MAJOR ARTICLE
Determinants of Influenza Vaccination, 2003–2004:
Shortages, Fallacies and Disparities
Timothy F. Jones,1,2 L. Amanda Ingram,1 Allen S. Craig,1,2 and William Schaffner1,2
1
Tennessee Department of Health and 2Department of Preventive Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee
Background. The influenza outbreak of 2003–2004 received substantial media attention, including widespread
reports of a severe season and vaccine shortages. Understanding the determinants of vaccine receipt is important
for guiding immunization policies.
Methods. From February through June 2004, we administered a structured telephone survey to Tennessee
residents, using random-digit dialing methodology.
Results. Questionnaires were completed by 4028 persons, of whom 2077 (52%) had received influenza vaccination during the previous outbreak season. Of these 2077 vaccine recipients, 63% received vaccine at a private
medical clinic, 14% at a workplace, 11% at a health department, and 7% at a pharmacy. Three-fourths of
respondents reported a risk factor for which the Centers for Disease Control and Prevention recommends vaccination; of those, 41% went unvaccinated, including 26% who had seen a medical provider for other reasons
during the influenza season. More than 40% of persons aged ⭓50, more than half of health care workers, and
70% of pregnant women were not immunized. Blacks, rural residents, and lower-income respondents were significantly less likely to be immunized than were comparison groups. Of respondents who were vaccinated, 6%
reported difficulties obtaining vaccination (most commonly, they reported that vaccine was not readily available).
One-fourth of unvaccinated persons had been offered vaccination but had declined it; of these, 35% thought it
unnecessary and 33% believed it would make them ill. Of those not vaccinated, 8% reported requesting vaccination
but not receiving it, most commonly because it was unavailable.
Conclusions. Many barriers contribute to disparities in rates of influenza vaccination, of which inadequate
supply is only one component. Myths regarding influenza vaccination persist tenaciously. A multifaceted approach
to increasing immunization rates is critical.
Influenza causes substantial annual morbidity and mortality, including an estimated 36,000 influenza-associated deaths in the United States each year [1]. The large
majority of those deaths are of persons at high risk and
for whom influenza vaccination is recommended, particularly persons aged ⭓65 years [2]. The Centers for
Disease Control and Prevention (CDC) estimates that
65% of the US population falls into groups targeted for
influenza vaccination [3]. Despite substantial evidence
that vaccination is efficacious, safe, and cost-effective
[2, 4], many persons at elevated risk for the complications of influenza are not vaccinated appropriately.
Received 22 September 2004; accepted 28 October 2004; electronically
published 18 November 2004.
Reprints or correspondence: Dr. Timothy Jones, Communicable and
Environmental Disease Services, Tennessee Dept. of Health, 4th Fl., Cordell Hull
Bldg., 425 5th Ave. N., Nashville, TN 37247 (tim.f.jones@state.tn.us).
Clinical Infectious Diseases 2004; 39:1824–8
2004 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2004/3912-0013$15.00
1824 • CID 2004:39 (15 December) • Jones et al.
In 2001–2002, only two-thirds of persons ⭓65 years of
age and fewer than one-fourth of those under age 50
with a high-risk medical condition received influenza
vaccine [2].
The reasons for persistent under-immunization of
critical populations are legion and complex [5–10].
Substantial racial disparities in immunization rates
have persisted, despite efforts to reduce the gap [11–
13]. Recent problems with widespread shortages of
vaccine have contributed to concerns about problems
of access to vaccination [14, 15]. If efforts to attain
the goals of Healthy People 2010 [16]—annually immunizing 60% of adults with high-risk medical conditions and 90% of persons ⭓65 years—are to be
successful, improved understanding of the barriers to
successful vaccination will be critical to the development of effective strategies [2].
During the 2003–2004 influenza season in the United
States, widespread reports of an early and severe season
contributed to the substantially increased demand for
vaccine and the resulting shortages [17–19]. We performed a
telephone survey to assess people’s knowledge, attitudes, and
beliefs about vaccination and to assess how access issues may
have impacted other determinants of vaccination rates.
METHODS
From February through May 2004 a survey of Tennessee residents was administered using the computer assisted telephone
interviewing method. To reach a sample representative of the
state population, respondents were identified through random
selection of residential telephone numbers across all counties
in Tennessee. A maximum of 5 attempts were made to reach
each potential respondent, including telephone calls during evenings and weekends. The household member with the most
recent birthday was selected as the subject for interview. After
obtaining consent, trained interviewers administered the questionnaire. Persons over the age of 18 were interviewed directly.
For persons under the age of 18, a primary caregiver was interviewed. Respondents were excluded if their primary residence was not in Tennessee or if they did not speak English.
The survey included questions focused only on demographic
characteristics, influenza vaccination history and related experiences, and risk factors for influenza disease. Data were analyzed using EpiInfo, version 6.2 (CDC) and SAS, version 8.0
(SAS Institute) software. Multivariate logistic regression was
used to control for potential confounders and covariates and
to identify variables independently associated with vaccination;
adjusted odds ratios and 95% confidence intervals were used.
The study was approved by the institutional review board of
the Tennessee Department of Health.
RESULTS
Of persons successfully contacted by telephone, 88% agreed to
participate. Fewer than 1% of interviews were terminated before
completion. Interviews with 4028 respondents were completed.
The mean age of respondents was 51 years (range, 0–102 years).
Of these 4028 respondents, 2077 (52%) had been immunized
against influenza during the 2003–2004 outbreak season, including 736 (78%) of 938 persons aged ⭓65 years, and 1769
(59%) of 3020 persons who met at least 1 criterion for which
CDC recommends targeted vaccination [2]. A substantial proportion of respondents (3016 [75%]) indicated that they had
⭓1 condition for which the CDC recommends routine influenza vaccination [2], though many persons for whom it was
indicated remained unvaccinated (table 1). Of the 2077 vaccinated respondents, 1839 (89%) reported a history of previous
influenza vaccination.
The majority of immunized respondents were vaccinated in
October (58%) and November (24%). Persons who were vaccinated in November or earlier had significantly fewer diffi-
culties getting vaccine than those vaccinated later in the season
(5.7% vs. 9.3%; P p .017).
Of the 2077 vaccinated respondents, 1312 (63%) were vaccinated in a physicians’ office or medical clinic, 279 (14%) in
the workplace, 232 (11%) at a health department, and 145 (7%)
at a pharmacy. Only 10 (0.5%) of the vaccinated respondents
had received live attenuated intranasal influenza vaccine.
One-third of vaccinated persons said they were offered vaccination while seeing a medical provider for another reason;
650 (31%) of 2077 made a separate visit to the provider specifically for vaccination. Nine hundred seventeen (44%) stated
that they had requested vaccination from the provider. Of the
2077 respondents ultimately vaccinated, 128 (6%) reported encountering difficulties getting the vaccine; 71 (55%) were told
initially that vaccine was not available, and 41 (32%) were told
to return later for vaccination. First-time vaccinees were no
more likely to encounter problems than were repeat vaccinees,
and children encountered no more problems than did adults.
Of the 1951 respondents who were not vaccinated in 2003–
2004, there were 263 (13%) who reported that they had wanted
to be vaccinated. Five hundred twenty-four (27%) of these 1951
said that a medical provider had offered vaccination to them.
One hundred fifty-one (8%) reported having asked a provider
for it, of whom 82 (54%) were told vaccine was unavailable
and 12 (8%) were told vaccination was unnecessary. Two-thirds
of all survey respondents stated that they wanted to be vaccinated against influenza the next year, including 476 (24%) of
the 1951 who were not vaccinated in the 2003-2004 outbreak
season.
Persons who were offered vaccine by a provider were more
likely to have been vaccinated than those who were not offered
it (55% vs. 50%; P p .004). Immunization rates did not differ
significantly according to the sex or educational level of the
respondents. After adjusting for other variables, we found that
Table 1. Influenza vaccination status of 4028 respondents reporting that they had a risk factor for complications of influenza
infection for which the Centers for Disease Control and Prevention
recommends routine vaccination.
No. (%) of patients
Risk factor
Nursing home resident
Age of ⭓65 years
Age of 50–64 years
Age of !50 years
Has a high-risk medical condition
Cares for person at high risk
Is a health care worker
Is pregnant
No identified risk factor
Vaccinated
Not
vaccinated
54 (72)
21 (28)
736 (78)
665 (57)
202 (22)
504 (43)
169 (46)
195 (54)
184 (37)
133 (49)
25 (30)
319 (63)
138 (51)
58 (70)
308 (30)
704 (70)
Influenza Vaccination Determinants • CID 2004:39 (15 December) • 1825
Table 2. Influenza vaccination status of 4028 respondents, according to demographic and other characteristics.
No. (%) of patients
Characteristic
Sex
Male
Not
a
Vaccinated vaccinated aOR
592 (48)
636 (52)
Female
Education level
No college (adults only)
1483 (53)
1314 (47)
941 (52)
852 (48)
Any college (adults only)
Race
White
Nonwhite
1052 (51)
996 (49)
1821 (53)
249 (43)
Geographic residence
Rural
Nonrural
P
0.91
.30
0.83
.06
1617 (47)
326 (57)
1.51
.002
677 (49)
1393 (53)
709 (51)
1232 (47)
0.74
.0017
1769 (59)
308 (30)
1247 (41)
704 (70)
1.88 !.0001
1401 (66)
706 (34)
2.42 !.0001
676 (35)
1245 (65)
CDC risk factorb
Yes
No
Age
⭓50 years
!50 years
a
Adjusted OR calculated from logistic regression analysis, adjusting for
other variables
b
Any risk factor for which the Centers for Disease Control and Prevention
specifically recommends vaccination.
the likelihood of having been vaccinated was greater among
white persons than among respondents of other races, greater
among urban/suburban residents than among rural residents,
greater among persons aged ⭓50 years than among persons
aged !50 years, and greater among persons with risk factors
for which CDC recommends vaccination than among persons
without such risk factors (table 2). Among persons aged ⭓50
years, those with annual household incomes of !$30,000 were
more likely to get vaccinated (adjusted OR [adjusted OR], 1.3;
P p .028) than were those with higher incomes, whereas the
opposite was true among persons aged !50 (aOR, 0.58; P !
.001).
The most common reasons unvaccinated respondents cited
for not receiving vaccination included the belief that it was
unnecessary, the belief that vaccination would cause illness, and
failure to think about it (table 3). After adjusting for other
variables, unvaccinated respondents who had a high-risk medical condition for which CDC recommends vaccination were
less likely to believe vaccination was unnecessary (aOR, 0.69;
P p .003) or to have never thought about it (aOR, 0.68;
P p .007), and more likely to cite the belief that it would cause
illness (aOR, 1.73; P ! .0001) or cost barriers (aOR, 1.88;
P p .038), compared with respondents who did not have such
medical conditions. Unvaccinated persons with any condition
for which CDC recommends vaccination were more likely than
1826 • CID 2004:39 (15 December) • Jones et al.
persons for whom vaccination is not routinely encouraged to
cite the belief that vaccination would cause illness (aOR, 1.55;
P p .004) but were significantly less likely to cite that they had
never thought about vaccination (aOR, 0.67; P.0004).
DISCUSSION
It is unnecessary for 36,000 Americans to die each year from
a potentially vaccine-preventable disease. This study demonstrates the complex reasons why so many people who should
be immunized against influenza are not. Barriers to improving
immunization rates include patient-specific factors and medical-care provider issues, as well as health care delivery system
characteristics, all of which must be addressed in order to optimize delivery of one of the most fundamental and effective
public health interventions available.
Patient-centered determinants are perhaps the most vexing
of the myriad challenges to improving influenza immunization
rates. Although vaccination rates among persons in this study
aged ⭓65 or with high-risk medical conditions were higher
than were recent estimates of national rates [2], there remain
substantial disparities in coverage among important target populations. Significant racial disparities exist in all high-risk categories; for example, white respondents aged ⭓65 were 130%
more likely to have been immunized than were black respondents. Strikingly, the belief that vaccination would cause illness
was cited by significantly more unvaccinated persons in highrisk groups than by their counterparts in lower-risk groups.
Earlier studies have found that as many as half of respondents
believed that they would get influenza disease from the vaccine
[5, 10, 20]. Although our findings suggest modest improvement, there remain substantial misperceptions about the risks
of vaccination, which widespread educational efforts have not
successfully overcome. Paradoxically, these misconceptions are
more common among those most in need of vaccination.
Similarly, the belief that immunization was not necessary
(cited by a third of unvaccinated persons), concerns about side
effects, and simply never considering it were commonly cited
reasons for lack of vaccination, which suggests that important
educational messages are not successfully reaching at-risk populations, despite the efforts of the medical establishment. In
contrast, although there has been recent attention to potential
“mismatches” between the vaccine being given and the influenza strains in circulation [21], perceived ineffectiveness was
cited as a concern by relatively few respondents, with no differences between various risk groups. Extensive reviews of strategies for implementing vaccine recommendations emphasize
the need for continued evaluation of multifaceted approaches
[22, 23]. These data suggest that particular attention must be
given to targeting specific populations, and that a “one message
fits all” approach will not be sufficient.
Health care delivery system characteristics also substantially
Table 3.
Primary reasons cited by respondents for not receiving influenza vaccination during the 2003–2004 outbreak season.
No (%) of unvaccinated patients, by demographic or clinical characteristic
High-risk
medical condition
Age
Reason cited
CDC risk factora
All
⭓65 years ⭐64 years
Yes
No
Yes
No
White
Nonwhite
(n p 1951) (n p 202) (n p 1749) (n p 426) (n p 1525) (n p 1247) (n p 704) (n p 1617) (n p 326)
Thought vaccination not
necessary
641 (33)
55 (27)
596 (34)
116 (27)b
535 (35)
399 (32)
Believed vaccination would
cause illness
409 (21)
59 (29)
350 (20)
122 (29)b
b
287 (19)
335 (22)
158 (10)
292 (23)
227 (18)b
126 (10)
133 (9)
Never thought about it
Side effects not worth it
403 (21)
195 (10)
40 (20)
23 (11)
363 (21)
172 (10)
68 (16)
37 (9)
Believed vaccination not
effective
Cost barrier
171 (9)
50 (3)
18 (9)
4 (2)
153 (9)
46 (3)
38 (9)
17 (4)b
No time
98 (5)
7 (3)
91 (5)
19 (4)
33 (2)
79 (5)
Vaccine was not available
Never saw provider to ask
95 (5)
32 (2)
9 (4)
1 (0)
85 (5)
31 (2)
27 (6)
4 (1)
67 (4)
28 (2)
a
b
Race
252 (36)
545 (34)
101 (31)
117 (17)
176 (25)
69 (10)
321 (20)
318 (20)
152 (9)
87 (27)
84 (26)
42 (13)
108 (9)
63 (9)
140 (9)
30 (9)
36 (3)
61 (5)
14 (2)
37 (5)
36 (2)
81 (5)
13 (4)
17 (5)
66 (5)
21 (2)
28 (4)
11 (2)
86 (5)
29 (2)
8 (2)
3 (1)
b
Any risk factor for which the Centers for Disease Control and Prevention specifically recommends vaccination.
Significant difference (P ! .05) compared to the group without the specified characteristic on multivariate analysis.
impact influenza immunization rates. More than half of unvaccinated persons who reported requesting vaccine were told
it was unavailable, though this accounted for only 5% of all
unvaccinated persons. These data suggest that vaccine shortages
in 2003–2004 impeded at least some motivated persons from
receiving it. Studies of prior vaccine shortages have suggested
that overall vaccination rates were not adversely affected [14,
15]. These observations must be interpreted in the context of
the unusual amount of media attention generated by perceived
vaccine shortages, compounded by reports of an early and severe epidemic. The adage that the “best way to stimulate demand is to tell someone that they can’t have something” appears to have been borne out, with an “unprecedented demand”
for influenza vaccine observed in 2003–2004 [2, p. 21]. In
response to recent problems, the CDC, beginning this year, is
contracting to purchase several million doses of vaccine and is
developing a national influenza vaccine stockpile [24]. Although manufacturers anticipate production of substantially
more vaccine doses in 2004–2005, the challenge to the medical
establishment will be to ensure that available supplies are administered. In typical influenza outbreak seasons in the past,
10%–20% of vaccine produced has gone unused [25]. It is
reassuring that cost barriers were cited by few respondents to
this survey. The efforts of influenza vaccination programs to
overcome the financial barriers that hinder access to many other
aspects of medical care appear to have been successful.
Provider-related factors, particularly a medical provider’s
recommendation to get vaccine, also substantially affect immunization rates [10, 22, 26]. We observed slightly higher immunization rates among persons to whom a medical provider
offered vaccine, compared with persons to whom it was not
offered. Disturbingly, well over a third of persons at high risk
had seen a medical provider during the influenza season but
remained unvaccinated. Other studies have focused on missed
opportunities [6, 27] and the effect of medical-care provider
attitudes and practices on vaccination rates, and, in response,
these studies have advocated interventions such as automatic
prompts to clinicians, standing orders, and reminders sent to
patients [8, 22, 26, 28–30]. Evidence (including the fact that
140% of health care workers in our survey had not been vaccinated) suggests that there is ample room for improvement
in education and practices among providers that could increase
immunization rates [2, 31].
This study had a number of limitations. Assessment of immunization status was based on self-report; however, this approach has been demonstrated to be reliable and is commonly
used [26]. Our survey was limited to persons accessible by landline telephone, which may have excluded a certain segment of
the population. Persons who use cellular telephones exclusively,
for example, were not contacted. Nursing home residents were
only contacted if they had personal telephones. Strengths of
this study include the large and diverse population sampled.
Although the study was limited to a single state, responses
compared favorably with those of previous studies, and it is
likely that these observations are generalizable. Intranasal vaccine was used by so few respondents that we were unable to
assess the impact of the availability of this new product on
attitudes toward and beliefs about vaccination. Introduction of
a live-virus vaccine, with its occasional side effects of sore throat
and nasal stuffiness, may complicate efforts to reassure the
public that influenza vaccination cannot cause infection.
This study demonstrates that high-risk populations in the
Influenza Vaccination Determinants • CID 2004:39 (15 December) • 1827
United States remain underimmunized against influenza, and
substantial disparities in vaccination rates exist in certain
groups. Although the public health establishment is very concerned about perceptions of vaccine effectiveness, this survey
suggests that the population appears not to have major concerns in this area. In contrast, important misperceptions about
adverse effects, which many may have hoped had been put to
rest, persist. The notion that one can “get flu from the flu
vaccine” remains common. Lack of a perceived need for vaccination also remains commonplace, even among high-risk persons. Studies such as this are necessary to identify the issues
of true concern to vaccination candidates.
Acknowledgments
We thank Celia Larson, Ph.D., for her assistance with survey administration.
Potential conflicts of interest. All authors: no conflicts.
References
1. Thompson WW, Shay DK, Weintraub E, et al. Mortality associated
with influenza and respiratory syncytial virus in the United States.
JAMA 2003; 289:179–86.
2. Centers for Disease Control and Prevention. Prevention and control
of influenza. Recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Morb Mortal Wkly Rep
2004; 53(RR-6):1–40.
3. Centers for Disease Control and Prevention (CDC). Interim estimates
of populations targeted for influenza vaccination from 2002 National
Health Interview Survey data and estimates for 2004 based on influenza
vaccine shortage priority groups. Atlanta: National Immunization Program, CDC, 2004.
4. Nichol KL, Margolis KL, Wuorema J, von Sternberg T. The efficacy
and cost effectiveness of vaccination against influenza among elderly
persons living in the community. New Engl J Med 1994; 331:778–84.
5. Centers for Disease Control and Prevention. Reasons reported by Medicare beneficiaries for not receiving influenza and pneumococcal vaccinations—United States, 1996. MMWR Morb Mortal Wkly Rep
1999; 48:886–90.
6. Centers for Disease Control and Prevention. Missed opportunities for
pneumococcal and influenza vaccination of Medicare pneumonia inpatients—12 western states, 1995. MMWR Morb Mortal Wkly Rep
1997; 46:919–23.
7. Armstrong K, Berlin M, Schwartz JS, Propert K, Ubel PA. Barriers to
influenza immunization in a low-income urban population. Am J Prev
Med 2001; 20:21–5.
8. Evans MR, Watson PA. Why do older people not get immunized against
influenza? A community survey. Vaccine 2003; 21:2421–7.
9. Doran T, McCann R. Obstacles to influenza immunization in primary
care. J Pub Health Med 2001; 23:329–34.
10. Zimmerman RK, Santibanez TA, Janosky JE, et al. What affects influenza vaccination rates among older patients? An analysis from innercity, suburban, rural and Veterans Affairs practices. Am J Med 2003;
114:31–8.
1828 • CID 2004:39 (15 December) • Jones et al.
11. Schneider EC, Cleary PD, Zaslavsky AM, Epstein AM. Racial disparity
in influenza vaccination: does managed care narrow the gap between
African Americans and whites? JAMA 2001; 286:1455–60.
12. Ostbye T, Taylor DH, Lee AMM, Greenberg G, van Scoyoc L. Racial
differences in influenza vaccination among older Americans
1996–2000: longitudinal analysis of the Health and Retirement Study
(HRS) and Asset and Health Dynamics Among the Old (AHEAD)
survey. BMC Public Health 2003; 3:41.
13. Egede LE, Zheng D. Racial/ethinic differences in adult vaccination
among individuals with diabetes. Am J Pub Health 2003; 93:324–9.
14. Santibanez TA, Nowalk MP, Zimmerman RK, Bruehlman RD. Effects
of the year 2000 influenza vaccine delay on elderly patients’ attitudes
and behaviors. Prev Med 2003; 37:417–23.
15. Zimmerman RK, Nowalk MP, Santibanez TA, Jewell IK, Raymond M.
Shortage of influenza vaccine in 2000–2001—did it change patient
beliefs? Am J Prev Med 2003; 24:349–53.
16. US Department of Health and Human Services (DHHS). Healthy people 2010. Washington, DC: DHHS, 2000.
17. Charatan F. Widespread flu in United States exposes shortage of vaccine. BMJ 2004; 328:8.
18. Nelson R. Influenza vaccine shortage hits the USA. Lancet 2003; 362:
2075.
19. Stephenson J. Health officials warn flu season could be severe. JAMA
2003; 290:3052–3.
20. Fiebach NH, Viscoli CM. Patient acceptance of influenza vaccination.
Am J Med 1991; 91:393–400.
21. Centers for Disease Control and Prevention. Assessment of the effectiveness of the 2003–04 influenza vaccine among children and adults—
Colorado, 2003. MMWR Morb Mortal Wkly Rep 2004; 53:707–10.
22. Shefer A, Briss P, Rodewald L, et al. Improving immunization coverage
rates: an evidence-based review of the literature. Epidemiol Rev
1999; 21:96–142.
23. Sarnoff R, Rundall T. Meta-analysis of effectiveness of interventions
to increase influenza immunization rates among high-risk population
groups. Med Care Res Rev 1998; 55:432–56.
24. Centers for Disease Control and Prevention. Update: influenza activity—United States and worldwide 2003–04 season, and composition
of the 2004–05 influenza vaccine. MMWR Morb Mortal Wkly Rep
2004; 53:547–52.
25. O’Mara D, Fukuda K, Singleton JA. Influenza vaccine: ensuring timely
and adequate supplies. J Infect Med 2003; 20:548–54.
26. Nichol KL, Mac Donald R, Hauge M. Factors associated with influenza
and pneumococcal vaccination behavior among high-risk adults. Gen
Intern Med 1996; 11:673–7.
27. Williams WW, Hickson MA, Kane MA, Kendal AP, Spika JS, Hinman
AR. Immunization policies and vaccine coverage among adults—the
risk for missed opportunities. Ann Internal Med 1988; 108:625.
28. Litt JCB, Laker PB. Improving influenza vaccine coverage in at-risk
groups: good intentions are not enough. Med J Aust 1993; 159:542–7.
29. Frank JW, Henderson M, McMurray L. Influenza vaccination in the
elderly. 1. Determinants of acceptance. Can Med Assoc J 1985; 132:
371–5.
30. Nichol KL, Zimmerman RK. Generalist and subspecialist physicians’
knowledge, attitudes and practices regarding influenza and pneumococcal vaccinations for elderly and other high-risk patients- a nationwide survey. Arch Intern Med 2001; 161:2702–8.
31. Martinello RA, Jones L, Topal JE. Correlation between healthcare workers’ knowledge of influenza vaccine and vaccine receipt. Infect Control
Hosp Epidemiol 2003; 24:845–7.