Vaccine
Volume 30, Issue 22 pp. 3249-3350 (9 May 2012)

Regular Papers
A comparison of parent and provider reported influenza vaccination status of adolescents
Original Research Article
Pages 3278-3285
Peng-jun Lu, Christina Dorell, David Yankey, Tammy A. Santibanez, James A. Singleton

Abstract
Objective
To compare parent and provider reported influenza vaccination status among adolescents.

Methods
Data from the 2009 National Immunization Survey-Teen (NIS-Teen) were analyzed. The NIS-Teen is a nationally representative random-digit-dialed telephone survey of households with adolescents 13–17 years at the time of interview, followed by a mail survey to the adolescent’s vaccination providers to obtain provider-reported vaccination histories. During the interview a parent or guardian was asked if the adolescent had received an influenza vaccination and whether their response was based upon recall only or from consulting a parent-held vaccination record (i.e., shot card) with recall of additional vaccinations not recorded on the shot card. Parent-reported influenza vaccination status was compared with provider-reported vaccination status by calculating various validity measures (sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV], and kappa), overall and stratified by several demographic characteristics. In the main analysis, provider-reported vaccinations were considered the gold standard. To evaluate the completeness of provider-reporting, we conducted additional analysis that also considered vaccinations reported by parents from the shot card or reported received in a non-medical setting as “true” vaccinations.

Results
During the 2008–2009 season, influenza vaccination coverage among adolescents based on provider report was 11.3%. Based on parent report, influenza vaccination coverage was 21.7%. Twenty-two percent of parents retrieved and referred to a shot card during the interview. In the shot card group, provider versus parent reported coverage was 12.5% versus 18.2% while among the recall only group coverage was 10.9% versus 22.7%, respectively. Overall, compared to provider report as the gold standard, parental report of influenza vaccination had a sensitivity of 86.7%, a specificity of 86.2%, a positive predictive value (PPV) of 43.1%, and a negative predictive value (NPV) of 98.0%. Among the shot card group, of vaccinations reported either by provider or by parent reading vaccination off shot card, only 66% were reported by providers. In the shot card group, the “true” vaccination level (16–17%) was closer to the parent reported coverage when it was assumed that vaccinations read by the parent from a shot card but not reported by a provider were considered true vaccinations. Overall, assuming that providers reported 64% of “true” vaccinations, sensitivity increased to 91%, specificity to 93%, and PPV to 71%.

Conclusions
Overall estimated influenza vaccination coverage was more than ten percentage points higher based on parental report than on provider report, with the difference between provider and parent report greater among the recall only group. The two estimates are closer for those with shot cards, but few parents utilized shot cards in our study and most national surveys do not ask parents to consult shot cards when responding about their adolescent’s vaccination. The actual vaccination coverage of adolescents studied is likely between coverage estimates obtained from parent report and provider report.