MCV4 coverage during the first ten months after its licensure (March-December 2005) was about 11% among targeted adolescents. Non-targeted adolescents (13 and 16 year olds) received a sizeable portion (27%) of the total administered doses. Of the five MCOs, the two largest reported vaccine shortages and had the lowest coverage. The surge in MCV4 uptake during the summer months coupled with the uptake among non-targeted adolescents likely contributed to the vaccine ordering limits and reported shortages. Seasonal uptake patterns of MPSV4 and Td (prior to MCV4 availability) show a similar increase in uptake during the summer, though the month of the peak uptake varied.
Prior to the availability of national coverage data,[14
] most vaccination coverage information for adolescents came from cross-sectional or before-and-after studies designed to evaluate the impact of immunization requirements, voluntary school-based programs, and intervention designed to increase the delivery of clinical preventive services [15
]. In general, these studies have revealed only limited uptake of newly recommended vaccines for adolescents before the implementation of interventions that were being studied. Other studies of vaccine uptake in adolescents have shown similarly modest uptake[8
]. The first nationally representative physician-validated adolescent immunization survey (National Immunization Survey-Teen) found that MCV4 coverage among US adolescents 13-17 years was 12% during 2006[14
]. Our results demonstrate that the VSD can be an important data source for timely evaluation of new or expanded vaccine recommendations given their use of a real-time surveillance system[12
Vaccine uptake among adolescents appears to have a seasonal pattern. MCV4 uptake started increasing in May and peaked in July. Examination of MPSV4 and Td also shows increases in uptake during summer months. The earlier peak for MPSV4 suggests that different factors may play a role; however, given the small number of MPSV4 vaccinations (compared to Td and MCV4), it is difficult to hypothesize possible reasons. It is possible that MCV4 would have peaked in August (before the start of school), like Td, if not for the vaccine limits placed by the manufacturer (in July). Michigan's vaccine registry data showed a similar surge in MCV4 uptake during summer months suggesting that this seasonal increase is not limited to adolescents enrolled in managed care [21
]. The GBS investigation was announced on September 30, 2005, and by then vaccine uptake had declined substantially. Although we cannot ascertain whether the decline between September and October was related to the GBS investigation, the impact of the investigation, if any, was likely minimal in this study population given the similar trend observed for Td.
Coverage was lower in the three MCOs reporting shortages (MCO C, D, and E) than in the other two. This included the two largest MCOs where implementation of a new vaccine recommendation is likely more challenging compared to smaller MCOs. Coverage differences within and between MCOs could have been affected by the nature of the shortage and/or the implementation plan at the MCO; we were unable to gather detailed information from all MCOs. The low coverage among the targeted adolescents in MCO D was likely related to their change in policy due to the shortage which instructed providers to defer vaccinating 11-12 and 14-5 year olds and preferentially vaccinate college freshman. School immunization laws have had a marked impact on both the incidence of vaccine preventable disease and immunization coverage in the United States. A limited number of studies suggest that much of the success of adolescent immunization programs in the United States is a direct result of these requirements[16
]. During our study period, none of the MCOs were affected by statewide MCV4 legislation (Table )[24
In February 2005, the ACIP carefully considered supply, anticipated demand and disease epidemiology when recommending routine use of MCV4 for adolescents 11-12 and 14-15 years of age[25
]. At the same meeting, the manufacturer projected producing about 5 million doses in 2005[26
] and the 2005 CDC Annual Biosurveillance data (CDC unpublished data) show that the manufacturer reported distributing about 3.1 million doses in 2005. The 2005 MCV4 mismatch between supply and demand and the reported shortages likely resulted from a combination of factors, including the increase in uptake over the summer, vaccination of non-targeted groups, and a supply shortfall possibly related to manufacturing capacity. Additionally, analysis of insurance claims data (presented by the MCV4 manufacturer at the June 2006 ACIP meeting) showed that from March-August 20, 2005, among adolescents, Menactra was administered more commonly to 18 year olds than other age groups [27
]. This also likely contributed to the reported shortages. Uptake among 18 year olds could not be assessed in this study because not all MCOs collected data for this group. Based partly on our findings, a supply-demand imbalance was anticipated the following summer (2006) leading ACIP to recommend in May 2006 that MCV4 administration to 11-12 year olds be deferred; that supply limitation was resolved in November 2006[28
]. In August 2007, ACIP changed their MCV4 recommendation to include routine vaccination of all adolescents 11-18 years partly to simplify provider decisions to vaccinate[30
This study has some limitations. First, these results based on the VSD data may not be generalizable to uninsured adolescents and coverage during 2005 was likely lower in the overall US adolescent population; however, with standardized protocols, VSD data can be used to obtain timely vaccination coverage of recently and newly recommended vaccination. Second, we could not account for vaccinations that might have been administered outside the VSD MCO; however, considering the high price of MCV4 (list price in 2005 was $82), it is unlikely that many adolescents received this vaccine outside the plan. Third, this study is unable to account for the effect of any local events on seasonality of vaccination trends such as a case of meningococcal disease. Finally, we could have underestimated coverage among the target groups due to misclassification; some adolescents included in the 13+16 year-old, "non-target" age group, may in fact have been vaccinated according to recommendations if they were starting high school at those ages. However, the impact on our estimates is likely negligible given US school enrollment estimates.