Interim Results: State-Specific Seasonal Influenza Vaccination Coverage --- United States, August 2009--January 2010

Interim Results: State-Specific Seasonal Influenza Vaccination Coverage --- United States, August 2009--January 2010
Weekly
April 30, 2010 / 59(16);477-484


The advent of the 2009 influenza A (H1N1) pandemic in April 2009 made the 2009--10 influenza season highly unusual. Public awareness of the potential seriousness of influenza was heightened by media coverage of pandemic-associated hospitalizations and deaths, especially among younger persons. In the fall, the distribution of two separate influenza vaccines began, with distinct, although overlapping, recommendations from the Advisory Committee on Immunization Practices (ACIP) (1,2). In addition, 2009--10 was the first full season in which ACIP's recommendation to vaccinate all children aged 5--18 years (2) was implemented. To provide preliminary state-specific estimates of seasonal influenza vaccination coverage, CDC analyzed Behavioral Risk Factor Surveillance System (BRFSS) and National 2009 H1N1 Flu Survey (NHFS) data collected during October 2009--February 2010. By January 31, estimated state seasonal influenza vaccination coverage among persons aged ≥6 months ranged from 30.3% to 54.5% (median: 40.6%). Median coverage was 41.2% for children aged 6 months--17 years, 38.3% for adults aged 18--49 years with high-risk conditions, 28.8% for adults aged 18--49 years without high-risk conditions, 45.5% for adults aged 50--64 years, and 69.3% for adults aged ≥65 years. These results, compared with the previous season (3), suggest large increases in coverage for children and a moderate increase for adults aged 18--49 years without high-risk conditions. Health departments should identify best practices that lead to higher vaccination coverage and should support effective vaccination services (e.g., school-located vaccination programs and office-based protocols, such as reminder/recall and standing orders).

CDC used data collected during October 2009--February 2010 from two separate surveys, BRFSS and NHFS. BRFSS respondents in 50 states, the District of Columbia, and two territories were asked if they (or their children, in 43 states) had been vaccinated for the "seasonal flu" in the past 12 months, and if so, in which month.* NHFS respondents aged ≥18 years were asked whether they (or their children) had received "seasonal flu vaccination" since August, and if so, in which month.† Similar questions were asked about "H1N1 flu" vaccination in both surveys (4). The Council of American Survey and Research Organizations (CASRO) NHFS response rates were 35% for landline and 27% for cellular telephones; cooperation rates were 45% and 57%, respectively. The median state CASRO BRFSS response and cooperation rates were 54% and 76%, respectively.

To improve the precision of state-level estimates from each survey, CDC first combined the monthly data collected during October 2009--February 2010 and estimated the cumulative proportion of persons vaccinated with at least 1 dose during August--January by using the Kaplan-Meier survival analysis procedure. To improve precision for subgroups within states, particularly children, CDC then combined the estimates from BRFSS and NHFS (5).§ The 2009--10 BRFSS survey included 18,368 children and 152,128 adults; the NHFS included 60,786 children and 31,135 adults. CDC performed Pearson correlations among children and adults to determine if state-specific seasonal coverage was associated with state-specific H1N1 coverage or with coverage during past seasons. For comparison to estimated national vaccination coverage for the 2008--09 influenza season, overall, and by month, CDC analyzed the BRFSS adult and child data available from 19 states that fully participated in a special early survey conducted January--February 2009 (3). Student t-tests were used to determine statistical difference between groups.

Among children aged 6 months--17 years, estimated national 2009--10 seasonal coverage was 40.0% (16 percentage points higher than in 2008--09) (3). State-specific 2009--10 coverage for children ranged from 23.6% (Nevada) to 67.2% (Hawaii) (U.S. median: 41.2%); coverage among adults aged ≥18 years ranged from 32.4% (Nevada) to 52.5% (Minnesota) (U.S. median: 40.6%) (Table 1). State-specific child and adult (age ≥18 years) coverages were correlated positively (r = 0.68), with a wider range of coverage among state-specific child levels compared with adult levels (Figure 1). Seasonal state coverage also correlated positively with 2009 H1N1 state coverage (5) among children (r = 0.72) and adults (r = 0.72). Among adults, state coverage in 2009--10 was correlated positively with coverage in 2007--08 (r = 0.81) (CDC, unpublished data, 2010).

Among adults aged 18--49 years, national seasonal vaccination coverage was higher for persons with high-risk conditions (36.2%) than without high-risk conditions (27.6%). This difference also was statistically significant in 10 states (Table 1). Coverage ranged from 21.2% (Mississippi) to 63.4% (Minnesota) for the high-risk group, and for the persons not at high risk, 19.4% (Florida) to 43.1% (South Dakota). Seasonal coverage for adults aged 18--49 years with high-risk conditions was similar to coverage in 2008--09. Among younger adults without such conditions, coverage was 30% higher in 2009--10 than in 2008--09 (20.8%) (3; CDC, unpublished data, 2010).

National seasonal influenza vaccination coverage was 45.0% for adults aged 50--64 years and 68.0% for adults aged ≥65 years, similar to previous seasons. Coverage ranged from 33.7% (Florida) to 56.3% (Hawaii) for adults aged 50--64 years and from 59.3% (Idaho) to 78.6% (New Mexico) for adults aged ≥65 years (Table 1).

Among all persons aged ≥6 months, coverage was higher among non-Hispanic whites (42.5%) compared with non-Hispanic blacks (32.2%) and Hispanics (33.6%) (Table 2), and was similar to Asians (41.7%; 95% confidence interval [CI] = 38.1--45.3) and American Indians/Alaska Natives (40.1%; CI = 35.7--44.5). Among adults, blacks (31.1%; CI = 29.1--33.1) and Hispanics (30.9%; CI = 28.2--33.6) had lower coverage than whites (43.0%; CI = 42.4--43.6). Among children, coverage among blacks was lowest (32.7%; CI = 29.5--35.9), followed by Hispanics (39.3%; CI = 36.0--42.6) and whites (40.3%; CI = 38.9--41.7).

Coverage for persons aged ≥6 months was higher during September and October and lower in November compared with 2008--09 (Figure 2). One or more seasonal doses reportedly were administered to 29.1 million children (CI = 28.1--30.0 million) and 89.7 million adults (CI = 88.3--91.2 million), for an estimated 118.8 million vaccinees (CI = 117.1--120.5 million) during August 2009--January 2010.

Reported by
H Ding, MD, PJ Lu, MD, PhD, GL Euler, DrPH, C Furlow, PhD, LN Bryan, MS, B Bardenheier, MA, MPH, E Monsell, AG Gonzalez-Feliciano, MPH, C LeBaron, MD, PM Wortley, MD, JA Singleton, MS, Immunization Svc Div; M Town, MS, L Balluz, ScD, Div of Behavioral Science, Office of Surveillance, Epidemiology, and Laboratory Svcs, CDC.

Editorial Note
Seasonal influenza vaccination during 2009--10 occurred under unique circumstances. In October 2009, the distribution of influenza A (H1N1) 2009 monovalent vaccine was added to the ongoing distribution of the seasonal influenza vaccine that started in August. The populations recommended for 2009 H1N1 vaccination were different from, but overlapped, those recommended for seasonal vaccination (1,2). For this unusual season, coverage for seasonal vaccination of children aged 6 months--17 years (40%) was higher than coverage estimates for the previous season (24% from BRFSS [3] and 30% from NHIS [CDC, unpublished data, 2009]). Coverage for healthy adults aged 18--49 years also was moderately higher (3). These higher coverage levels for 2009--10 might reflect increased awareness of the seriousness of influenza associated with the H1N1 pandemic. Influenza activity due to 2009 H1N1 virus increased during summer and early fall, when only seasonal vaccine was available, and was associated with a threefold higher levels of hospitalizations and deaths among persons aged <65 years compared with previous seasons (6).¶ Higher coverage in children also might reflect the first full year of expansion of ACIP recommendations for all children aged 6 months--18 years to receive annual influenza vaccination (2). Nevertheless, despite increased attention to influenza during the fall, no significant increase in coverage occurred among those aged 18--49 years with high-risk conditions, underscoring the challenges associated with increasing coverage in this group.

Seasonal vaccination coverage varied widely among states, particularly among children. The strong positive correlations between 2009--10 seasonal state levels with both current 2009 H1N1 vaccine coverage and with previous season coverage suggests that certain factors that differ among states (e.g., medical-care delivery infrastructure, population norms, and effectiveness of state and local immunization programs) might explain at least part of the state-level variation in vaccination levels. Increased coverage in September likely reflects the early availability of vaccine supplies, coupled with increased demand, compared with previous seasons.

The lower seasonal influenza vaccination coverage for Hispanics and blacks observed during 2009--10 compared with whites among persons aged ≥6 months also has been observed in previous seasons among adults (3). The reasons for racial/ethnic disparities and their variations by state are multifactorial. Broad use of interventions to systematize offering of vaccination (e.g., routine offering of vaccine to all patients) is one important component of efforts to reduce these disparities (7).

The estimate of 119 million persons in the United States receiving at least 1 dose of seasonal vaccine certainly is an overestimate because only 114--115 million doses were distributed in the United States. The actual number of first doses available to be administered to the surveyed population was even lower because of unused doses, a second dose administered to children aged <9 years, and institutional use of doses (e.g., military and long-term care facilities not surveyed by BRFSS). The overestimate based on the surveys likely was caused by nonresponse bias, which has been observed for telephone surveys conducted during previous seasons, and recall bias for self-report or parental report of vaccination.

The findings in this report are subject to at least six other limitations. First, misclassification of H1N1 for seasonal vaccine, unique to this season, might have contributed to some overreporting. However, the high seasonal vaccination coverage in September before 2009 H1N1 vaccine was available suggests that survey respondents generally were able to distinguish between the two types of influenza vaccinations. Second, both BRFSS and NHFS are telephone-based surveys and thus do not include persons without telephone service. Also, BRFSS is limited to households with landlines. Third, response rates for both surveys were low (5), which increases the risk for nonresponse bias. Fourth, for these surveys, self-reported vaccination status is not validated with medical records and is subject to respondents' recall bias. Fifth, combining BRFSS and NHFS estimates increased the sample to approximately 260,000 persons, but differences in survey methods (e.g., different sampling frame, survey questions, operations, response rates, and weighting) might lead to different levels of bias that are averaged in the combined estimates of this report (5). Given all of these factors combined, the net bias (estimated coverage minus actual coverage) range likely is 10%--15%.** This range was probably higher than in previous seasons due, in part, to higher media coverage of the pandemic. Nevertheless, because demand for seasonal influenza vaccination in 2009--10 remained strong even after vaccine was in short supply in some places, a higher proportion of available doses likely were used this season compared with previous seasons. Finally, the 2008--09 BRFSS child coverage data slightly underestimated coverage through January and were based on only 19 states reporting (3). However, 2008--09 estimates of overall coverage levels and distribution by month matched those of the 2007--08 season.

Even with increased demand for vaccination this season, influenza vaccination levels were well below Healthy People 2010 targets of 60% for noninstitutionalized adults aged 18--64 years with high-risk conditions and 90% for adults aged ≥65 years (objectives 14-29a and 14-29c) (8). School-located vaccination was implemented in many communities for 2009 H1N1 vaccination (CDC, unpublished data, 2010), and the experience acquired might lead to greater use of this strategy in fall 2010, and thereby contribute to ongoing gains in influenza vaccination.

Acknowledgments
The findings in this report are based, in part, on contributions by KR Copeland, N Ganesh, M Stanislawski, and N Davis, National Opinion Research Center, Chicago, Illinois; state BRFSS coordinators; the Div of Behavioral Science, Office of Surveillance, Epidemiology, and Laboratory Svcs; and the H1N1 Vaccine Coverage Monitoring Team, CDC.

References
1.CDC. Use of influenza A (H1N1) 2009 monovalent vaccine: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2009. MMWR 2009;58(No. RR-10).
2.CDC. Prevention and control of seasonal influenza with vaccines. Recommendations of the Advisory Committee on Immunization Practices (ACIP), 2009. MMWR 2009;58 (No. RR-8).
3.CDC. Influenza vaccination coverage among children and adults---United States, 2008--09 influenza season. MMWR 2009;58:1091--5.
4.CDC. Interim results: influenza A (H1N1) 2009 monovalent vaccination coverage---United States, October--December 2009. MMWR 2010;59:44--8.
5.CDC. Interim results: state-specific influenza A (H1N1) 2009 monovalent vaccination coverage---United States, October 2009--January 2010. MMWR 2010;59:363--8.
6.CDC. Update: influenza activity---United States, August 30, 2009--March 27, 2010, and composition of the 2010--11 influenza vaccine. MMWR 2010;59:423--30.
7.Fiscella K. Anatomy of racial disparity in influenza vaccination. Health Serv Res 2005;40:539--50.
8.US Department of Health and Human Services. Healthy people 2010 (conference ed., in 2 vols.). Washington, DC: US Department of Health and Human Services; 2000. Available at http://www.health.gov/healthypeople. Accessed April 22, 2010.
* Respondents were asked, "Now I will ask you questions about seasonal flu. A flu shot is an influenza vaccine injected into your arm. During the past 12 months, have you had a seasonal flu shot? During what month and year did you receive your most recent seasonal flu shot? The seasonal flu vaccine sprayed in the nose is also called FluMist. During the past 12 months, have you had a seasonal flu vaccine that was sprayed in your nose? During what month and year did you receive your most recent seasonal flu vaccine that was sprayed in your nose?" Additional information about BRFSS is available at http://www.cdc.gov/brfss.

† Respondents were asked, "Since August 2009, have you had a seasonal flu vaccination? There are two types of seasonal flu vaccinations. One is a shot and the other is a spray, mist or drop in the nose. During what month did you receive your most recent seasonal flu vaccine? Was your most recent seasonal flu vaccine a shot or the spray in the nose? The seasonal flu vaccine can be given either as a shot or a nasal spray, also called FluMist." The landline sample was augmented with a sample of children aged <18 years identified during screening for the National Immunization Survey. Additional information about NHFS is available at http://www.cdc.gov/nis/h1n1_introduction.htm and http://www.cdc.gov/nis/data/h1n1_flu_survey.pdf .

§ Combined estimates were weighted averages of the BRFSS and NHFS estimates, with weights being determined by the effective sample sizes. The effective sample sizes take into account the design of each survey and are determined as the unweighted sample size divided by the design effect. The design effect is the ratio of the variance of a survey estimate to the variance had the survey design used a simple random sample; surveys with large design effects are less efficient. CDC estimated state and age-group--specific design effects based on estimated proportions vaccinated each month, using data from each survey from October 2009 through February 2010. Among states, the median design effects for children were 1.9 for NHFS and 1.3 for BRFSS, with the BRFSS estimate receiving a median of 34% of the weight in the combined average estimate. For adults, median design effects were 1.2 for NHFS and 1.5 for BRFSS, with BRFSS estimates receiving a median of 80% of the weight. The NHFS estimate was used alone when no data were available from BRFSS.

¶ Additional information available at http://www.cdc.gov/h1n1flu/estimates_2009_h1n1.htm.

** Before estimating the net overreporting bias observed, subtractions from the amount of doses distributed were made for estimates of 6 million unused doses and for additional coverage not included in this report (e.g., an estimate of 2 million second doses to children aged <9 years and approximately 2 million doses administered to persons in the military and nursing homes).

open here to see the full-text:
Interim Results: State-Specific Seasonal Influenza Vaccination Coverage --- United States, August 2009--January 2010