2009/H1N1 infection in pregnancy association with adverse perinatal outcomes
- Correspondence to May Li Lim
Department of Maternal Fetal Medicine, Level 7 Women's Tower, KK Women's & Children's Hospital, 100 Bukit Timah Road, Singapore 229899;
Implications for practice and research
■ 2009/H1N1 infection is associated with increased risk of perinatal death and preterm birth.
■ Infected women delivering preterm were more likely to be infected in the third trimester, be admitted into intensive care unit and have secondary pneumonia.
■ Further evaluation using larger sample size, rigorous matching of controls and consideration for variables lacking in current study will be advantageous.
Since identification of the 2009/H1N1 influenza virus, the increased risk of maternal morbidity and mortality from this novel infection has been highlighted.1 Past pandemics and seasonal flu have shown inconsistent perinatal effects. Some data have shown increased risk of preterm birth and fetal death.2 Other studies have indicated no increased fetal risk, supported by the absence of transplacental transmission of influenza virus in the umbilical cord sera of women with laboratory-confirmed flu infection.3 Studies evaluating the effect on perinatal outcomes during the 2009/H1N1 pandemic have tended to focus on maternal outcomes.
This was a case-control study conducted in the UK, aimed at assessing the risk of adverse perinatal outcomes following 2009/H1N1 flu infection in pregnancy. ‘Cases’ were pregnant women with confirmed 2009/H1N1 infection and who were admitted into hospital during the second wave of the pandemic between September 2009 and January 2010. Data were collected via a national registry (UK Obstetric Surveillance System, ‘UKOSS’). Clinicians from maternity units were requested to report all ‘cases’ from their institutions to the registry. Information relating to diagnosis, management and outcomes were obtained from clinicians. ‘Controls’ were historical, comprising women whose data were collected between February 2005 and February 2006, and had delivered before start of the 2009/H1N1 pandemic. Outcome variables assessed were preterm birth, perinatal death and low or very low birthweight. Confounding factors such as maternal age, ethnicity and socioeconomic status were adjusted for. To help evaluate the representativeness of the comparison group's data, national data from year 2008 were utilised. Statistical methods for data analysis were described. Missing data were appropriately handled using the statistical method.
There was good representation from hospitals in the UK (221 of 223 hospitals). Outcome data were available in 94% of the infected cohort. In the 2009/H1N1 group, stillbirth rate was 27 per 1000 births and perinatal mortality was 39 per 1000 births. In comparison, the uninfected cohort had a stillbirth rate of 6 per 1000 births and perinatal mortality of 7 per 1000 births. This was statistically significant. The infected group had a higher likelihood of delivery before 37 weeks gestation (adjusted OR 4.0, 2.7 to 5.9) and before 32 weeks gestation (adjusted OR 4.9, 2.4 to 10.0). The association between preterm birth and infection remained significant following exclusion of preterm deliveries for maternal indications such as secondary pneumonia and admission into intensive care unit. There was no difference in mean birthweight following adjustment for gestation at delivery. Congenital anomaly rate at birth in the infected group was 32 per 1000 births (national rate 17 per 1000 births). Admission into the intensive care unit occurred in 23% of infants born to infected mothers. There was a much larger proportion of women not immunised for H1N1 in the infected cohort.
This is an important study that highlights potential adverse perinatal effects of 2009/H1N1 infection. The use of case-control methodology has facilitated a relatively quick evaluation of a virus where knowledge has been sparse. The high percentage of complete data is noteworthy. Explicit description of criteria for hospital admission, how diagnosis of 2009/H1N1 infection was made and method of sampling would be helpful. The authors have nevertheless presented a balanced discussion.
The findings indicate that 2009/H1N1 infection in pregnant women requiring hospitalisation is associated with an increased risk of preterm delivery, stillbirth and perinatal mortality. Information relating to stillbirth such as postmortem result and classification for determining cause of stillbirth would assure against other causative factors. The possibility of seasonal flu in the control group may need considering as this may influence a magnitude of observed effects. The low rates of immunisation are noteworthy. Although this may reflect the availability of vaccine and the study has not set out to evaluate the protective effect of flu vaccine, these data may serve as an impetus to increase the immunisation effort to the public and to guide policy development.