84 Chapter 4 Statistical analysis We integrated dates of infection and vaccination with the E-diary data, removing months in which ≥1 infection and/or vaccination date was present and months with ≤80% compliance. Baseline months were defined as 1 month (28 days) before the reported infection or vaccination date. Our primary outcome was change in number of monthly migraine days (MMD) between baseline month and one month (28 days) after the reported infection and/or vaccination date. Secondary outcome variables were change in monthly non-migrainous headache days (MHD) and monthly acute medication days (MAMD). Subgroup analyses for change in monthly triptan days and monthly analgetic days were performed. We predefined that, if a significant change in the outcome was found in the first month, we would investigate whether this change in month 1 was significantly declined in month 2. Additionally, we aimed to evaluate the effect of type of vaccination (mRNA or viral vector) on our outcome measures. Baseline patient characteristics were described using descriptive statistics. Chronic migraine was defined as ≥15 headache days per month with ≥8 migraine days per month. For each outcome variable, a linear mixed model was fitted for the baseline month, and both the first (month 1) and the second month (month 2) after infection/ vaccination. The patient was included as random effect to account for withinpatient correlation. Possible confounding factors, including age, sex, body mass index, chronic migraine, changes in preventive medication and number of previous infections or vaccinations, were treated as fixed effects. We evaluated differences at a significance level of 5%, and examined the marginal effect on MMD. The marginal effect represents the mean difference in MMD between the baseline month, month 1 and month 2, while holding all other variables constant. All analyses were performed using RStudio (R Foundation for Statistical Computing, Vienna, Austria), version 4.0.5, and the lmer package was used to fit linear mixed-effects models. RESULTS Out of 12,843 patients in the Leiden Headache Center database at the time of selection, we identified 547 clinically diagnosed migraine patients who fulfilled the initial inclusion criteria, out of whom 336 (61.4%) responded to the questionnaire. For the infection analysis, we excluded patients who did not experience any infection or had insufficient compliance in the 2 months surrounding the infection date, resulting in a representative group of n = 59 migraine patients with a combined total of 150 months of E-diary data (supplementary Figure 1). Concerning the vaccination analysis, 189 patients out of the group of 336 responders were excluded as a result of either lack
RkJQdWJsaXNoZXIy MjY0ODMw