Perinatal outcomes after maternal 2009/H1N1 infection: national cohort study
Cite this as: BMJ 2011;342:d3214
Rapid responses are electronic letters to the editor. They enable our users to debate issues raised in articles published on bmj.com. Although a selection of rapid responses will be included as edited readers' letters in the weekly print issue of the BMJ, their first appearance online means that they are published articles. If you need the url (web address) of an individual response, perhaps for citation purposes, simply click on the response headline and copy the url from the browser window.
Displaying 1-2 out of 2 published
We thank Professors Steinhoff, Omer and MacDonald for their interest in our paper. We read with interest their analysis showing that maternal influenza immunisation is associated with reduced likelihood of prematurity and small for gestational age births (1). This, coupled with our observations noting an increased risk of perinatal mortality and preterm delivery associated with maternal 2009/H1N1 influenza infection (2) highlights again to women, clinicians and policy makers, the importance of preventing influenza in pregnancy through immunisation.
Professors Steinhoff, Omer and MacDonald make a valid point that some of the comparison cohort we used might have been infected with seasonal influenza. We do not have information about whether any of the cohort was tested for influenza, however, none of the women included were reported to have had seasonal influenza during their pregnancies. None of these women could have had 2009/H1N1 influenza, delivering as they did in 2005-6, and therefore they do represent a valid uninfected cohort for our study assessing the perinatal outcomes of women with 2009/H1N1 infection in pregnancy. As the letter authors quite rightly point out, the fact that some women in the comparison cohort may have had seasonal influenza infection during their pregnancies would simply decrease the effect sizes we observed, compared to those which may have been seen if we had used a comparison cohort uninfected by any form of influenza. Readers should therefore interpret the associations we have documented as conservative estimates of the true impact of 2009/H1N1 influenza on perinatal outcomes. Nonetheless, treating our estimates as conservative reinforces even further the importance of prevention of infection through immunisation.
The letter authors suggest that we should reanalyse our data restricting the comparison cohort to a period when influenza viruses are not circulating. We assume this means to restrict the analysis to comparison women who deliver outside the influenza season. Such a restriction assumes that it is only influenza infection shortly before delivery which impacts on perinatal outcomes; this may possibly be the case for outcomes such as preterm birth, but is very unlikely to be the case for outcomes such as small for gestational age (SGA), which is clearly likely to develop over a longer period following infection. We are not therefore convinced of the benefits of such an additional analysis, given that the likely effect is to increase the estimates of what we have already demonstrated are significant risks.
We do not have any information on the sex of the infants born to either infected or comparison women, hence the approach we used, analysing differences in birthweight after adjusting for gestational age. Therefore we cannot make an estimate of the proportion of SGA infants based on sex- specific growth centiles. Taking this into consideration, our conservative estimate of the proportion of SGA infants in the infected cohort, is 11% (n=28), compared with 7% (n=90) in the comparison cohort (p=0.051, odds ratio 1.56, 95% confidence interval 0.96-2.47). This proportion of SGA infants in the infected cohort is not statistically significantly higher than the expected population proportion (10%) (p=0.301), but note the limited validity of this comparison due to the fact that we cannot undertake sex-specific SGA estimations.
1. Omer SB, Goodman D, Steinhoff MC, et al. Maternal influenza immunization and reduced likelihood of prematurity and small for gestational age births: a retrospective cohort study. PLoS Med 2011;8:e1000441.
2. Pierce M, Kurinczuk JJ, Spark P, Brocklehurst P, Knight M. Perinatal outcomes after maternal 2009/H1N1 infection: national cohort study. BMJ. 2011;342:d3214.
Competing interests: None declared
National Perinatal Epidemiology Unit, University of Oxford
Click to like:
We read with interest the recent paper by Pierce et al. "Perinatal outcomes after maternal 2009/H1N1 infection: national cohort study." BMJ 2011; 342: d3214. The observations they report are important because they assessed neonatal outcomes as well as maternal outcomes.
We are writing to point out a design consideration regarding their selection of the "uninfected comparison cohort". They selected comparison deliveries nonrandomly from the UKOSS database between February 2005 and February 2006. Although they referred to these deliveries as "uninfected", they were very likely exposed to the seasonal influenza epidemics in the winters of 2005 and 2006(1,2).We understand that a UK recommendation for immunization of pregnant women was not made until 2009, suggesting that few pregnant women in the earlier years would have been protected against the circulating seasonal influenza viruses.
Seasonal influenza has also been associated with adverse birth outcomes (3,4). Since women infected with seasonal influenza were included in the control comparison cohort, the effect sizes for various birth outcomes may have been underestimated by the authors. Responses to the following questions will clarify this issue: Was influenza testing done in the UKOSS cohort in 2005 - 2006? What were the characteristics of influenza epidemics in those two influenza seasons, regarding severity, increased hospitalizations in pregnancy, and perinatal outcomes? What is the effect of analyzing a truly "uninfected"cohort only, say in the period April-November, when no influenza viruses were circulating? Moreover, exclusion of any influenza virus infection should strengthen the analysis of the effect of 2009/H1N1 influenza infection.
Table 2 provides a variety of perinatal outcomes, for the infected and comparison cohorts. While mean birth weight, low and very low birth weight are displayed, we are puzzled why weight for gestational age, which provides a biological classification of birth weights related to gestational age, was not included. Since birth weights are related to gestational age, and the authors showed an increase of prematurity, it seems that a routine analysis of the proportion infants less than 10th percentile (classified as small for gestational age -SGA) using relevant reference data should have been carried out. We note that the authors' adjusted models (which did not include an assessment of proportion SGA) suggested that after adjustment, there was no effect of H1N1 on mean birth weights. However, without an analysis of SGA (that directly accounts for gestational age when assessing birth weight), and the fact that the estimates for low birth weight were statistically significant, it is difficult to accept that "after adjustment for gestational age at delivery, we find no evidence of a difference in mean birth weight between infected and comparison women, suggesting that 2009/H1N1 infection had no effect on fetal growth." This point is of particular interest to clarify as several other recent reports of the effect of epidemic or pandemic maternal influenza infection on newborn outcomes suggest that there is an effect on fetal growth manifested by a lowering of mean birth weights, and the proportion of SGA infants (3,4). Furthermore, influenza immunization appears to be protective for SGA. (5)
We agree with the authors' assessment that their data overall suggests that influenza infection in pregnancy influences both maternal and neonatal outcomes. We suggest Pierce and colleagues should consider a revised analysis specifically excluding seasonal epidemic influenza infections in the comparison cohort, as well as a direct analysis of the proportion of SGA infants in the format of table 2. If individual-level data on infection are not available, the control cohort could be restricted to a period when seasonal influenza viruses were not circulating (i.e. outside the influenza season).
1. Zhao H, Cooke M, Joseph C, et al. Influenza and other respiratory viruses surveillance in the United Kingdom: October 2004 to May 2005. CDR Weekly 2005 Supplement.
2. Desai S, Zhao H, Cooke M, et al. Surveillance of influenza and other respiratory viruses in the United Kingdom: October 2005 to May 2006. CDR Weekly 2006 Supplement.
3. McNeil SA, Dodds LA, Fell DB, et al. Effect of respiratory hospitalization during pregnancy on infant outcomes. American Journal of Obstetrics and Gynecology 2011;204:S54-S7.
4. Mendez-Figueroa H, Raker C, Anderson BL. Neonatal characteristics and outcomes of pregnancies complicated by influenza infection during the 2009 pandemic. Am J Obstet Gynecol. 2011 Jun;204(6 Suppl 1):S58-63.
5. Omer SB, Goodman D, Steinhoff MC, et al. Maternal influenza immunization and reduced likelihood of prematurity and small for gestational age births: a retrospective cohort study. PLoS Med 2011;8:e1000441.
Competing interests: None declared
Cincinnati, OH 45229, USA
Click to like: