Various Measures of the Effectiveness of 2009 H1N1 Influenza Vaccination Among Children Vaccinated at School During the 2009 Influenza Pandemic In Maine

Background: Unlike many other areas in the United States, the state of Maine, through school-located vaccination (SLV) campaigns, was able to distribute H1N1 vaccine before or concurrent with circulation of the 2009 pandemic influenza A/H1N1 virus (pH1N1), providing an opportunity to estimate the effectiveness of H1N1 vaccine among school-aged children.

Objectives: We assessed single-dose H1N1 vaccine effectiveness (VE) among school children during the 2009 influenza pandemic using various outcome measures.

Methods: Individual-level data on student attendance and school vaccination records were obtained from four counties in Maine, October-December 2009. We estimated VE based on differences in single-day and multiple-day absenteeism rates between vaccinated and non-vaccinated students (grades K-8).  In addition, a case-control study was performed using only Cumberland County data.  We interviewed parents of students with laboratory-confirmed pH1N1, reported to the state health department, and their classmates to identify potential cases and controls based on school attendance, date and cause of absenteeism, and episodes of acute respiratory illness (ARI).  VE, with adjustments for selected potential confounders, was estimated using logistic regression for three case definitions: laboratory-confirmed (pH1N1), probable (laboratory-confirmed plus physician-diagnosed cases), and suspected (probable plus ARI).

Results: In total, 93 schools (89%) agreed to participate.  Average school vaccination coverage was 55% (range 17%-83%).  VE against absenteeism was 13% (95% CI: -32, 53%) for single-day absences and 38% (95% CI: -90, 100) for 4-day absences.  Fifty-nine schools in Cumberland County were included in the case-control analysis; 1,206 students were interviewed.  Preliminary estimates of adjusted VE against illness ranged from 19% to 69%. VE based on laboratory-confirmed cases only was 69% (95% CI: 12, 89%).

Conclusions: Our results suggest that SLV helped reducing influenza illness and related student absenteeism in Maine.  VE estimates in our evaluation varied widely; more specific outcomes to define influenza illness provide more accurate VE estimates.