key: cord-0717789-we6wyjaj
authors: PAPAGEORGHIOU, Aris T.; GUNIER, Robert B.; VILLAR, José
title: The link between COVID-19 and Preeclampsia
date: 2021-09-06
journal: Am J Obstet Gynecol
DOI: 10.1016/j.ajog.2021.08.047
sha: 27fb7a0fc7e789d181fb695ce55bd5a4959dfaee
doc_id: 717789
cord_uid: we6wyjaj

nan

is not a strong source of bias because including infected women in the reference group 36 would dilute, rather than strengthen, the observed association. Secondly, although it is 37 possible that preeclamptic women admitted to hospital were more likely to be diagnosed 38 with COVID-19, the study design 2 avoided such systematic bias by selecting two 39 women immediately after a diagnosed woman at the same level of care as the 40 reference group. Thirdly, the study ended in February 2021, when vaccine use in 41 pregnancy was still uncommon, i.e. case numbers would be largely unaffected. Finally, 42 adjustment by study site as a covariate and using mixed-effects models with random 43 slopes by site were conducted in the study, and the results were very similar 44 (Supplemental Table 3 ). 45

We have now undertaken further analyses, restricted to undiagnosed women who had a 46 negative PCR or antibody test result, reducing the total sample size to 1359 women. 47

The association between COVID-19 diagnosis and preeclampsia (compared to Table 2  48 in the original report) had a similar but slightly reduced risk ratio (RR) of 1.71 (95% 49 Confidence Interval (CI) 1.14 to 2.56) in the unadjusted and 1.52 (95% CI = 1.01 to 50 2.31) in the full model (adjusted for maternal age, previous parity, tobacco use during 51 pregnancy, overweight status, and history of diabetes, cardiac disease, hypertension, 52 kidney disease, or adverse pregnancy outcomes). The associations with Hypertensive 53 Disease in Pregnancy and Gestational Hypertension (GH, previously reported in Table  54 4) were also similar, with a slightly increased RR for GH: RR were 1.61 (95% CI = 1.21 55 to 2.13) and RR 1.80 (1.21 to 2.68), respectively in the unadjusted model; and 1.47 implemented to ensure enrolment of not-diagnosed women was unbiased and the data 62 have been explored for possible selection bias using several strategies. The results 63 remain largely unchanged, suggesting the association between Covid-19 and 64 preeclampsia is not due to confounding by common risk factors. 65 66 67 J o u r n a l P r e -p r o o f

from the INTERCOVID prospective longitudinal study

Preeclampsia, a risk factor to get infected with COVID or a selection 73 bias?

INTERCOVID prospective longitudinal study