1383 - Characterizing Real-world Patterns of Early Childhood Vaccination

Background: Vaccine hesitancy is increasingly common, but more information is needed on patterns of childhood vaccination. We characterized patterns of vaccine delay among commercially-insured children in the U.S.

Methods: Using the IBM MarketScan Commercial Database, we identified infants who received a timely first dose of diphtheria-tetanus-acellular pertussis (DTaP) vaccine from October 2009 to June 2017. We used CPT codes to collect vaccine administration history on antigen, formulation, dose, and date. We ascertained injectable and oral vaccine antigens (DTaP, polio, pneumococcal conjugate, rotavirus, Haemophilus influenza type b (Hib), measles, mumps, rubella, varicella). Timely receipt was defined as concomitant administration of the CDC-recommended number of antigens during the following time windows: 2, 4, 6, and 12-15 months of age (grace period: -7, +21 days). We generated heat maps to illustrate age distributions at receipt of specific antigens and doses. We created Sankey diagrams to illustrate the number of antigens received concomitantly during each time window and depict transitions to different states over time (e.g., no vaccine delay to vaccine delay). For each antigen and dose, we estimated the cumulative incidence of receipt.

Results: Among 1,066,216 eligible infants, the majority (84%) concomitantly received all 5 CDC-recommended antigens at 2 months of age while others only received 1 (1%), 2 (2%), 3 (4%) or 4 (9%) antigens. Many vaccinations were delayed – 30% and 39% of children did not receive all recommended antigens concomitantly at 4 and 6 months, respectively. The heat map shows wide variation in age at vaccination. For several antigens including Hib, measles, mumps, rotavirus, rubella, and varicella, the cumulative incidence increased steeply at ≥2 time points, suggesting vaccine delay for some infants (e.g., the first dose of Hib was administered to 85% of infants by 2 months of age, with subsequent small but distinct increases at 4, 6, 12, and 15 months of age).

Conclusion: Using real-world data to study early childhood vaccination patterns, we observed evidence of substantial deviation from the CDC-recommended schedule. These results expand current knowledge on the magnitude and timing of antigen- and dose-specific vaccine delay on a population level.