Surveillance of antenatal influenza vaccination: validity of current systems and recommendations for improvement.

<bold>Background: </bold>Although influenza vaccination is recommended during pregnancy as standard of care, limited surveillance data are available for monitoring uptake. Our aim was to evaluate the validity of existing surveillance in Western Australia for measuring antenatal influenza immunisations.<bold>Methods: </bold>The self-reported vaccination status of 563 women who delivered between April and October 2013 was compared against three passive data collection sources: a state-wide antenatal influenza vaccination database maintained by the Department of Health, a public maternity hospital database, and a private health service database. Sensitivity, specificity, and positive and negative predictive values were calculated for each system using self-report as the "gold standard."<bold>Results: </bold>The state-wide antenatal vaccination database detected 45.7 % (95 % CI: 40.1-51.4 %) of influenza vaccinations, the public maternity hospital database detected 66.7 % (95 % CI: 55.1-76.9 %), and the private health service database detected 29.1 % (95 % CI: 20.5-39.4 %). Specificity exceeded 90 % and positive predictive values exceeded 80 % for each system. Sensitivity was lowest for women whose antenatal care was provided by a private obstetrician.<bold>Conclusions: </bold>Existing resources for surveillance of antenatal influenza vaccinations detect 29-67 % of vaccinations. Considering the importance of influenza immunisation as a public health intervention, particularly in pregnant women, improvements to routine monitoring of influenza vaccination is warranted. [ABSTRACT FROM AUTHOR]