Abstract
Introduction
The incidence of pertussis in the United States has been rising since 2007, with more than 27 500 cases reported in 2010. 1 Furthermore, the actual number of cases is likely higher, as some are unreported. One contributing factor is that the presentation of pertussis illness in adults is often mild and may be unrecognized, resulting in underreporting and complicating efforts to prevent transmission. 2 Adult vaccination is important to reduce disease burden and also to protect infants who have not completed the primary vaccination series. In the United States, most pertussis-related deaths occur in infants younger than 3 months. 3
In 2006, the Advisory Committee on Immunization Practices recommended the tetanus–diphtheria–acellular pertussis (Tdap) vaccine for adults aged 18 to 64 years as a single replacement dose for the routine tetanus–diphtheria (Td) booster. 4 More recently, they recommended Tdap for individuals who have close contact with infants younger than 12 months, including those aged 65 years and older. 3 Unfortunately, many adults who get the routine Td booster do not receive Tdap. It is a matter of concern that the proportion of tetanus vaccinations that are Tdap has remained essentially unchanged at approximately 50% for several years.5-7
Most current literature evaluates provider-level factors in relation to Tdap vaccination.8-10 There is only limited information regarding patient-level factors. In this regard, one study suggests that individuals with a higher education level and those who received a recent influenza vaccination were more likely to report receiving Tdap. 11 Given what is currently known about predictors of Tdap, there is a need for more studies that focus on individuals who report receiving Td to determine what additional factors are related to receiving the acellular pertussis formulation. The aim of this study was to describe these factors among adults aged 18 to 64 years in the United States.
Methods
This was a cross-sectional study using data from the 2008 National Health Interview Survey. The survey methodology is described in detail elsewhere. 12 Briefly, the survey is conducted annually to monitor the health of the civilian, noninstitutionalized US population. It uses a multistage sampling plan to identify a representative subset of households covering the 50 states and the District of Columbia. Personal household interviews are conducted by agents of the US Bureau of the Census. Participation in the survey is voluntary.
The current study identified respondents who were aged 18 to 64 years and reported having a tetanus shot since 2005. A total of 1336 subjects had a yes/no response to the survey question, “Did the vaccine include pertussis/whooping cough?” All information related to study variables was obtained from public-use data files made available by the Centers for Disease Control and Prevention. The study was determined to meet oversight exemption criteria by the Institutional Review Board for BayCare Health System and the University of South Florida.
The primary dichotomous outcome variable was Tdap vaccination based on self-report. The potential predictor variables were chosen based on studies suggesting that barriers to vaccination are often related to sociodemographics, utilization of health care services, health status, and an individuals’ preventive behavior.13-16 Sociodemographic variables included age (18-24, 25-49, and 50-64 years), race-ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other), gender, language (English, non-English), US census region, dependent children <18 years old (yes/no), marital status (married, not married), education level (high school or less vs more than high school), employment status (employed vs unemployed), and total family income level (<$35 000, $35 000 up to $74 999, $75 000 up to $99 999, and ≥$100 000).
Variables for utilization of health care services included insurance status (private, public, none), office visits within 12 months (0, 1, 2-3, 4-9, ≥10), and presence of any barrier to access (yes/no). Barriers included inability to obtain medical care in the previous 12 months due to cost, waiting too long in the doctor’s office, not able to get through on the telephone, unable to obtain an appointment in a timely manner, office not being open when respondent could attend, or lack of transportation. Physician preventive behavior (yes/no) was defined as a recommendation for Pap smear, mammogram, prostate-specific antigen testing, or colorectal cancer screening.
Health status was assessed by examining coexisting illness, limitations of activity due to illness, and self-reported rating of health. Coexisting medical illnesses (yes/no) included coronary heart disease, angina pectoris, heart attack, heart condition/disease, diabetes, cancer, asthma, emphysema, chronic bronchitis, or weak/failing kidneys in the past 12 months. Psychiatric illness (yes/no) included depression or generalized anxiety. Other health status variables included functional limitation (yes/no) and self-reported health status (very good, good, or not good).
Variables assessing respondent preventive behavior included receiving influenza vaccine (yes/no), ever having pneumococcal vaccine (yes/no), and ever having received hepatitis A, hepatitis B, human papillomavirus, or varicella zoster vaccines (yes/no). Other recommended preventive services included use of sunscreen on a warm sunny day (always or most of the time), ever having a Pap smear, mammogram, prostate-specific antigen test, colorectal screening exam, or ever following advice to change diet, exercise, or reduce alcohol use for high blood pressure (yes/no).
The frequency of Tdap vaccination in relation to predictor variables was described taking into account the complex survey methodology of the National Health Interview Survey. Univariate analysis used logistic regression by class (SAS SURVEYLOGISTIC procedure) to determine the relationship between categorical predictors and the primary outcome. This procedure fits linear logistic regression models for discrete response survey data by the method of maximum likelihood and incorporates the sample design, such as stratification, clustering, and unequal weighting into the analysis. Tests of trend for ordinal variables (age, income, office visits in the past 12 months, education, and self-reported health status) were conducted using logistic modeling without the class statement. Multivariate logistic regression examined adjusted odds ratios (ORs) and 95% confidence intervals (CIs). As an exploratory study, all potential predictor variables were included in the multivariate model. Analyses were conducted using SAS version 9.2. A P value ≤.05 was considered statistically significant.
Results
Sample characteristics and Tdap vaccination rates are shown in Table 1. Of 1336 respondents, 51.1% who received tetanus vaccination reported receiving Tdap. The unadjusted odds of receiving Tdap are shown in Table 2. In univariate analysis, characteristics significantly associated with higher odds of Tdap included female gender (OR = 1.27, 95% CI = 1.02-1.59), Asian and other race-ethnicity (OR = 1.86, 95% CI = 1.14-3.05), college education (OR = 1.67, 95% = CI 1.32-2.12), 2 to 3 office visits in the past year (OR = 1.88, 95% CI = 1.26-2.82), and receiving other vaccines (OR = 1.40, 95% CI = 1.13-1.74). Tests of trend were significant for increasing level of education (P < .001) and increasing level of income (P = .017).
Percentage of Respondents Receiving Tdap by Selected Characteristics: NHIS 2008.
Abbreviations: Tdap, tetanus–diphtheria–acellular pertussis; NHIS, National Health Interview Survey.
Weighted proportion.
Military coverage includes TRICARE, VA (Veterans Affairs), or CHAMP-VA (Civilian Health and Medical Program of the Department of Veterans Affairs) health plans.
Public health care coverage includes Medicare, Medicaid, Indian Health Service, other government programs, or state-sponsored health plans.
Unadjusted and Adjusted Odds Ratios of Receiving Tdap by Selected Respondent Characteristics: NHIS 2008.
Abbreviations: Tdap, tetanus–diphtheria–acellular pertussis; NHIS, National Health Interview Survey; OR, odds ratio; 95% CI, 95% confidence interval; ref, reference.
P < .05.
Military coverage includes TRICARE, VA (Veterans Affairs), or CHAMP-VA (Civilian Health and Medical Program of the Department of Veterans Affairs) health plans.
Public health care coverage includes Medicare, Medicaid, Indian Health Service, other government programs, or state-sponsored health plans.
Characteristics significantly associated with lower odds of Tdap in univariate analysis included age 25 to 49 years (OR = 0.63, 95% CI = 0.45-0.88), age 50 to 64 years (OR = 0.47, 95% CI = 0.33-0.66), non-English language (OR = 0.57, 95% CI = 0.33-0.99), barrier to access (OR = 0.77, 95% CI = 0.60-0.98), medical illness (OR = 0.76, 95% CI = 0.60-0.96), functional limitations (OR = 0.60, 95% CI = 0.47-0.76), good health status (compared with very good; OR = 0.70, 95% CI = 0.53-0.92), and not good rating of health (OR = 0.63, 95% CI = 0.44-0.91). Tests of trend were significant for increasing age (P < .001) and decreasing health status (P < .002).
In multivariate analysis, several characteristics remained significantly associated with Tdap vaccination (Table 2). College education predicted receiving Tdap (OR = 1.55, 95% CI = 1.16-2.07), as did 2 to 3 office visits (OR = 2.01, 95% CI = 1.32-3.06) or 4 to 9 office visits (OR = 1.60, 95% CI = 1.06-2.42) in the past year. Age 50 to 64 years (OR = 0.61, 95% CI = 0.38-0.96) and functional limitations (OR = 0.70, 95% CI = 0.54-0.91) were associated with lower odds of receiving Tdap.
Discussion
In this study, 51.1% of adults aged 18 to 64 years who received tetanus vaccination during 2005-2008 received Tdap. For several reasons, we believe vaccination rates should be higher. First, contraindications are rare and unlikely to contribute to low coverage. 17 Also, coverage has continued to be about 50%, indicating minimum influence of recent introduction.6,7 Last, this study does not show that pertussis vaccination is less likely for the uninsured, suggesting cost does not play a role. Therefore, other factors must contribute to undervaccination.
As age increased, the odds of Tdap utilization decreased. While young parents are a source of pertussis transmission to infants, grandparents and other adults may also be sources and should be vaccinated. 18 Another predictor is education, which has been shown to be associated with vaccination.11,19,20 This study builds on previous studies by suggesting that higher education predicts receiving acellular pertussis vaccine in those receiving Td, possibly related to greater awareness of pertussis and its risks.
This study uses patient-level data; the impact of provider-level factors on pertussis vaccination is also important. Provider barriers might include doubt regarding the seriousness of pertussis and lack of knowledge about immunization guidelines.21,22 These barriers reduce the likelihood of provider recommendation, which influences patients’ decisions regarding vaccination.11,22,23 Providers may also miss opportunities to vaccinate because of competing demands, which could explain why respondents in this study were not more likely to receive pertussis vaccination with 10 or more yearly office visits.21,24 In the setting of complex office visits, there may be little time to fully explain the benefit of pertussis vaccination and its rationale over routine tetanus vaccine.
It was interesting to find that participants reporting functional limitation were 30% less likely to be vaccinated. Other studies have shown increased vaccination with chronic illness, which may be related in part to patient perception of susceptibility to disease and potential severity of the outcome.25-27 In our analysis, perhaps patients with more severe illness and associated limitations have concerns about pertussis vaccination worsening their condition. Misconceptions of this sort are not uncommon and will require further study.
We recognize the limitations of this analysis. Because multivariable models require complete data on included variables, missing data may reduce statistical power by reducing the effective sample size. We recoded variables that would have naturally led to missing data (eg, men responding to questions about mammography) in order to include them in our analysis. Also, the data are subject to respondent recall bias. Tdap vaccination was self-reported and was not verified by medical records. Systematic misclassification is possible, which could affect our conclusions.
Our study sample was selected based on a yes/no response to the question, “Did the doctor tell you the vaccine included the pertussis or whooping cough vaccine?” Individuals with responses of doctor did not say, not ascertained, and don’t know were excluded from analysis (n = 3189). Compared with this excluded group, our sample was more likely to be white, English-speaking, female, and married. They were also more likely to have college education, be employed, have higher income, and have private insurance. Based on these differences, we suspect that our study overestimated pertussis vaccination rates and the actual proportion of persons receiving Tdap is likely less than the 51% found in our study.
In 2008, approximately half of adult Td vaccinations included acellular pertussis, suggesting a need for increased awareness regarding pertussis, its consequences, and methods of prevention. Older individuals, those who are educationally disadvantaged, and those with functional impairment may be especially suited for supplemental immunization initiatives.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
