Poor pregnancy outcomes associated with maternal infection with the A(H1N1) 2009 virus during the pandemic – findings from a European cohort studyArchived

Perinatal outcomes after maternal 2009/H1N1 infection: national cohort study Pierce M, Kurinczuk JJ, Spark P, Brocklehurst P, Knight M on behalf of UKOSS .BMJ 2011; 342:d3214

Accompanying editorial: H1N1 influenza in pregnant womenJoseph KS, Liston RM BMJ 2011; 342; d3237

The outcomes of pregnancy for a cohort of pregnant women admitted to UK hospitals with influenza A(H1N1) 2009 infection were compared with those of an uninfected cohort of women who delivered in the UK before the start of the 2009 pandemic. The settings were over 200 hospitals with obstetrician led maternity units in the United Kingdom (UK). The exposed cohort were pregnant women admitted to hospital with laboratory confirmed influenza A(H1N1) 2009 infection between the 1st September 2009 and the 31st January 2010. The cohort was identified prospectively through the ‘UK Obstetric Surveillance System’ (UKOSS).This is a well-established national network of collaborating clinicians established to provide a routine system for conducting parallel natural studies of severe and uncommon complications of pregnancy throughout the UK.(1)  Clinicians were asked to report all pregnant women with confirmed influenza A(H1N1) 2009 infection admitted to their unit through a web-based reporting system. This was active surveillance as units return “nil reports”, confirming by doing so the denominator population for the study.  If a woman’s pregnancy was continuing at the time of discharge a copy of the initial data collection form was sent to the reporting clinician two weeks after the expected date of delivery to obtain information on the outcome of the pregnancy. The comparison cohort data had been previously collected between February 2005 and February 2006 using the same UKOSS methodology for women who delivered in UK hospitals between those dates and who therefore represented a cohort that could not be infected with A(H1N1)2009.

Outcome data were reported for 256 (94%) infected women in the cohort. These women had 249 live births, including five pairs of twins. In addition, five pregnancies were lost or terminated before 24 weeks’ gestation. Perinatal mortality was significantly higher in infants born to infected women (10 deaths among 256 infants; rate 39 per 1000 births) than in infants of uninfected women (9 deaths among 1233 infants; rate 7 per 1000 births). After statistical analysis allowing for known confounders the odds ratio with 95% confidence interval was 5.7 (2.2 to 15.1).  This was principally explained by an increase in the rate of stillbirth (27 per 1000 births versus 6 per 1000 births – odds ratio 4.2 [1.2 to 12.4]). The stillborn infants were delivered a median of 12 days after the onset of symptoms of influenza. Women with influenza A(H1N1) 2009 infection were more likely to deliver before 37 weeks’ gestation (adjusted odds ratio: 4.0 (2.7 to 5.9) and before 32 weeks’ gestation (adjusted odds ratio: 4.9 [2.4 to 10]) than were uninfected women. Infected women who delivered preterm were more likely to be infected in their third trimester, to have been admitted to an intensive care unit and to have a secondary pneumonia than were those who delivered at term. The risk of preterm birth associated with A(H1N1) 2009 infection persisted even after the role of secondary pneumonia was allowed for, suggesting that the excess risk could not be explained by pneumonia. After adjusting for gestational age at delivery, the authors found no evidence of a difference in mean birth weight between infected and comparison women, suggesting that influenza A(H1N1) 2009 infection had no effect on foetal growth. Though eight infants born to women after influenza A(H1N1) 2009 infection had congenital anomalies diagnosed at birth, representing a birth prevalence of 32 per 1000 total births, compared with a national rate of 17 per 1000 total births, a difference that did not quite reach statistical significance.(p=0.08) A sensitivity analysis assuming an optimal outcome in the small proportion of pregnancies where the outcome was unknown (6% of the mothers were lost to follow up) did not change the findings significantly. The paper contains a useful table of similar studies from the 2009 pandemic though only one of these was from Europe and its loss to follow-up was substantial.

The authors highlighted limitations of their study. First, perinatal mortality remains a rare outcome, and the statistical uncertainty that surrounds this estimated increased mortality risk should be noted. Second, the outcomes they presented are likely to represent the severe end of the disease spectrum as, from previous estimates in the UK, up to 30 women presented with influenza-like illness in pregnancy for each one admitted to hospital with confirmed influenza A(H1N1) 2009 infection.(1) Thirdly the numbers of outcomes was small thus the results must be interpreted with caution. Fourthly confounding for other perhaps social factors that may have caused a negative outcome could not be excluded. Though notably there was no over-representation of South Asian ethnic minority groups in the infected women. Association of severe influenza disease due to A(H1N1) 2009 had been observed in two other studies of the 2009 pandemic in the UK and might have explained the finding.(2,3)  The study was not designed to elucidate whether or not immunisation of pregnant women or giving antiviral medications was protective.  Only 6% of the cohort were immunised consistent with late availability of vaccine in Europe.(4)  More than half the infected women were treated influenza antivirals before within two days of infection but this was not associated with a reduced likelihood of pre-term delivery. Notwithstanding a few limitations, this study indicated a substantial increase in the risk of poor outcomes of pregnancy in women infected with the influenza A(H1N1) 2009 virus.

ECDC Comment (15th July 2011):This is an important European study.  Testing for influenza in hospitalised patients in Europe was uncommon before the 2009 pandemic. Hence the fact that there are very few published papers on the broad topic of influenza infection in pregnant women in Europe has been difficult to interpret in the light of outputs from North America even before the 2009 pandemic.(5,6)  Was there not much infection in pregnant women in Europe or were people not looking for it There are some indications of differences in experience between Europe and the Americas where severe disease in pregnant women due to A(H1N1)2009 seems to have been more common .(1,5,6)   Also the very few published European studies that exist have usually not included perinatal outcomes and have not always achieved such good follow-up as is the case here.(1,7)   It is important not to overstate the risk from influenza to pregnant women, hospitalization is an uncommon outcome in pregnancy (it is estimated that for every pregnant woman hospitalised by influenza A(H1N1)2009 there are thirty symptomatic infections that do not need such care(1).  However this study emphasises the threat to the unborn child.

The authors mention the difference between the pandemic and seasonal influenza implying that seasonal influenza may be less threatening to pregnancy. But seasonal influenza is changed by each pandemic.(8)  In Europe there has been a predominance of A(H1N1)2009 since the pandemic in the first post pandemic winter with higher levels of deaths than in the pandemic in one country.(9)  A dominance of A(H1N1) 2009 is not inevitable. Influenza A(H3N2) predominated as the A viruses in North America in 2010-11. But certainly pre-2009 studies may be less valid for the future. This is of particular interest to ECDC as it and the research organisation PALLAS are undertaking a systematic review of risk factors for severe influenza (including the impact on pregnancy). This is for the work ECDC has to do supporting the European Commission and Member States in implementing the 2009 EU Health Council Recommendation on seasonal influenza immunisation.(10).  The World Health Organisation is also pulling together data ahead of a review of the 2005 World Health Assembly recommendation in the same area.(11)   What is not looked at is the impact of maternal immunisation on early childhood and whether there is protection of the newborn beyond the perinatal period.  Little impact would be expected here as in Europe transmission of influenza declined to almost nothing in the late winter of 2009/10 and did not rise again until the late autumn of 2010. There are studies that have looked at the outcome on the newborn with the strongest being a randomized controlled trial  of seasonal influenza immunisation in Bangladesh (where influenza transmission is more continuous than seasonal) which was found to be also beneficial for the newborn child’s health.(12) 

Which leads onto the “so… what next” question – should pregnant women be routinely immunised   An accompanying editorial makes a strong case for routinely immunising pregnant women at least against A(H1N1)2009.(5)  The UK paper makes a distinction between pandemic and seasonal influenza.  As mentioned above that distinction was non-existent last season and seasonal influenza may be changing.  In the pandemic  many countries followed  WHO SAGE guidance and immunised pregnant women. (13)  Data on the safety of seasonal vaccine in pregnant women is encouraging.(14)  Some obstetricians and authorities avoid immunising in the first trimester because of the possible coincidence of immunisation with miscarriage that were going to happen anyway. That may not be pragmatic in midst of influenza transmission but sensible at other times of year. However certainly this paper strengthens the evidence for offering immunisation to pregnant women in Europe.