RE: “DETECTABLE RISKS IN STUDIES OF THE FETAL BENEFITS OF MATERNAL INFLUENZA VACCINATION”
Bibliographic record
Abstract
In the August 1, 2016, issue of the Journal, Hutcheon et al. (1) nicely showed that adequately powered studies often require enormous sample sizes to demonstrate plausible benefits of maternal immunization to the fetus, and they cautioned that published studies based on smaller sample sizes may have given spurious results. The same sample size considerations also apply to studies of potential risks of maternal immunization to the fetus. Unfortunately, however, the authors did not demonstrate their own recommended caution when asserting in the introduction to their paper that maternal influenza immunization causes no apparent harm to the developing fetus (1). Although it is generally understood that very large populations are needed to detect an increase in adverse events after immunization that are rare (e.g., Guillain-Barré Syndrome, which has a risk of approximately 1 per 1 million influenza vaccine doses), it is less well recognized that when the outcome of interest is frequent among nonvaccinated individuals, large populations are then also needed to detect vaccine-associated risks that are several thousand times greater than 1 per 1 million doses (e.g., 1 per 200 or 1 per 500 doses). For example, without vaccination, the baseline frequency of low birth weight (<2,500 g) is approximately 7%, that of premature birth is 9%, and that of miscarriage is 15%. An additional risk of 1 child with a low birth weight, 1 preterm birth, or 1 miscarriage for every 200 vaccinated pregnant women is likely to be considered an unacceptable vaccine-associated risk for most mothers or obstetricians. However, this meaningful absolute increase in risk of 0.5% translates into minuscule relative risks that study investigators are required to measure (i.e., for an increase in low birth weight from 7% to 7.5%, relative risk = 1.07; for an increase in preterm birth from 9% to 9.5%, relative risk = 1.06; and for an increase in miscarriage from 15% to 15.5%, relative risk = 1.03), with very important sample size implications. As shown in Table 1, at this level of vaccine-associated increased risk, adequately powered studies (80% power and α = 0.05) would require 84,438, 105,410, and 162,300 women for baseline risks of 7%, 9%, and 15%, respectively, if 50% of mothers were vaccinated; still more participants would be needed with lower vaccine coverage. If the “acceptable” level of vaccine-associated risk were instead assumed to be lower, at 1 per 500 immunizations (i.e., an absolute increase of 0.2% and relative risks for the same pregnancy outcomes varying from 1.01 to 1.04), then the required sample sizes would increase nearly 6-fold to as much as 517,696, 649,243 and 1,006,210, respectively. These sample sizes are several times larger than those included in the largest cohort study of the safety of influenza vaccine in pregnancy to date (75,000 vaccinated and 145,000 unvaccinated pregnant women; vaccine coverage ≈35%) (2, 3). Sample Size Requirements for 80% Statistical Power and α of 0.05 Given Specified Baseline Risk, Vaccine Coverage, and Assumed Absolute Risk Increase Sample Size Requirements for 80% Statistical Power and α of 0.05 Given Specified Baseline Risk, Vaccine Coverage, and Assumed Absolute Risk Increase These sample size requirements would seriously limit the ability of available epidemiologic studies to rule out unacceptable influenza vaccine–associated risks to the fetus that are as low as 1.01- to 1.1-fold above baseline. This concept equally applies to maternal immunization for other diseases, like pertussis. Given the acknowledged bias and confounding inherent to observational studies, as reported by Vazquez-Benitez et al. (4) in the same issue of the Journal, and other methodological considerations unique to studies conducted during pregnancy (5), greater caution should therefore also be applied before asserting that maternal vaccination is safe. Conflict of interest: none declared.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.019 | 0.113 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.004 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.006 | 0.006 |
| Scholarly communication | 0.007 | 0.006 |
| Open science | 0.006 | 0.003 |
| Research integrity | 0.088 | 0.071 |
| Insufficient payload (model declined to judge) | 0.010 | 0.013 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".