Promoting Tdap immunization in pregnancy: Associations between maternal perceptions and vaccination rates

Abstract

Objective:

Tdap vaccine uptake among US pregnant women is low despite current recommendations. This study evaluated if a Tdap vaccine information statement (VIS) affected overall perception, vaccination intention, and components of a health behavior model associated with Tdap vaccination rates.

Methods:

A randomized, prospective study was conducted among pregnant women receiving care at two women’s clinics in May-August 2014. Verbally consented participants were randomized to receive either the standard CDC Tdap VIS (sVIS) or a modified version (mVIS) before completing the first multi-part survey (T1). After T1, participants read their assigned VIS then completed the second part (T2). A 2015 chart review identified vaccinated participants. A health behavior model was hypothesized using the Reasoned Action Approach and Health Belief Model. Logistic regression, path analysis, and chi-square tests were used in the analysis.

Results:

279 surveys were analyzed. Average age of the participants was 26.4 years (SD=5.7) with average gestational age of 25.9 weeks (SD=9.2). 13% self-reported receiving Tdap vaccine prior to the survey. Overall perception scores significantly increased (3.1 to 3.4, p<0.001) after VIS review. A chart review showed that 131 (47%) received the vaccine post study. There was no significant difference in vaccination rates between the sVIS and mVIS groups (45% vs. 49%). Perceived benefits (B=0.315) and self-efficacy (B=0.197) were positively associated with the overall perception (T1), while perceived barriers (B=−0.191) were negatively associated with the overall perception (T1). Social norms (B=0.230), self-efficacy (B=0.213), and perceived benefits (B=0.117) were positively associated with vaccination intention (T1). The vaccination intention (T2) was positively associated with participants’ decision to receive Tdap vaccine (B=0.223).

Conclusion:

A VIS improved overall perception of the Tdap vaccine. Vaccination intention was a predictor of Tdap vaccination. It is crucial to provide information about immunization benefits to promote maternal Tdap vaccination.

1. Introduction

Maternal Tdap (tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis absorbed) vaccination during pregnancy is the best way to protect newborns from pertussis infection until they can start their own DTaP (diphtheria, tetanus, pertussis) vaccination series at two months of age [1;2]. A report from the Centers for Disease Control (CDC) showed that Tdap vaccination rates during pregnancy between 2011-2013 were only 14.3% in a Michigan Medicaid population, which were comparable to the report from the Vaccine Safety Datalink (13.7% in 2012) [3;4]. However, the most recent reports showed increased maternal Tdap vaccine rates in the US ranges from 51% to 56% in 2014 [5;6] due to national immunization guidelines.

In 2011, the CDC’s Advisory Committee on Immunization Practices (ACIP) recommended Tdap vaccination be offered to pregnant women after 20 weeks gestation if not previously immunized [7]. Postpartum immunization continued to be endorsed for women who had not been immunized during pregnancy. The stronger guidelines were published in February 2013, when the ACIP advised that all pregnant women receive pertussis vaccine during every pregnancy regardless of vaccination history because of waning maternal antibodies for pertussis in subsequent pregnancies [8]. The optimal time period for immunization is 27 to 36 weeks gestational age, but it may be administered earlier in pregnancy in patients at high-risk for pertussis. To allow time for antibody production and facilitate passive immunity to the fetus, the vaccination is ideally administered at least 2 weeks prior to delivery.

These recommendations are in response to the re-emergence of pertussis in the US, which has become a major public health concern. Despite vaccine availability since the 1940s, the reported incidence of this highly contagious disease continues to rise. A fivefold increase in pertussis rates occurred from 1991 to 2005 (26.4 per 100,000 persons to 103.5 per 100,000 persons) [9]. In 2012, the CDC reported approximately 50,000 cases of pertussis in the US, which was the highest reported since 1955 [10;11]. Pertussis has earned the distinction of being “the most poorly controlled vaccine-preventable disease in the developed world” [12]. Improved diagnosis, reporting, and awareness of pertussis as well as the waning immunity after vaccination have been suggested as contributing to the resurgence in recent years [13;14].

Populations at risk of transmitting or acquiring pertussis include adolescents, healthcare workers, and newborns [13]. Data from the 2012 National Health Interview Survey showed that 25.6% of adults aged 19-64 years who reported living with an infant younger than 1 year of age received a Tdap vaccine [15]. Infants younger than three months old are of particular concern due to their extreme vulnerability. This age group has the highest mortality (85% of deaths), morbidity, hospitalization rate, and complication rate [12]. The infection rate in all infants less than six months old is 20 times higher than the rate in the total population [16]. The 2015 CDC’s child and adolescent immunization schedules recommend a series of five DTaP vaccinations beginning at 2 months of age which creates a window of pertussis susceptibility in newborns [17].

Tdap vaccination is safe in pregnancy and early data suggest that it is effective at preventing pertussis in newborns, yet robust data regarding efficacy is limited. A 64% vaccination rate was reported in the United Kingdom (UK) after the UK Department of Health recommended that pregnant women be offered a single injection of acellular pertussis vaccine between 28 and 38 weeks gestation in October 2012, a year after the ACIP recommendations. The nationally implemented vaccination strategy was in response to increased pertussis-related infant morbidity and mortality [18]. There are several potential barriers relating to the relatively low Tdap vaccination rate in US pregnant women [19–21]. From a provider perspective, financial barriers to administration of Tdap vaccine in pregnancy and inadequate reimbursement for the cost of vaccine were identified [22]. From a patient view, perceived high barriers and low perceived susceptibility have been discussed as key considerations in preventative health behaviors such as vaccination [23]. A meta-analysis showed that risk perception is crucial for adult vaccination behavior [24].

In this study, we aimed 1) to evaluate if a Tdap vaccine information statement (VIS) affected overall perception and intention to receive the vaccine; 2) to explore if there was any difference between two VIS versions on overall perception and intention to receive the vaccine; and 3) to determine associations between health perceptions with Tdap vaccine receipt.

2. Methods

2.1. Study population and procedure

A randomized, prospective cohort study design was utilized in this study. The target population was women who received care at two women’s clinics in an academic medical center between May and August of 2014. Inclusion criteria to participate in this study were (1) at least 18 years of age or older at enrollment, (2) currently pregnant, and (3) able to speak, read, and communicate in English. Pregnant women who lacked the cognitive ability to make decisions concerning research participation or were unable to complete the survey independently were excluded.

A minimum sample size of 250 was determined, based on a guideline for conducting a simple structural equation modeling (SEM) approach [25]. To assure that the study would have at least 250 usable surveys, an additional 50 participants were added, which yielded a target sample of 300. A research announcement describing the project was posted in the clinics. Patients were recruited by research assistants while waiting to receive care at the two clinics. Inclusion and exclusion criteria were verbally determined and patients were informed that participation in the study was voluntary. Verbal consent was obtained and recorded for patients who met the inclusion criteria agreed to participate. A package of diapers (approximate $10 value) was offered to participants as an incentive.

Consented participants were randomly assigned, by a coin toss, into two groups before completing the first multi-part electronic survey (T1) on a tablet device. After finishing the first part of the survey, participants then reviewed either the standard CDC Tdap VIS (sVIS) or a modified version using plain language VIS (mVIS), based on the prior randomization. There was no limit on the participant VIS reading time. When the reading was complete, the second part of the survey (T2) was administered. The process of consenting, completing the pre-survey, reading the VIS, and completing the post-survey took approximately 50 minutes.

A chart review was conducted in the first week of July 2015 to identify if participants received Tdap vaccines during the recent pregnancy. The study protocol and materials were approved by the study Institutional Review Board.

2.2. Health behavior conceptual model

This study modified and integrated the Reasoned Action Approach (RAA) and Health Belief Model (HBM) framework to explore how perceptions of Tdap vaccine are associated with Tdap vaccine uptake among pregnant women. The RAA assumes that the motivation for actual behavior is reflected in individual’s intention [26]. We used the HBM components to serve as the beliefs and attitudes toward the Tdap vaccine component in the RAA. We chose to do so because the HBM components serve well for the beliefs and attitudes toward the vaccine component and helped us frame item descriptions by modifying items from the study of maternal perceptions toward flu vaccination [27]. The HBM is a framework used to identify psychological factors that predict health behavior [28]. Four constructs are central to the HBM including perceived susceptibility (beliefs concerning the likelihood of getting the condition), perceived severity (beliefs of the seriousness of the condition), perceived benefits (beliefs regarding benefits of available actions to reduce threat of the condition), and perceived barriers (negative beliefs impairing the recommended behavior).

Table 1 presents survey items used to capture participant perceptions and intention to receive the Tdap vaccine. The risk-benefit perception items of the HBM were adapted from Fridman (2011) and measured the following components: perceived susceptibility (2 items), perceived severity (2 items), perceived benefits (2 items), and perceived barriers (3 items) [27]. The RAA components adapted from Fishbein (2008) were social norms, self-efficacy and intention to receive the Tdap vaccination. Overall perception toward Tdap vaccine was measured using two items at the pre-survey (T1) and another two items on the post-survey (T2). A 4-response Likert scale (strongly disagree, disagree, agree, and strongly agree) was used for the HBM components and the overall perception items (Table 1). Each component score was calculated by summing the scores and dividing by the number of items in the component. The higher score represents a higher trait. Figure 1 presents the hypothesized relationships among HBM and RAA components and how they were associated with the decision to receive Tdap vaccine.

Table 1.

Perception items in the survey.

Health behavior perception		Item description	Pre-survey	Post-survey	Response options
Health Belief Model	Perceived susceptibility	• If I don’t get the whooping cough shot, I may catch whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• If I don’t get the whooping cough shot, my baby may catch whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
	Perceived severity	• I could die from whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• My baby could die from whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
	Perceived benefits	• Getting the whooping cough shot will protect me from getting whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• Getting the whooping cough shot while pregnant protects my baby from getting whooping cough.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
	Perceived barriers	• Getting the whooping cough shot while pregnant could harm me.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• Getting the whooping cough shot while pregnant could harm my baby.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• It’s too much trouble to get the whooping cough shot.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
Reasoned Action Approach	Social norms	• I would get the whooping cough shot if my family or friends thought I should.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• I would get the whooping cough shot since other pregnant women are getting it.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
	Self-efficacy	• I am confident I can get the whooping cough shot, even if I have to come back to the clinic.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• I am confident in my ability to get the whooping cough shot even if the shot hurts a little bit.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
	Intentions	• If you were offered the whooping cough shot during your current pregnancy, how likely are you to get the shot?	X		I would not get the shot; I probably would not get the shot; I probably get the shot; I would get the shot
		• Do you plan to get the whooping cough shot during your current pregnancy?		X	No, Probably not; probably would; Yes
Overall Perception		• Receiving the whooping cough shot is good for my health.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• Receiving the whooping cough shot will help protect my baby’s health.	X		strongly disagree; disagree; agree; strongly agree; I don’t know
		• What do you think the whooping cough shot will do for your health?		X	It will be bad for my health; It may be bad for my health; It may be good for my health; I will be good for my heatlh
		• What do you think the whooping cough shot will do for your baby’s health?		X	It will be bad for my baby’s health; It may be bad for my baby’s health; It may be good for my baby’s health; I will be good for my baby’s heatlh

Figure 1.

Health behavior conceptual framework (the vertical dash line represents the exposure of Tdap vaccination information sheet).

2.3. Materials and survey

The CDC Tdap VIS was printed directly from the CDC website [29]. The mVIS was created by the Plain Language Training Program at the study institution using guidelines from the National Institutes of Health and others for plain language as well as health literacy best practices [30–32]. The mVIS was composed at a 6^th grade reading level compared to the 10th grade measured readability of the sVIS (Fry Graph Readability Calculator) and printed in color.

The first part of the survey consisted of a 10-item assessment of health literacy (Health Literacy Skills Instrument, HLSI-SF [22]), demographic and characteristics (age, race, ethnicity, education level, gestational age, annual household income, and insurance coverage), a 9-item pre-test assessing participant knowledge of pertussis, and perceptions of the vaccine (four HBM components, perceived norms, self-efficacy, overall perception, and intention to receive the vaccine). Health literacy skills measured within the HLSI-SF included five skill dimensions consistent with the 25-item version: print prose (2 questions), print document (3 questions), oral (2 questions), print quantitative (2 questions); and internet-based information seeking (1 question) [22;23]. The HLSI-SF score ranges from 0 to 10. A score of 7 or higher represent adequate health literacy. The HLSI-SF instrument took 20-25 minutes to complete.

The 9-item knowledge test was developed by the research team. The knowledge score ranges from 0 (lowest possible score) to 9 (highest possible score). The knowledge scores were analyzed separately and were not included in this current report.

The second part of the survey was administered immediately after participants finished reading their assigned VIS. The post-assessment survey was composed of the same knowledge questions as in the pre-assessment survey and also evaluated overall perception and intention to receive the vaccine. After the survey was complete, participants were given an opportunity to ask questions and receive additional vaccine counseling if desired. Research assistants were instructed to read the survey (except the HLSI-SF) to participants if requested.

The survey was reviewed by the research team and a faculty member from another college for content validity and was administered to five pharmacy faculty and three students via electronic device to identify any technical problems. The survey was then administered to 15 participants in the clinics as a pre-testing step to identify technical problems regarding internet connection and any software issue or unclear items. The significant change we made on the survey after the pre-testing was adding the response option of “I don’t know” in most of the items. Results from the pre-testing was not used in the final analysis.

2.4. Statistical analysis

We used the Wilcoxon signed rank test to examine if reading a VIS improved overall perception and intention to receive a Tdap vaccine, before (T1) and after (T2) VIS exposure in the sample (matched paired). The Wilcoxon rank sum test was used to test differences on overall perception (T2) and intention to receive the vaccine (T2) between the group who received Tdap vaccine after the survey and the group who did not. We examined factors associated with Tdap vaccination uptake using a path analysis. A multiple logistic regression was conducted to examine associations of demographics (i.e., age, race, education level, annual household income) and a VIS version on participants’ decision to receive Tdap vaccination. Odds ratios were calculated and reported if any statistically significant differences were found.

The path analysis with either maximum likelihood or quasimaximum likelihood estimation was utilized to determine associations among HBM and RAA components as well as their impact on Tdap vaccine uptake (Figure 1), depending on distribution of the perception data. Modification indices were analyzed in order to improve the original proposed model. Fit statistics and cutoff values for good fit used in this study included (1) the model chi-square ratio of < 2, (2) the Steiger-Lind root mean square error of approximation (RMSEA) of ≤ 0.05 with its 90% confidence interval, (3) the Bentler comparative fit index (CFI) of ≥ 0.95 [25]. Standardized coefficients were provided for the final model.

Surveys with >50% missing were excluded from the analysis. Multiple imputation was used for missing data if necessary. The criterion for significance in this study was set at an alpha level of p<0.05. All statistical analyses were conducted in Stata/SE 13.0 (StataCorp LP, College Station, TX).

3. Results

A total of 561 pregnant women were approached, 238 (42%) declined, 16 (3%) did not meet the age inclusion criteria, 16 (3%) did not complete the survey due to technical problems with the electronic device (i.e., connection failure), and 291 consented and were randomized. 152 were randomized to receive the sVIS and 139 received the mVIS. Surveys from eight participants in the sVIS and four in the mVIS were excluded from the final analysis because of >50% missing items. The missing data were due to internet connection failure and/or survey participants were called from the waiting room to see their healthcare providers. 279 usable survey sets did not have any missing data and were included in the final analysis. Figure 2 shows the study enrollment.

Figure 2.

A flow diagram of recruited participants in the study.

Table 2 presents the demographics and characteristics of the study participants. Average participant age was 26.4 years (SD=5.7). Similar proportions of whites (46%) and blacks (45%) participated. Approximately 54% reported an educational level of 12^th grade or lower, and 65% reported annual household incomes less than $20,000. Most participants reported having health insurance with 73.5% on Medicaid. 8.7% did not have health insurance. Health literacy scores averaged 5 (SD=2.0) with 73% scoring less than 7, which qualifies as having “inadequate health literacy.” Average gestational age at the time of survey was 25.9 weeks (SD=9.2) and 53.4% were at 27 weeks gestational age or later. Only 13% self-reported receiving a Tdap vaccine during the current pregnancy prior to the survey. Those who had received the vaccine had an average gestation age of 34 weeks (SD=4.1).

Table 2.

Demographics and characteristics of pregnant women in the study (n=279).

Age (years), mean±SD	26.4 ± 5.7
Race (%)
White	46.6%
Black	45.2%
Others	8.2%
Number of Children, mean±SD	1.3 ± 1.3
Education (%)
Less than 12th grade	14.7%
GED or High school graduate	39.1%
Some Technical or Community College	18.3%
Graduated Technical or Community College or higher	38.9%
Annual household income (%)
<$20,000	64.5%
≥$20,000	35.5%
Insurance status (%)
Health Insurance from job/spouse’s job/parents job	13.3%
Health Insurance someone else paid for (not job-related)	4.6%
Medicaid	73.5%
No Health Insurance	8.7%
Gestational age (weeks), mean±SD	25.9 ± 9.2
Gestation age ≥27 weeks	53.4%
Had already received Tdap vaccine during current pregnancy at the time survey was conducted (%)
Yes	13.2%
No	86.8%
Health Literacy Score (HLSI-SF), mean±SD	5.0 ± 2.0
Limited health literacy (total score <7)	73%

A retrospective chart review showed documentation of Tdap vaccination in 131 participants (47%). Of the 131 participants, 102 received the Tdap vaccine for their current pregnancy after the survey was conducted. Eight participants reported receiving the Tdap vaccine during the current pregnancy, but we did not find their Tdap vaccination records. It was possible that they received the vaccine at other clinics other than the study sites. Average gestational age at Tdap vaccination per the chart review was 31.8 (SD=4.9).

Table 3 provides characteristics, scores of overall perception and intention to receive the vaccine, and vaccination rates between the two groups of participants who received a sVIS or mVIS. Both groups were similar in age, health literacy scores, and gestational age. The average scores of the overall perception and intention to receive the vaccine were similar for both VIS groups at T1 and T2. There was no significant difference in the proportions of participants who received a Tdap vaccine between the sVIS and mVIS groups (45% vs. 49%, p=0.53).

Table 3.

Characteristics, overall perception, intention to receive Tdap vaccine, and vaccination rates for the two groups.

	Standard CDC VIS group (n=144)	Modified VIS group (n=135)
Age (years), mean±SD	26.5±5.3	26.2±6.1
Health literacy score, mean±SD	5.0±2.1	5.1±1.9
Gestation age when the survey was conducted, mean±SD	26.1±5.3	26.5±8.7
Overall perception score, mean±SD
Before reading the VIS	3.1±0.5	3.0±0.5
After reading the VIS	3.5±0.6	3.4±0.7
Intention to receive the vaccine score, mean±SD
Before reading the VIS	2.9±0.9	2.8±0.8
After reading the VIS	2.9±0.9	2.9±0.9
Tdap vaccination rates when the survey was conducted, % (95% CI)	14% (8% - 20%)	13% (7% - 18%)
Tdap vaccination rates from chart review in 2015, % (95% CI)	45% (37% – 53%)	49% (40% - 57%)

Note: No statistically difference was found on any variables; VIS = vaccine information sheet; SD = standard deviation; CI = confidence interval

The overall perception scores significantly increased (p<0.001) after reading a VIS, yet the intention to receive vaccine scores remained the same (p=0.07). We compared the overall perception and intention to receive the vaccine after reading a VIS between the group who received Tdap vaccine after the survey (n=102) and the group who did not (n=148). Both overall perception (T2) and intention to receive the vaccine (T2) of the group who received the vaccine were statistically significantly higher than the group who did not (p< 0.05). Age, health literacy score, and post-test knowledge score were comparable between the two groups.

The proposed health behavior conceptual framework did not fit well (RMSE=0.08, CFI=0.78). Thus, we modified the model by adding two direct pathways from perceived benefits to intention to receive vaccine (T1) and from self-efficacy to overall perception (T1) as suggested by modified indices analysis. We also allowed correlations between error terms of the two overall perceptions (T1 and T2) as well as error terms of the two intention to receive vaccine (T1 and T2). The modified health behavioral model (Figure 3) yielded a good fit with model chi-square/df of 1.35 (chi-square=32.49; df=24), CFI of 0.98, and RMSE of 0.036 (90% CI: 0.000-0.064). All components in the modified model have correct signs. Table 4 presents standardized coefficients with robust standard errors of the revised model. Perceived benefits and self-efficacy were significantly positively associated with the overall perception (T1), while perceived barriers were significantly negatively associated with the overall perception (T1). Social norms, self-efficacy, and perceived benefits were significantly positively associated with the intention to receive vaccine (T1). Intention to receive vaccine (T1) and overall perception (T2) were positively significantly associated with the intention to received vaccine (T2). Finally, the intention to receive vaccine (T2) was significantly positively associated with decision to receive Tdap vaccine.

Figure 3.

Path analysis of the modified health behavior conceptual framework with standardized coefficients (*p<0.05, **p<0.001).

Table 4.

Standardized coefficients from a path analysis of the revised model.

	Standardized coefficient	Robust Standard Error	p-value	95% Confidence Interval
				Lower	Upper
Overall Perception (T1) ←
Perceived susceptibility	0.025	0.072	0.722	−0.116	0.167
Perceived severity	0.056	0.076	0.462	−0.093	0.204
Perceived benefits	0.315	0.093	0.001	0.133	0.497
Perceived barriers	−0.191	0.068	0.005	−0.323	−0.058
Self-efficacy	0.197	0.084	0.018	0.033	0.361
Intentions (T1) ←
Overall Perception (T1)	0.198	0.057	<0.001	0.087	0.309
Perceived benefits	0.117	0.056	0.037	0.007	0.227
Self-efficacy	0.213	0.063	0.001	0.089	0.337
Social norms	0.230	0.058	<0.001	0.116	0.345
Overall Perception (T2) ←
Overall Perception (T1)	0.691	0.165	<0.001	0.368	1.015
Intentions (T2) ←
Overall Perception (T1)	0.167	0.057	0.003	0.055	0.279
Intentions (T1)	0.865	0.110	<0.001	0.648	1.082
Tdap vaccine receipt ←
Overall Perception (T2)	0.067	0.067	0.320	−0.065	0.198
Intentions (T2)	0.223	0.062	<0.001	0.102	0.344

4. Discussion

In our study, 47% of pregnant women received Tdap vaccine, which was at least three times higher than the reports from Michigan Medicaid population and the Vaccine Safety Datalink Sites [3;4]. The Tdap vaccination rate recorded in our study was relatively lower than the reports from Healy et al. (2015) and Koepke et al. (2015) [5;6]. However, the socio-demographics of our study participants were different from the previous studies. The majority of our participants were Medicaid beneficiaries or uninsured, 45% were black, and reported lower education levels, when compared to the privately insured sample in the previous reports. Literature showed that low education level was associated with low influenza vaccination rates [33]. We believe the discrepancy in our Tdap vaccination rates compared to those of the previously mentioned studies is due, in part, to different study populations.

We did not find association between reading VIS and Tdap vaccination uptake in our study. It could be explained by inadequate health literacy in the majority of our sample. Inadequate health literacy may imply that the participants may also have inadequate overall literacy. Although, we were able to lower a reading level of the sVIS from grade 10 to grade 6 in the mVIS, the content of the mVIS was the same as the sVIS. Thus, participants with inadequate overall literacy may be not fully understand the health terms and health statistics presented in either VIS version. Although we did not find significant change in intention to receive the vaccine, we found that both Tdap VIS versions improved overall perception of the vaccine. We also found that the overall perception (T2) and intention to receive the vaccine (T2) after reading a VIS were significantly higher in the group who received Tdap vaccine after the study was conducted, when compared to the group who did not receive the vaccine.

This study was the first to examine associations of health behavior perceptions and decision to receive Tdap vaccine in pregnancy. We found that the components of HBM (perceived susceptibility, perceived severity, perceived benefits, perceived barriers) and RAA (social norms, self-efficacy) were indirectly associated to the decision to vaccinate via overall perception and intention. Thus, we are convinced that in order to promote Tdap vaccination in pregnancy it is crucial to provide information about the benefits of the vaccine and clarify that the vaccine will not cause maternal or fetal harm. These findings were consistent with the literature regarding influenza vaccination in pregnancy, human papillomavirus (HPV) vaccination, and the recent literature on Tdap vaccination [5;27;34–36]. In addition, improving access to Tdap vaccination would potentially increase Tdap vaccination rate, which is in alignment with the call for the universal use of Tdap vaccine in pregnancy [2;20].

Researchers have found that “intention to receive vaccine” is a predictor of HPV vaccine uptake [35;37], which is similar to our findings. We found that the “intention to receive vaccine” was a strong predictor of Tdap vaccination during pregnancy in our path analysis. Psychological components that increased intention to receive vaccine included perceived benefits, self-efficacy, and social norms. Pregnant women are strongly influenced by social norms and peer pressure, which implies that if health care providers reach out to family members and explain the importance of a Tdap vaccine, it is more likely to increase intention of receiving the vaccine. More importantly, a survey on influenza vaccine found that pregnant women whose providers recommended and offered influenza vaccine were the most likely to be vaccinated (73.6%), compared to women who received a recommendation but no offer of vaccination (47.9%) and women who did not receive recommendation or an offer (11.1%) [38]. We believe that health care providers play an important role in increasing Tdap vaccine uptake in pregnancy [36;39;40]. In our study, the vaccination rate increased from 13% to 47% after the date the study was conducted. A possible reason for high Tdap uptake reported in our study may be because the study research assistants spent time talking to participants about Tdap vaccine after the survey.

There are limitations to be addressed. Because our study was conducted in only one public institution and the majority of patients were of low socioeconomic level and had limited health literacy, the findings have limited generalizability to other pregnant women in different US regions. The Tdap vaccine uptake could be underestimated because participants may have received Tdap vaccines from clinics outside the study institution. The majority of our participants had inadequate health literacy, which may imply that they also had inadequate overall literacy. Thus, we did not find significant association between reading a VIS and Tdap vaccine uptake. We believe that verbal discussion between the participants and study personnel about the vaccine could have played a significant role in vaccine uptake. Lastly, a path analysis requires multivariate normality to satisfy normal theory estimators. Our perception data were distributed non-normally, but we used quasimaximum likelihood which relaxes the normality assumptions when estimating the standard errors (robust). Future studies to confirm our proposed model appears warranted.

In conclusion, this study found that reviewing a VIS could improve overall perception of the Tdap vaccine. Health education and vaccine promotion among pregnant women as well as providers may potentially increase maternal Tdap vaccine uptake. We recommend focusing on the potential benefits of the vaccine regarding both mother’s and baby’s health. Health care providers should offer the vaccine to all pregnant women to protect infants from serious infection and possible death from pertussis. While our results do not isolate the VIS as a contributing factor to vaccine uptake, providing patients with vaccine, and discussing the information, could contribute to the beliefs, knowledge and attitudes that drive vaccine behavior.

Highlights.

• A Tdap Vaccine Information Sheet (VIS) improved participants’ overall perception.

• Both VIS versions improved participants’ overall perceptions similarly.

• Intention to receive vaccine is a predictor of maternal Tdap vaccination.

• Tdap vaccination rates increased from 13% to 47% after participants read a VIS and discussed with the research team.