that most newborn babies with head circumferences between −2 SD and −3 SD below the mean range do not have any evidence of malformation.16 The −3 SD cutoff has been used traditionally by the Latin American Collaborative Study of Congenital Malformations.4 A WHO manual on surveillance of birth defects defines microcephaly as a head circumference below −2 SD of sex-specific and gestational-age specific curves, but only if accompanied by structural abnormalities of the brain.17 The WHO manual explicitly states that absence of such abnormalities rules out a diagnosis; it does not recommend diagnostic workouts for all children with small heads.17

We derived preliminary estimates of sensitivity from 31 cases of confirmed microcephaly from ten states in Brazil, including eight northeastern states.18 The newborn babies had head circumferences below the Ministry of Health cutoff at the time of birth and had brain abnormalities confirmed by imaging that were compatible with congenital infection (mainly brain calcifications, lissencephaly, and ventriculomegaly). These are the first consecutive confirmed cases available to our coauthors who are involved in surveillance in the ten states. In these 31 babies with microcephaly there was no evidence of intrauterine exposure to other infectious diseases, such as syphilis, toxoplasmosis, rubella, cytomegalovirus, and herpesvirus. The average head circumference was 28·4 cm (SD 2·3), with a mean InterGrowth Z score of −3·5 (SD 1·4). Only three newborn babies had head circumferences greater than or equal to 32 cm, all with mild radiological signs of brain abnormalities.

The table shows the sensitivity of the different diagnostic criteria for suspected microcephaly based on our case series. These preliminary results must be interpreted with caution given the small number of cases, the fact that most measurements were rounded to full cm instead of mm, and the possibility of notification bias. Although the specificity of the cutoffs used so far seem high, in a country with almost 2·9 million annual births there would still be many newborn babies classified as suspected cases who would require in-depth investigation. The table shows that the number of cases of suspected microcephaly range from more than 600 000 with the initial criterion from the Ministry of Health6 to just over 3000 if −3 SD is used as the cutoff. The fact that 3670 of the 4783 suspected cases of microcephaly identified in the past months are still under investigation is not surprising given delays in obtaining access to advanced diagnostic facilities in Brazil’s national health system.19 This backlog of cases under investigation may become worse as increasing numbers of suspected cases are reported and babies’ head circumferences are measured more often than in the past. Another consideration is the emotional stress for parents whose healthy babies with small heads are incorrectly screened as positive.

There is a trade-off between specificity and sensitivity. Since there is no effective treatment for congenital microcephaly, there is a strong argument to prioritise specificity over sensitivity. Increasing specificity would reduce the iatrogenic potential of radiation during brain tomography, which is 100 times higher than that for a chest x-ray,20 and would help alleviate the emotional effects on parents of healthy children who are given a false-positive result in the screening assessment. This approach would also reduce the burden and costs to an already overstretched health system. Although true cases of microcephaly that are missed by a less sensitive cutoff could benefit from early intellectual stimulation, there would still be opportunities to detect their condition later during infancy.

Arguments for prioritising sensitivity include the fact that we are in the early stages of a new epidemic about which little is known. It is also conceivable that babies with microcephaly who would profit most from early intellectual stimulation could be exactly those whose head circumference is closer to the normal range.

On the basis of our results (table), we recommend use of a consistent set of diagnostic criteria for suspected microcephaly that take into account gestational age for term and preterm newborn babies; such criteria are provided by the InterGrowth standards.11 These are preliminary recommendations that can be revised once a larger case series is accrued. The sensitivity of a cutoff for head circumference of −3 SD seems to be too low, particularly during what seems to be a new epidemic of microcephaly, when one does not want to miss many cases. We favour a cutoff of −2 SD, which has similar sensitivity to the current Ministry of Health8 and PAHO9 recommendations, with the advantage of greatly reducing the number of newborn babies who will need investigation. Although −2 SD and the 3rd percentile seem to be close, the latter cutoff would classify an additional 0·7% of the population as suspected cases, or about 20 000 Brazilian newborn babies per year. The recent availability of a computer and mobile-phone based application for the estimation of head circumference Z scores by gestational age and sex will contribute to the field implementation of the InterGrowth standards;21 Portuguese and Spanish versions have also been made available.

Although there is evidence of an increased number of cases of microcephaly in Brazil, we show that the number of suspected cases relied on a screening test that had very low specificity and therefore overestimated the actual number of cases by including mostly normal children with small heads. We recommend that national and international agencies should refrain from reporting suspected cases and speed up investigation to report on confirmed cases with laboratory or radiological evidence. It is also important that health workers measure head circumference in all newborn babies using standardised anthropometric techniques, and report results in mm. The present situation in Brazil is certainly a severe public health challenge. Better measurement and use of the appropriate growth standards are essential for the continued surveillance of microcephaly cases that are potentially associated with the Zika virus infection.

CGV and FCB were part of the team of researchers that created the InterGrowth standards; the standards are in public domain and the researchers involved in their creation do not receive any financial benefits from their use.


  1. WHO. WHO statement on the first meeting of the International Health Regulations (2005) (IHR 2005) Emergency Committee on Zika virus and observed increase in neurological disorders and neonatal malformations. http://www.who.int/mediacentre/news/statements/2016/1st-emergency-committee-zika/en/; Feb 1, 2016. ((accessed Feb 3, 2016).)
  2. Brasil Ministério da Saúde. DATASUS. Tabulação de anomalia ou defeito congênito em nascidos vivos (in Portuguese). http://www2.aids.gov.br/cgi/tabcgi.exe?caumul/anoma.def. ((accessed Feb 4, 2016).)
  3. Brasil Ministério da Saúde. Centro de Operações de Emergências em Saúde Pública sobre Microcefalias. Monitoramento de Casos de Microcefalia no Brasil. Informe epidemiológico 11—semana epidemiológica 04/2016 (24 a 30/01/2016). Ministério da Saúde, Brasília; 2016 ((in Portuguese).)
  4. Latin American Network of Congenital Malformations. Microcefalias en el ECLAMC y en Brasil (in Portuguese). http://www.eclamc.org/microcefaliaarchivos.php; Dec 5, 2015. ((accessed Feb 2, 2016).)
  5. Butler, D. Brazil’s surge in small-headed babies questioned by report. Nature. 2016; 530: 13–14
  6. Brasil Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância Note Informativa 01/2015. COES Microcefalias. Ministério da Saúde, Brasília; 2015 ((in Portuguese).)
  7. Fenton, TR and Kim, JH. A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants. BMC Pediatr. 2013; 13: 59
  8. Brasil Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância das Doenças Transmissíveis. Protocolo de vigilância e resposta àocorrência de microcefalia relacionada à infecção pelo vírus Zika. Ministério da Saúde, Brasília; 2015 ((in Portuguese).)
  9. Pan American Health Organization. Lineamientos preliminares de vigilancia de microcefalia en recién nacidos en entornos con riesgo de circulación de virus Zika. Jan 22, 2016. Pan American Health Organization, Washington, DC; 2016
  10. WHO. The WHO child growth standards. www.who.int/childgrowth/en/. ((accessed Feb 4, 2016).)
  11. Villar, J, Cheikh Ismail, L, Victora, CG…, and for the International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21st). International standards for newborn weight, length, and head circumference by gestational age and sex: the Newborn Cross-Sectional Study of the INTERGROWTH-21st Project. Lancet. 2014; 384: 857–868
  12. Gibbons, L, Belizan, JM, Lauer, JA, Betran, AP, Merialdi, M, and Althabe, F. Inequities in the use of cesarean section deliveries in the world. Am J Obstet Gynecol. 2012; 206: e1–19
  13. Giuliani, F, Ohuma, E, Spada, E et al. Systematic review of the methodological quality of studies designed to create neonatal anthropometric charts. Acta Paediatr. 2015; 104: 987–996
  14. Brasil Ministerio da Saude. DATASUS. Sistema Nacional de Nascidos Vivos (in Portuguese).http://tabnet.datasus.gov.br/cgi/deftohtm.exe?sinasc/cnv/nvuf.def; 2016. ((accessed Feb 2, 2016).)
  15. Villar, J, Papageorghiou, AT, Pang, R et al. Monitoring human growth and development: a continuum from the womb to the classroom. Am J Obstet Gynecol. 2015; 213: 494–499
  16. Woods, CG and Parker, A. Investigating microcephaly. Arch Dis Child. 2013; 98: 707–713
  17. WHO. Birth defects surveillance: a manual for programme managers. World Health Organization,Geneva; 2014
  18. Schuler-Faccini, L, Ribeiro, EM, Feitosa, IM…, and Brazilian Medical Genetics Society–Zika Embryopathy Task Force. Possible association between Zika virus infection and microcephaly—Brazil, 2015. MMWR Morb Mortal Wkly Rep. 2016; 65: 59–62
  19. Solla, J and Chioro, A. Atenção ambulatorial especializada. in: L Giovanella, S Escorel, LVC Lobato, JC Noronha, AI Carvalho (Eds.) Políticas e sistema de saúde no Brasil. Editora Fiocruz, Rio de Janeiro; 2008: 627–663 ((in Portuguese).)
  20. Davies, HE, Wathen, CG, and Gleeson, FV. Risks of exposure to radiological imaging and how to minimize them. BMJ. 2011; 342: 589–593
  21. InterGrowth21st. Oxford University, Oxford;2016http://intergrowth21.ndog.ox.ac.uk/en/ManualEntry. ((accessed Feb 4, 2016).)
Posted in AYNLA, Uncategorized.