Josephine van Dongen

5 Safety and Tolerability of Human Rotavirus Vaccination 135 Exploration of potential bias induced by complete case analyses Comparing infants for whom complete data were available on vaccination status to those with missing information, we only observed differences in variables that were not associated with occurrence of solicited AEs in univariate analysis (propor tion of infants from multiple bir th, median parental age and family educational level, table S5 ); we therefore conclude missing information is independent of the outcome of solicited AE and complete case analysis induced little bias in our study results. Table S5 . Infants with missing data versus complete cases Characteristic Missing (N=59) Complete (N=960) p-value Gestational age in weeks+days (mean, SD) 32+1 (2+3) 32+2 (2+4) 0.59 Bir thweight in grams (median, IQR) 1640 (825-2465) 1780 (930-2630) 0.42 Small for gestational age (yes,%) £ 25 (42.2) 288 (30.0) 0.05 Sex (male, %) 28 (47.5) 532 (55.4) 0.23 Multiple birth (yes, %) 36 (61.0) 714 (74.4) 0.02 Congenital pathology (yes, %) 3 (5.1) 82 (8.5) 0.35 Age at first vaccination in days (median, IQR) 60 (41-79) 61 (45-77) 0.77 Sibling (yes, %) 6 (13.6) 210 (21.9) 0.20 Parental education # High Medium Lower 30 (62.5) 14 (29.2) 4 (8.3) 728 (75.8) 206 (21.5) 26 (2.7) 0.03 Parental background ^ European Non-European Mixed 42 (84.0) 4 (8.0) 4 (8.0) 823 (85.7) 46 (4.8) 91 (9.5) 0.57 Average parental age in years (median, IQR) 30.5 (25.0-36.0) 33.0 (27.5-38.5) 0.03 Percentages are derived excluding subjects with missing data on the variable. Statistical significance (p-value <0.05) is highlighted in bold. £ Based on 10th percentile perinatal growth curves. # Based on highest parental educational level. ^ Based on parental background and categorized according to world population by country. Abbreviations: HRV = human rotavirus vaccine, N = number in group, SD = standard deviation, IQR = interquar tile range.