UK Vaccines & Pregnancy: Updating the Update

In my post from yesterday,

Vaccines & Pregnancy in the UK- An Update

I estimated that in the month of December in the UK that unvaccinated women had a stillbirth rate of 3.89 per 1,000 while vaccinated women had a rate of 4.30 per 1,000 (11% higher). I also reference my posts from January that argued vaccinated women should have a rate ~33% lower than unvaccinated women after correcting for age & time of vaccination in pregnancy. A short update post then explained that the 33% number may be inaccurate for December because women giving birth in December would likely have been vaccinated earlier in pregnancy. The remainder of this post tries to answer the question of what should we have expected the relative rates to have been in December for unvaccinated vs. vaccinated women.

This estimate will include three factors that I believe must be considered:

1. When in pregnancy a vaccinated woman was vaccinated

2. Age of vaccinated women vs. unvaccinated women

3. Socio-economic status of vaccinated women vs. unvaccinated women (Note: this was not included in my previous posts when creating an estimate)

Timing of vaccination in pregnancy

From week 11 UKHSA report:

From week 13 UKHSA report:

By subtracting the figures in the former from the figures in the latter, we can calculate the numbers for vaccinated women giving birth in December:

• 29,544 women vaccinated where we know when they were vaccinated.

• 6,712 before pregnancy (22.7%)

• 6,542 in 1st trimester (22.1%)

• 9,448 in 2nd trimester (32.0%)

• 6,842 in 3rd trimester (23.2%)

This post from January explained the methodology for now estimating the impact of this timing of vaccination, I will not rehash it here. Here are the numbers run for December births given that 76.8% of vaccination occurred before the 3rd trimester:

Intuition would lead me to believe that reality is probably somewhere between Scenario 1 (where we assume all 3rd trimester vaccinations occur in the 1st week of the trimester) and Scenario 3 (where we assume 3rd trimester vaccinations are evenly distributed in the trimester). I believe this because given that women giving birth in December were more likely to have been vaccinated earlier in pregnancy, it would reason they were more likely vaccinated early 3rd trimester than late 3rd trimester. This would then argue that the expected reduction in stillbirth rate among vaccinated women, based on time of vaccination, is between 6% (Scenario 1) and 13% (Scenario 3). I propose we use an estimate of 10% reduction… it is roughly in the middle, a nice round number, and also I have not included an estimate of the women who were vaccinated in weeks 24-26 (2nd trimester) who also should have had a reduced risk of stillbirth, leading to a larger estimate of reduced rate. Estimate: Based on timing of vaccination, all else equal, vaccinated women giving birth in December should have had stillbirths at a rate ~10% lower than unvaccinated.

Age of vaccinated women vs. unvaccinated women

Estimating the impact of age requires combining two sets of data. First, we have information on the age profile by vaccination status from the UKHSA report:

Next, from ONS birth characteristics data we have information on how mother’s age impacts risk of stillborn births:

Some spreadsheet work, then calculating weighted averages yields the following:

We see that based on age, all else being equal, we would expect vaccinated women to have higher rates of stillborn births because vaccinated women tended to be older. Estimate: Based on age of mother and vaccination status, all else equal, vaccinated women giving birth in December should have had stillbirths at a rate ~2% higher than unvaccinated.

Socio-economic status of vaccinated vs. unvaccinated women

While vaccinated women tend to be older, they also tend to be less deprived than unvaccinated women. To estimate the impact, we will do something very similar to the previous section, combining data from the UKHSA report with data from the ONS birth characteristics data.

From the UKHSA report:

From the ONS birth characteristics data we have estimate of impact of deprivation level of mother on stillbirths:

After mapping the 5 level index of the UKHSA report to the 10 level index of the ONS data and doing some spreadsheet work, we have the following:

Because vaccinated women tend to be much less deprived than unvaccinated women, all else being equal, we would expect lower stillbirth rates among the vaccinated (3.60 vs. 3.97). Estimate: Based on deprivation level of the mother and vaccination status, all else equal, vaccinated women giving birth in December should have had stillbirths at a rate ~10% lower than unvaccinated.

Summary

1. When in pregnancy a vaccinated woman was vaccinated → Vaccinated women giving birth in December should have stillbirths ~10% less often

2. Age of vaccinated women vs. unvaccinated women → Vaccinated women giving birth in December should have stillbirths ~2% more often

3. Socio-economic status of vaccinated women vs. unvaccinated women → Vaccinated women giving birth in December should have stillbirths ~10% less often

Combining the three factors, my best estimate is vaccinated women giving birth in December should have a stillbirth rate ~18% lower than unvaccinated women. As shown in my post from yesterday, it would appear that they actually had stillbirths in December at a rate of ~11% higher than the unvaccinated. My hope is that someone with full access to all the data is looking into this as it seems alarming based on my best effort of analyzing the data available to me.

Addendum

Would like to point out something quickly- In the above, you will see that if I just attempted to predict stillbirth rates among the unvaccinated based on age, I would have come up with 3.88 stillbirths/1,000. If I had tried predict that based on deprivation level, we have 3.97. The fact that the UKHSA published the rate of 3.90 for unvaccinated women, firmly between these two numbers, is support that this methodology is sound. Therefore, even if we ignored section 1 above and only based our estimate for vaccinated women based on age and deprivation level, their rate should be somewhere between 3.60 and 3.96… the 4.30 we see in December would seem noteworthy.