Swine flu risks - what pregnant women need to know

The risks to pregnant women from swine flu are real. Yet many may still resist vaccination because they fear that carries an even greater risk. What do the data so far tell us?

Chris Robertson and I summarised the recent-past risk (mid April to 12 October 2009) to Scotland’s pregnant women of being hospitalized for confirmed swine flu as three times greater than the 8 hospitalisations per 1 million per month for non-pregnant women in the same age-range (16-44 years).

Sir Liam Donaldson has rightly advised that pregnant women’s unvaccinated risk from swine-flu is real. The southern hemisphere’s critical care admissions of pregnant women with confirmed H1N1 virus speak for themselves; and confirm the CMO’s advice.

First, recall that, as in Australia and New Zealand, about 1 per cent of the UK population is pregnant at any time. In Australia and New Zealand, without vaccination and during their three winter months of June to August, 66 pregnant women (that is: 26 per 100,000 pregnant women; 95 pr cent CI from 20 to 33 per 100,000 pregnant women) were admitted to intensive care units with confirmed swine flu.   

For this 3-month period, the total ICU-admission rate for confirmed H1N1v in Australia and New Zealand was 722 out of 25.2 million (or 2.9 per 100,000 of the general population). Pregnant women’s risk was therefore increased by a factor of 10 compared to the non-pregnant population, as a consequence of both age-related exposure to swine-flu and, having been exposed, the risk of severe-H1N1v-illness-requiring-critical-care-admission.
Let’s now consider how the winter season of H1N1v in Australia and New Zealand compares to UK so far. Scotland’s total ICU admission rate for confirmed swine flu in the six months from mid April to 12 October 2009 was 0.55 per 100,000 of the general population (based on 28 ICU admissions in a population of 5.1 millions; 95 per cent CI from 0.35 to 0.75).
By comparison with Australia and New Zealand, the UK - were we also to be without immunization - could face, in the next three months, a toll five times greater than to date. To cope, the southern hemisphere had 5.8 staffed beds equipped for mechanical ventilation per 100,000 of population and, during their winter peak, 10 per cent of these ICU-beds were accounted for by patients with swine-flu (NEJM 2009: ANZIC investigators).
Thus, as the Chief Medical Officer has advised, the severe-H1N1v-illness-risk to pregnant women is real. In the next three months, without immunization, we could find ourselves with at least five times as many pregnant women being admitted to intensive care units as we have had in the six months to date.
Over the next three months, the prospect that 20 to 30 per 100,000 unvaccinated pregnant women may need ICU-admission for confirmed swine-flu is a realistic prior estimate of their risk. Actual risk that can, and of course should, be formally measured (during mid October to mid January; and in subsequent three months) by recording the proportion, if not the identities, of pregnant women who decline immunisation - together with the immunisation status of pregnant women who are admitted to intensive care units with confirmed swine-flu.
So women's concerns about an unmeasured, unknown risk that might be associated with the new vaccines should be put into perspective by these real risks of H1N1v infection. Any serious vaccine-related harm that came close to the ICU-admission rate for unvaccinated pregnant women over the next three months would have to occur with a frequency of 2 to 3 per 10,000 vaccinees. Serious harm of that frequency, were it present, would be almost surely be evident in the first 30,000 vaccinees. Internationally, the number of recipients of the new vaccine formulations - whether with H5N1 or H1N1 on board - may already have passed this 30,000 threshold (in total, or even for pregnant women). It would be reassuring to know this.
Finally, pregnant women may also want to take account of some early data from the US suggesting that the swine-flu-related hospitalisation risk for pregnant women depends on trimester: of 17 hospitalisations of pregnant women for confirmed H1N1v, 12 were in the third trimester of pregnancy. Data on trimester of pregnancy are currently being checked for the 66 pregnant women with confirmed swine-flu who were admitted to ICUs during the winter months in Australia and New Zealand. The investigators’ informal impression is that they were weighted towards third trimester. So important are these data that their formal publication is essential.
Equally important, however, is that UK properly marshals its own past and prospective data on (the trimester of) pregnant women admitted to hospital, and especially to intensive care units, for confirmed swine-flu, as the Australian and New Zealand have demonstrated can readily be done.