Study population
The Norwegian Mother, Father and Child Cohort Study (MoBa) is a population-based cohort including about 280 000 participants based on written informed consent. The main aim is to understand the aetiology of complex diseases.3 Pregnant women and their partners were recruited from 1999 through 2008. Parents and children have been followed with questionnaires, registry linkages and a series of sub-studies.4 Since the end of March 2020, all active adult participants (initially about 149 000 subjects aged 25 to 65 years) have received biweekly mobile-phone questionnaires asking about symptoms related to COVID-19. Since vaccination against SARS-CoV-2 commenced in December 2020, we have included questions on vaccine uptake and side effects.
The vaccination policy in Norway was to start with the elderly population, and then to include critical health care personnel.5 The Pfizer-BioNTech (BNT162b2) mRNA vaccine was the first to be approved in December 2020 followed by the adeno-vectored vaccine from AstraZeneca (ChAdOx1 nCoV-19), and the Moderna mRNA vaccine in February 2021. The adeno-vectored vaccine was recommended for subjects below 65 years and was mainly distributed to health care personnel.
On March 11, the AstraZeneca vaccine was suspended in Norway following a report in Denmark of a death due to thrombosis after vaccination. On March 13, 2021, the Norwegian Medicines Agency was notified of blood clots and bleeding in younger people and also of severe cases of thrombosis and thrombocytopenia.6 Consequently, we included questions on bleeding episodes in the questionnaire round issued on March 17.
The response rate to the March 17 questionnaire was 58% (81267/138924). Eleven percent (n=8699) reported that they had received either an mRNA- or an adeno-vectored vaccine. The dataset was linked to the Norwegian Immunisation Registry (SYSVAK)4 using each citizen’s unique personal identification number. Registration in SYSVAK is mandatory for vaccinations against covid-19 and was confirmed for 8548 subjects in the dataset. This linked dataset defines the study population. The agreement between reported and registered information on vaccination against covid-19 was high and a discrepancy was found for only 28 subjects (kappa=0.99).
Exposure variables
The exposure variable was vaccination against covid-19. The vaccine registry holds information on date of vaccination, type of vaccine and vaccine doses for all vaccinations. 5132 subjects had received one dose of the adenovirus-vectored vaccine, while 3416 subjects had received one dose of an mRNA vaccine (3315 had received the Pfizer and 101 the Moderna vaccines). No subjects had received two doses of the adeno-vectored vaccine, and 3135 subjects had received two doses of an mRNA vaccine.
Outcome variables
The questionnaire issued to MoBa participants initially covered a standard set of side effects derived from the summary of product characteristics (SPC) for the respective COVID-19 vaccines.8,9 Bruising/skin bleeding, nose bleeding, and gingival bleeding when brushing teeth was included in the questionnaire round 26 issued on March 17 2021. These are the outcome variables. Participants were asked to check for presence for each potential side effect listed. Multiple checks were possible. If the respondents had received two doses of a vaccine, they were asked to report side effects after each dose.
Other variables
Information on occupation and chronic diseases were linked to the study population from questionnaire data collected from previous rounds during the covid-19 pandemic. Information on occupation was collected from questionnaire round 3, issued in May 2020. A list of nine choices was included and multiple checks were possible. Due to the vaccination policy, most of the vaccine recipients were health care personnel, and the variable was categorized into “health care personnel”, “other occupation” and “missing.”
Information on chronic diseases was collected from questionnaire round 2, issued in April 2020. We asked for presence of cancer, asthma, heart disease, diabetes, hypertension or other chronic diseases and multiple checks were possible. The current analyses were performed with each condition separately and combined. Since no association was found for separate conditions, “chronic disease” was categorized and reported as “any chronic disease” defined as “no” and “yes” or “missing.”
Information on having had SARS-CoV-2 confirmed in nasal /oral swabs or saliva samples was collected from the March 17 2021 questionnaire. Participants answered “no,” “yes” or “don’t know” if they had ever tested positive.
Statistical method
We calculated the prevalence of bleeding episodes and prevalence odds ratios with 95% confidence intervals, comparing bleeding episodes for the mRNA vaccines and the adenovirus-vectored vaccine. To adjust for potentially confounding variables, we used logistic regression. Analyses were performed using IBM SPSS 26.0.
Patient and Public Involvement
MoBa participants were recruited from the general population. Patients or the public were not involved in the design, or conduct, or reporting or dissemination plans of our research. However, we have interviewed focus groups consisting of participants for input to the maintenance and strategic development of the cohort.