Volume 162, Issue 1 p. 105-115
CLINICAL ARTICLE
Open Access

Pregnancy vaccination predictive factors and uptake profiles among Italian women: A cross-sectional survey study on a large population

Amerigo Ferrari

Corresponding Author

Amerigo Ferrari

Institute of Management, MeS (Management and Health) Laboratory, Sant'Anna School of Advanced Studies, Pisa, Italy

Correspondence

Amerigo Ferrari, Institute of Management, MeS (Management and Health) Laboratory, Sant'Anna School of Advanced Studies, Via San Zeno 2, 56127 Pisa, Italy.

Email: [email protected]

Search for more papers by this author
Giaele Moretti

Giaele Moretti

Institute of Management, MeS (Management and Health) Laboratory, Sant'Anna School of Advanced Studies, Pisa, Italy

Search for more papers by this author
Ilaria Corazza

Ilaria Corazza

Institute of Management, MeS (Management and Health) Laboratory, Sant'Anna School of Advanced Studies, Pisa, Italy

Search for more papers by this author
Paolo Mannella

Paolo Mannella

Division of Obstetrics and Gynecology, Department of Clinical and Experimental Medicine, University of Pisa, Pisa, Italy

Search for more papers by this author
Tommaso Simoncini

Tommaso Simoncini

Division of Obstetrics and Gynecology, Department of Clinical and Experimental Medicine, University of Pisa, Pisa, Italy

Search for more papers by this author
Manila Bonciani

Manila Bonciani

Institute of Management, MeS (Management and Health) Laboratory, Sant'Anna School of Advanced Studies, Pisa, Italy

Meyer Center for Health and Happiness, A. Meyer Children's Hospital, Florence, Italy

Search for more papers by this author
First published: 24 April 2023
Citations: 2

Abstract

Objectives

To assess influenza and Tdap (tetanus, diphtheria, pertussis) vaccine coverage during pregnancy, explore key socioeconomic and maternity pathway-related predictors of vaccination, and detect specific patterns of vaccination uptake.

Methods

The authors cross-sectionally analyzed self-reported data obtained from the systematic survey on the maternity pathways of Tuscany. They selected all pregnant women that completed from March 2019 to June 2022 the third-trimester questionnaire (n = 25 160), which included two dichotomous items on influenza and Tdap vaccination, as well as socioeconomic and pathway-related questions. Multilevel logistic models were performed to assess vaccination predictors and cluster analysis to identify vaccination patterns.

Results

Vaccination coverage was higher for pertussis (56.5%) than for influenza (18.9%). The main predictors of vaccination were high socioeconomic status, attending private gynecologists, and receiving vaccine information. Three clusters were identified: cluster 1 included women receiving both Tdap and influenza vaccines; cluster 2 included women receiving no vaccinations; and cluster 3 included women receiving only the pertussis vaccine. Although women from cluster 3 were of middle to low education status, vaccine information was the main adherence determinant also in this group.

Conclusions

Health workers and policymakers should focus on groups of pregnant women less prone to vaccination to promote vaccination information and encourage wider uptake and coverage.

1 INTRODUCTION

Immunization during pregnancy has been shown to be an important tool for providing protection for both mother and fetus against certain pathogens.1 The Italian National Vaccine Prevention Plan 2017–2019 recommended that pregnant women receive the combined Tdap (tetanus, diphtheria, pertussis) vaccine ideally between the 27th and 36th gestational weeks and be vaccinated against influenza before or during the flu season.

Indeed, while pertussis infection may be mild or asymptomatic in adults and older children, it can cause severe morbidity and mortality in infants too young to be vaccinated. Maternal immunization from pertussis allows the transfer of maternal antibodies to the fetus, thus ensuring protection of the infant in the first few months of life.2 On the other hand, influenza in pregnancy can cause serious complications for both mother and infant, such as cardiopulmonary hospitalizations and an increased risk of stillbirth, neonatal death, preterm delivery, and congenital heart disease.3

Despite the availability of free vaccines in Italy for pregnant women and numerous studies proving their safety and efficacy,4, 5 Italy has very low pregnancy vaccination rates mainly against influenza but also against pertussis, as previously demonstrated.6, 7 One of the main causes is a general unawareness about maternal immunization and its benefits for pregnant women, along with poor information, lack of recommendations by health professionals, or other factors related to the maternity pathway in general.8 In addition, vaccine adherence often depends on the socioeconomic background of pregnant women.9

A multicenter study from Italy10 showed that vaccine coverage rates in the 2018 to 2019 season for pertussis and influenza were 61% and 15%, respectively. Particularly, influenza vaccine coverage was still far behind coverage rates in neighboring countries and the average European rates (25%).

Several studies exploring socioeconomic and pathway determinants of vaccine adherence through surveys have been published,11-13 also in the Italian context.14 However, these studies have used medium-sized populations and only limited numbers of socioeconomic and pathway-related variables. In contrast, this study used a very large study population and a wide number of covariates, obtaining data from the systematic and longitudinal survey on the maternity pathways of Tuscany, Italy.15-17 We aimed to: (1) compute the influenza and Tdap vaccine uptake rate, (2) explore key socioeconomic and pathway-related predictors of influenza and Tdap vaccine uptake, and (3) identify subgroups of respondents with specific patterns of vaccination coverage.

2 METHODS

This is a cross-sectional survey study exploring patient-reported vaccine coverage during pregnancy and the barriers and facilitators to access to vaccination. As an observational study, it follows the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology)/RECORD (Reporting of Studies Conducted Using Observational Routinely Collected Data) guidelines. The study setting was the Italian Region of Tuscany, responsible for the health care services delivered to 3.7 million inhabitants. Tuscany is divided into three Local Health Authorities and 26 health districts and receives approximately 6% of the national health care fund (7 billion euros in 2019).

We used the same data source that we used in a previous paper.16 This source was the systematic survey on the maternity pathways of Tuscany, launched by Sant'Anna School of Advanced Studies and Tuscany Region Health Authority and embedded into the app hAPPyMamma (which includes the digital Pregnancy Booklet in Tuscany).15 The eligible population for the survey consisted of the general population of pregnant women receiving the Pregnancy Booklet in Tuscany. The booklet collects the list and due dates for all pregnancy examinations that are provided free of charge by the Regional Health Service of Tuscany. Pregnant women are invited to participate and, if they accept, are recruited via online invitation. Questionnaires are administered at eight different timepoints from the beginning of pregnancy up to 1 year post partum. As previously explained,16 informed consent and ethics approval were not necessary for using such data for research purposes.

This study comprised all women who completed the third-trimester questionnaire (T3g) from March 2019 to June 2022 (n = 25 160). For comparison, approximately 70 000 women gave birth in Tuscany from January 2019 to December 2022. The T3g questionnaire included dichotomous questions about vaccination during pregnancy, specifically on the influenza and pertussis vaccines, which in Italy are licensed only in combination with tetanus and diphtheria as the Tdap vaccine. It also included several questions about the maternity care pathway utilization and the vaccine information each woman received during pregnancy. Conversely, sociodemographic and childbirth data were obtained from the T0g (beginning of pregnancy) and T0p (hospitalization for childbirth) questionnaires, respectively, by merging the data sets through anonymized encrypted identifiers.

2.1 First aim

We used the dichotomous variables for Tdap and influenza vaccination to compute vaccine uptake rates as reported by respondents. For this aim, we used a two-step approach as a robustness check. Particularly, we obtained the general population of pregnant women during the 2-year period 2020 to 2021 from the Birth Assistance Certificate (n = 43 264), an administrative database including all women giving birth each year in Tuscany. Then, we selected from the entire survey population (n = 25 160) the subgroup of women who were beyond 40 weeks of gestation in 2020 to 2021 (n = 18 363). We stratified both the general population of pregnant women and the study subgroup by age (three-level variable), nationality (two-level variable), and education (three-level variable), resulting in 18 strata. We then calculated the number of women per stratum for each health district, both among the general population and in the study subgroup. After dividing the two values to obtain the “weight” variable, we computed vaccination uptake rates weighted on the general population.

2.2 Second aim

We used the dichotomous variables for vaccination uptake as outcomes. We ran multivariate multilevel logistic regression models (melogit command in Stata Software [StataCorp LLC, College Station, TX, USA]) to explore which women's characteristics and maternity pathway attributes were associated with higher odds of having pertussis and influenza vaccination. The variable “health district of residence” was used as the top-level grouping variable. The covariates were classified as follows: (1) respondents' sociodemographic features; (2) maternity care pathway variables; and (3) vaccine information during pregnancy. Also, for this objective, we adopted a two-step approach, running the same models described above, with the same approach, but weighted on the general population of pregnant women.

2.3 Third aim

Cluster analysis18 was performed in SAS software (SAS Institute, Cary, NC, USA) to identify subgroups of patients with distinct self-reported vaccination coverage patterns. The variables used to detect clusters were the dichotomous variables for pertussis and influenza vaccines. After running cluster analysis, we transformed the cluster variables into dichotomous variables, one for each cluster. Then, we used multilevel logistic regressions to investigate which characteristics were significantly associated with each cluster profile, using the dichotomous “cluster” variables as outcomes. The models were built as already described for the second aim.

3 RESULTS

Our study population consisted of 25 160 respondents. Their sociodemographic features are shown in Table 1. Most women were aged between 30 and 39 years (69.4%), highly educated (50.4%), and Italian (92.0%). Compared with the general population (Table S2), our study population included higher proportions of women with such characteristics. Moreover, 61.0% of participants were primiparous, while 98.5% of women had only one child. A total of 69.5% of women were of normal weight. Women who reported planning pregnancy were 56.0%, and 19.0% of the population had a high-risk pregnancy. During pregnancy, most women did not smoke (93.8%) and took folic acid (95.3%). Finally, 83.6% of women were employed, with 70.5% of the study cohort reporting a good income. See Table S1 for details on the maternity pathway-related variables.

TABLE 1. Sociodemographic features of our respondents.
Full study population (n = 25 160) Percentage (n)
Age
16–29 years 20.03 (5040)
30–39 years 69.38 (17453)
≥40 years 10.59 (2663)
Missing 4
Nationality
Italian 92.03 (23154)
Non-Italian 7.97 (2006)
Education level
Middle school or less 9.79 (2463)
High school 39.80 (10014)
University 50.41 (12683)
Parity
Primigravidae 61.00 (15348)
Multiparous 39.00 (9812)
Weight class
Normal 69.45 (17473)
Underweight 6.97 (1754)
Overweight 16.46 (4142)
Obesity 7.12 (1791)
Twin pregnancy
No 98.46 (24771)
Yes 1.54 (387)
Missing 2
High-risk pregnancy
No 80.97 (20371)
Yes 19.03 (4787)
Missing 2
Smoking during pregnancy
No 93.82 (23603)
Yes 6.18 (1554)
Missing 3
Folate intake during pregnancy
Yes 95.33 (23984)
No 4.24 (1066)
Never heard 0.44 (110)
Income level
High 70.53 (17743)
Medium 25.19 (6338)
Low 4.28 (1076)
Missing 3
Job
Employed 83.59 (21032)
Unemployed 5.95 (1496)
Student 1.67 (421)
Housewife 8.79 (2211)
Planned pregnancy
Yes 56.02 (14094)
Not excluded 33.66 (8470)
No 10.32 (2596)

3.1 First aim

The vaccination uptake rate for pertussis was 56.5% (n = 14 222) and for influenza was 18.9% (n = 4751). Only 15.6% (n = 3919) of respondents received both the influenza and pertussis vaccines. Furthermore, the weighted vaccine coverage was 49.7% for pertussis and 16.6% for influenza.

3.2 Second aim

As shown in Table 2, women who received the pertussis vaccine were most often aged 30 to 39 years, Italian, primigravidae, with good income status, and employed. They also reported that they did not smoke during pregnancy. They had planned pregnancy and visited gynecologists privately. Their examinations were booked directly by health workers. Moreover, they obtained more often vaccine information from counseling centers, the general practitioner (GP), other doctors, or the Pregnancy Booklet. Weighted models showed that the effect of smoking in pregnancy and visit setting lost significance, while educational level acquired significance, with high education as a predisposing factor.

TABLE 2. Multilevel models for pertussis vaccine uptake.
Pertussis vaccine uptake Full population Weighted subgroup
OR SE OR SE
Sociodemographic
Age 30–39 vs. <30 years 1.33 *** 0.10 1.36 *** 0.13
Age > 39 vs. <30 years 1.00 0.11 1.01 0.16
Non-Italian vs. Italian 0.65 ** 0.09 0.68 * 0.11
Medium vs. high education 0.93 0.06 0.99 0.07
Low vs. high education 0.91 0.12 0.75 * 0.08
Multiparous vs. primigravida 0.60 *** 0.04 0.61 *** 0.06
Underweight vs. normal weight 0.85 0.09 0.93 0.10
Overweight vs. normal weight 1.01 0.08 1.01 0.09
Obese vs. normal weight 1.13 0.14 1.20 0.18
Twin vs. single pregnancy 0.98 0.27 0.85 0.23
High-risk vs. low-risk pregnancy 1.12 0.09 1.22 0.17
Smoking during pregnancy vs. not 0.78 * 0.10 0.75 0.16
Folate not taken vs. taken 0.84 0.13 0.80 0.18
Never heard of folate vs. taken 0.71 0.37 1.01 0.56
Medium vs. good income level 0.87 * 0.06 0.78 * 0.08
Poor vs. good income level 0.77 0.13 0.65 0.19
Unemployed vs. employed 0.87 0.11 0.99 0.16
Student vs. employed 0.62 * 0.14 0.38 ** 0.14
Housewife vs. employed 0.95 0.14 0.94 0.21
Pregnancy not excluded vs. planned 0.82 *** 0.05 0.85 ** 0.06
Pregnancy not planned vs. planned 0.64 *** 0.07 0.64 *** 0.09
Maternity pathway attributes
Discomfort during pregnancy vs. not 1.00 0.06 1.08 0.11
Medium vs. not difficult access to services 1.14 0.10 0.90 0.08
Difficult vs. not difficult access to services 0.90 0.13 0.69 0.13
Some delays in examinations vs. not 1.07 0.07 1.08 0.08
Frequent delays in examinations vs. not 1.16 0.23 1.09 0.25
Medium involvement in choices vs. low 1.05 0.08 0.99 0.10
High involvement in choices vs. low 1.01 0.08 0.96 0.10
Hospital visits vs. counseling centers 1.05 0.10 1.01 0.13
Private visits vs. counseling centers 1.22 * 0.11 1.20 0.16
Gynecologist vs. midwife 1.52 *** 0.16 1.61 ** 0.28
Gynecologist and midwife vs. midwife 1.21 0.13 1.21 0.20
Another specialist vs. midwife 1.38 0.89 1.04 0.64
Sessions of BAC 1.00 0.00 1.00 0.00
Examination booking explained vs. made by health workers 0.80 * 0.08 0.76 * 0.09
Examination booking never heard vs. made by workers 1.37 0.41 1.70 0.61
Two to 5 vs. >5 close people during pregnancy 0.98 0.08 1.05 0.10
One vs. >5 close people during pregnancy 0.85 0.09 0.88 0.09
Zero vs. >5 close people during pregnancy 0.94 0.23 0.85 0.24
Enough vaccine information during BAC vs. a lot 1.10 0.08 1.13 0.11
Poor vaccine information during BAC vs. a lot 1.11 0.09 1.16 0.12
Vaccine information
Vaccine info from counseling centers vs. not 3.12 *** 0.21 3.20 *** 0.29
Vaccine info from the GP vs. not 2.45 *** 0.20 2.83 *** 0.34
Vaccine info from other physicians vs. not 3.44 *** 0.28 3.83 *** 0.44
Vaccine info from the Pregnancy Booklet vs. not 2.03 *** 0.22 1.91 ** 0.36
Vaccine info from hAPPyMamma vs. not 1.35 0.26 1.48 0.30
Constant 0.55 ** 0.11 0.47 ** 0.13
Health district variance 0.50 0.15 0.59 0.13
  • Note: Bold values indicate significance. Abbreviations: BAC, birth-accompanying course; GP, general practitioner; OR, odds ratio; SE, standard error.
  • * P < 0.05
  • ** P < 0.01
  • *** P < 0.001.

As for the influenza vaccine (Table 3), women who reported receiving this vaccine were more often aged 30 to 39 years and highly educated. They reported a high-risk pregnancy, discomfort during pregnancy, and difficulty in accessing health services. They were followed privately by a gynecologist and significantly more often received vaccine information from all sources analyzed. Interestingly, in the weighted models, age, level of risk of pregnancy, discomfort during pregnancy, and difficulty in booking examinations lost statistical significance, while being employed became a significant predictor of vaccination.

TABLE 3. Multilevel models for influenza vaccine uptake.
Influenza vaccine uptake Full population Weighted subgroup
OR SE OR SE
Sociodemographic
Age 30–39 vs. <30 years 1.21 * 0.10 1.26 0.20
Age > 39 vs. <30 years 1.04 0.13 1.08 0.19
Non-Italian vs. Italian 0.86 0.14 0.82 0.15
Medium vs. high education 0.74 *** 0.05 0.72 *** 0.06
Low vs. high education 0.66 ** 0.10 0.81 0.12
Multiparous vs. primigravida 1.02 0.08 1.21 0.13
Underweight vs. normal weight 1.09 0.12 1.38 0.25
Overweight vs. normal weight 1.11 0.09 0.91 0.09
Obese vs. normal weight 0.91 0.12 1.10 0.21
Twin vs. single pregnancy 1.15 0.29 1.07 0.30
High-risk vs. low-risk pregnancy 1.21 * 0.09 1.13 0.12
Smoking during pregnancy vs. not 0.84 0.12 0.77 0.15
Folate not taken vs. taken 1.07 0.18 1.53 0.41
Never heard of folate vs. taken 1.24 0.74 1.20 0.82
Medium vs. good income level 0.96 0.07 1.06 0.12
Poor vs. good income level 0.83 0.17 0.76 0.17
Unemployed vs. employed 1.05 0.14 1.13 0.20
Student vs. employed 0.76 0.21 0.38 ** 0.14
Housewife vs. employed 0.98 0.17 1.29 0.27
Pregnancy not excluded vs. planned 1.00 0.06 0.93 0.07
Pregnancy not planned vs. planned 0.79 0.10 0.93 0.17
Maternity pathway attributes
Discomfort during pregnancy vs. not 1.14 * 0.07 0.99 0.09
Medium vs. not difficult access to services 0.92 0.08 0.88 0.10
Difficult vs. not difficult access to services 0.70 * 0.12 1.08 0.40
Some delays in examinations vs. not 0.92 0.06 0.85 0.08
Frequent delays in examinations vs. not 1.10 0.23 0.72 0.23
Medium involvement in choices vs. low 0.97 0.07 1.02 0.08
High involvement in choices vs. low 0.91 0.07 0.95 0.10
Hospital visits vs. counseling centers 1.08 0.11 1.13 0.15
Private visits vs. counseling centers 1.15 0.11 1.17 0.15
Gynecologist vs. midwife 1.34 ** 0.16 1.47 ** 0.21
Gynecologist and midwife vs. midwife 1.32 * 0.16 1.32 * 0.18
Another specialist vs. midwife 1.00 0.61 0.80 0.66
Sessions of BAC 1.00 0.00 1.00 0.00
Examination booking explained vs. made by health workers 0.94 0.11 1.08 0.17
Examination booking never heard vs. made by workers 1.01 0.32 1.10 0.32
Two to 5 vs. >5 close people during pregnancy 1.12 0.09 1.12 0.12
One vs. >5 close people during pregnancy 1.09 0.12 1.09 0.18
Zero vs. >5 close people during pregnancy 1.02 0.26 1.22 0.42
Enough vaccine information during BAC vs. a lot 0.82 ** 0.06 0.86 0.09
Poor vaccine information during BAC vs. a lot 1.02 0.08 1.00 0.12
Vaccine information
Vaccine info from counseling centers vs. not 1.22 ** 0.08 1.54 *** 0.12
Vaccine info from the GP vs. not 1.76 *** 0.12 2.01 *** 0.20
Vaccine info from other physicians vs. not 1.49 *** 0.10 1.44 *** 0.14
Vaccine info from the Pregnancy Booklet vs. not 1.45 *** 0.13 1.42 ** 0.19
Vaccine info from hAPPyMamma vs. not 1.51 ** 0.24 1.47 * 0.29
Constant 0.12 0.02 0.10 0.03
Health district variance 0.20 0.07 0.25 0.06
  • Note: Bold values indicate significance. Abbreviations: BAC, birth-accompanying course; GP, general practitioner; OR, odds ratio; SE, standard error.
  • * P < 0.05
  • ** P < 0.01
  • *** P < 0.001.

3.3 Third aim

We identified three clusters (Table S3). Cluster 1 included 18.9% of women from the study population, who all reported receiving the pertussis vaccination, and—mostly (82.5%)—also the influenza vaccine. Women from cluster 2 (40.2%) received no vaccinations. Finally, all women from cluster 3 (40.9%) received the pertussis vaccine but not the influenza vaccine.

As shown in Table 4, women from cluster 1 (all vaccinated) were significantly more often aged 30 to 39 years and with a high education level. They were more likely to report high-risk pregnancy and discomfort during pregnancy. They mainly attended gynecologists, regardless of the visit setting, and obtained a lot of information on vaccination during birth-accompanying courses (BACs). They received vaccination information significantly more often from counseling centers, the GP, other physicians, or the Pregnancy Booklet.

TABLE 4. Multilevel models for cluster profiles.
Clusters Cluster 1 Cluster 2 Cluster 3
OR SE OR SE OR SE
Sociodemographic
Age 30–39 vs. <30 years 1.21 * 0.10 0.73 *** 0.06 1.13 0.07
Age > 39 vs. <30 years 1.05 0.13 0.98 0.11 1.02 0.10
Non-Italian vs. Italian 0.87 0.14 1.49 ** 0.21 0.80 0.11
Medium vs. high education 0.74 *** 0.05 1.10 0.07 1.15 ** 0.06
Low vs. high education 0.65 ** 0.10 1.09 0.14 1.25 * 0.15
Multiparous vs. primigravida 1.02 0.08 1.60 *** 0.12 0.69 *** 0.05
Underweight vs. normal weight 1.10 0.12 1.16 0.13 0.84 0.08
Overweight vs. normal weight 1.10 0.09 0.97 0.08 0.95 0.06
Obese vs. normal weight 0.90 0.12 0.97 0.12 1.10 0.11
Twin vs. single pregnancy 1.19 0.30 1.13 0.31 0.82 0.19
High-risk vs. low-risk pregnancy 1.20 * 0.09 0.92 0.07 0.93 0.06
Smoking during pregnancy vs. not 0.84 0.12 1.35 * 0.18 0.89 0.10
Folate not taken vs. taken 1.06 0.18 1.16 0.18 0.83 0.12
Never heard of folate vs. taken 1.24 0.73 1.19 0.63 0.70 0.34
Medium vs. good income level 0.96 0.07 1.17 * 0.08 0.91 0.06
Poor vs. good income level 0.88 0.17 1.25 0.22 0.93 0.15
Unemployed vs. employed 1.05 0.14 1.12 0.14 0.88 0.10
Student vs. employed 0.76 0.21 1.62 * 0.36 0.77 0.16
Housewife vs. employed 0.99 0.17 1.00 0.15 0.98 0.13
Pregnancy not excluded vs. planned 1.00 0.06 1.19 ** 0.07 0.89 * 0.05
Pregnancy not planned vs. planned 0.79 0.10 1.54 *** 0.17 0.84 0.09
Maternity pathway attributes
Discomfort during pregnancy vs. not 1.14 * 0.07 0.98 0.06 0.93 0.05
Medium vs. not difficult access to services 0.93 0.08 0.88 0.08 1.14 0.09
Difficult vs. not difficult access to services 0.70 * 0.12 1.12 0.17 1.14 0.15
Some delays in examinations vs. not 0.92 0.06 0.98 0.07 1.07 0.06
Frequent delays in examinations vs. not 1.10 0.22 0.87 0.18 1.00 0.18
Medium involvement in choices vs. low 0.97 0.07 0.96 0.07 1.06 0.07
High involvement in choices vs. low 0.90 0.07 1.06 0.09 1.02 0.07
Hospital visits vs. counseling centers 1.08 0.11 0.90 0.09 1.01 0.09
Private visits vs. counseling centers 1.15 0.11 0.82 * 0.07 1.04 0.08
Gynecologist vs. midwife 1.34 * 0.16 0.63 *** 0.07 1.18 0.12
Gynecologist and midwife vs. midwife 1.33 * 0.16 0.78 * 0.08 1.01 0.10
Another specialist vs. midwife 0.99 0.60 0.58 0.41 1.57 0.83
Sessions of BAC 1.00 0.00 1.00 0.00 1.00 0.00
Examination booking explained vs. made by health workers 0.93 0.11 1.17 0.12 0.92 0.08
Examination booking never heard vs. made by workers 1.00 0.31 0.89 0.26 1.09 0.28
Two to 5 vs. >5 close people during pregnancy 1.11 0.09 1.00 0.08 0.92 0.06
One vs. >5 close people during pregnancy 1.08 0.12 1.19 0.12 0.83 * 0.08
Zero vs. >5 close people during pregnancy 1.00 0.26 1.24 0.31 0.83 0.18
Enough vaccine information during BAC vs. a lot 0.83 * 0.06 0.96 0.07 1.19 ** 0.08
Poor vaccine information during BAC vs. a lot 1.04 0.08 0.88 0.07 1.07 0.07
Vaccine information
Vaccine info from counseling centers vs. not 1.23 ** 0.08 0.32 *** 0.02 1.89 *** 0.11
Vaccine info from the GP vs. not 1.76 *** 0.12 0.34 *** 0.03 1.24 ** 0.08
Vaccine info from other physicians vs. not 1.49 *** 0.10 0.29 *** 0.02 1.61 *** 0.10
Vaccine info from the Pregnancy Booklet vs. not 1.45 *** 0.13 0.44 *** 0.05 1.17 0.10
Vaccine info from hAPPyMamma vs. not 1.50 * 0.24 0.69 0.14 0.79 0.12
Constant 0.12 *** 0.02 1.66 * 0.35 0.53 *** 0.09
Health district variance 0.19 0.07 0.52 0.16 0.19 0.07
  • Note: Bold values indicate significance. Abbreviations: BAC, birth-accompanying course; GP, general practitioner; OR, odds ratio; SE, standard error.
  • * P < 0.05
  • ** P < 0.01
  • *** P < 0.001.

Women in cluster 2 (nonvaccinated) were more often younger than 30, non-Italian, and multiparous. They more frequently stated that they smoked during pregnancy. They had an average but not good income level. They were students more than employed women and they had less frequently planned pregnancy. They made their visits during pregnancy less often to private specialists than to free counseling centers and they referred more often only to a midwife. Women in this cluster reported receiving vaccine information from all sources analyzed significantly less than the other women.

Finally, women in cluster 3 (Tdap-vaccinated) had a low/middle education and were more often primigravidae. They had planned pregnancy and had more than five close people assisting them during pregnancy. Despite saying that they obtained enough but not a lot of vaccine information during BACs, they reported receiving vaccine information more frequently from family counseling centers, the GP, or other physicians.

4 DISCUSSION

In this study, we used data reported by a large population of pregnant women to analyze predisposing factors to pregnancy vaccination and identify vaccination patterns. We found that the self-reported vaccination coverage for pertussis (56.5%, weighted 49.7%) was higher than for influenza (18.9%, weighted 16.6%), in line with the literature.19 Indeed, the maternal influenza vaccine uptake rates have been reported to range between 6% and 15% in previous studies from Italy,3, 7, 10, 14 which are below the European average (25%) and similar to those found in Spain (7%–12%),20 while data from France are more variable (1%–35%).21, 22 On the other hand, pertussis vaccine uptake has been reported to be 61% in Italy,14 50% to 79% in Spain,20 54% in the United States,23 and 64% in the United Kingdom.24

The slightly higher influenza vaccine uptake observed in our study population may also depend on the higher proportion of Italian, highly educated, and 30- to 39-year-old women than in the general population of pregnant women (Table S2). Such characteristics are known to be predictors of vaccination uptake25 and our findings supported this evidence. In any case, our study confirmed that influenza vaccination coverage during pregnancy remains too low in Italy, despite being recommended for more than a decade.

The low vaccination adherence found in our study population could also be due to the effect of the coronavirus disease 2019 (COVID-19) pandemic, which hit Italy right in the middle of the study period. Indeed, many studies have shown increased vaccination hesitancy during the pandemic,26 especially among pregnant women who constitute a population particularly susceptible to potential adverse effects. This undesired influence seems more pronounced for the influenza than for the pertussis vaccine.27 Moreover, such a trend has been detected for pertussis and influenza vaccines, but it has also been shown that pregnant women who have greater hesitation to influenza and pertussis vaccination also have a greater propensity to refuse anti–COVID-19 vaccination.28 This evidence reinforces the need to promote vaccination by providing comprehensive information and recommendations to pregnant women.29

Our study also shows other predictive factors of vaccination in pregnancy. Multiparity was associated with lower odds of receiving the Tdap vaccine—in line with the literature30—but not the influenza vaccine. Pregnant women who smoked received the pertussis vaccine less often, probably due to less awareness of preventive behaviors to take during pregnancy. Similarly, the odds of receiving the Tdap vaccine were lower in women with a poor income level and who had not planned to become pregnant. All of these effects were not observed for the influenza vaccine, for which the only predisposing characteristic was high-risk pregnancy.

Visiting a gynecologist privately and having all examinations directly booked by health care workers increased Tdap vaccine uptake, while for influenza the effect of visit setting was absent. This may be related to the fact that gynecologists are the main source of information about pregnancy vaccination.30 Not surprising, we found that receiving vaccine information from counseling centers, the GP, other physicians, or by reading the Pregnancy Booklet instructions increased the odds of taking the Tdap and influenza vaccines, with a specifically pronounced effect on pertussis, as shown by the higher odds ratio values.

Clusters 1 and 2 seem to support these results. Indeed, the characteristics of women from cluster 1 (all vaccinated) almost reflected the predisposing factors to vaccination described above. Indeed, most women from cluster 1 received both the influenza and pertussis vaccine, probably because they were more aware of preventive behaviors to adopt during pregnancy, participated more in the maternity pathway utilization, and had high-risk conditions.

In contrast, women from cluster 2 (nonvaccinated) possessed all factors associated with lower odds of being vaccinated, and thus had no vaccination at all. These women might have given less attention to prevention, probably because, being non-Italian, they understood less information on the importance of vaccination due to linguistic barriers. Also, women from cluster 2 visited more public counseling centers than private gynecologists during pregnancy. This evidence raises the need for wider and proactive vaccine information and recommendation from health providers working in the public sector.

Finally, women from cluster 3 (Tdap-vaccinated) were low/medium educated but primigravidae, and they had planned pregnancy and more than five close people helping them throughout. Although they had enough but not completely satisfying information during BACs, they more often received vaccination information from all sources analyzed. Presumably, these women were advised to specifically take the pertussis vaccine but avoid the influenza vaccine, and they trusted the advice.

It is known that pregnant women are more predisposed to pertussis than to influenza vaccination, since they more often believe that influenza vaccination is unnecessary, are less aware of the benefits to the newborn, and wrongly think that this vaccine may result in risks to the newborn.14, 19 However, it is also known that recommendation by health workers is the main determinant of vaccination uptake.14 In this study, although we had no specific questions about vaccine recommendation, we demonstrated that women who were privately followed by gynecologists and had received vaccination information were more likely to be vaccinated. Even women with medium/low education level who had received vaccination information were specifically vaccinated against influenza and pertussis.

A limitation of this study is that our population was not fully representative of the general population of pregnant women due to selection bias. To mitigate this limitation, we weighted our models, confirming our results. Another limitation is that the data are self-reported and may contain unverifiable information. Also, we could not verify the actual vaccine coverage from administrative data since the databases recording vaccines are organized by drugs dispensed, not by individuals, and could not be cross-referenced with the birth database. In addition, we had no information on COVID-19 vaccination as we lacked specific questions. Finally, we did not consider seasonality, which may affect influenza vaccine uptake.

Nevertheless, the strength of this study is the simultaneous assessment of numerous factors potentially influencing vaccination uptake through a large study population obtained from a validated data source. This study can also provide guidance to health workers and policymakers to promote broader recommendation of vaccination in pregnancy especially in less predisposed subgroups and based on the different population's needs, provide more information about vaccination and the benefits for newborns, implement policies that stimulate vaccine uptake among the most fragile segments of the population, and generally increase pregnancy vaccine coverage.

In conclusion, we found that vaccination coverage in Tuscany against pertussis was higher than against influenza. Furthermore, this study confirmed the results of previous studies on predisposing factors to vaccination. Women with lower socioeconomic status had lower vaccination adherence, but, when they received information about vaccination and its benefits, they were more likely to vaccinate against pertussis and influenza. Therefore, health care providers and policymakers should promote vaccination information and recommendation through different strategies that meet the different needs of the population, focusing mainly on groups less prone to vaccination to improve vaccine uptake in pregnancy.

AUTHOR CONTRIBUTIONS

Amerigo Ferrari and Giaele Moretti participated in the design, methodology, implementation, conduct, monitoring, analysis, and writing of the study. Ilaria Corazza participated in the interpretation of the study and writing of the Discussion section. Professors Tommaso Simoncini and Paolo Mannella coordinated the collaboration between Pisa University and Sant'Anna School, reviewed the article drafts, and approved the final version. Manila Bonciani coordinated the design and development of the systematic survey on the maternity pathways of Tuscany and the web app hAPPyMamma, and participated in designing the study design, interpreting the results, and writing, reviewing, and editing the manuscript.

ACKNOWLEDGMENTS

The first author thanks both the MeS Laboratory of Sant'Anna School and the Obstetrics and Gynecology Division of Pisa University for the successful collaboration and support given for the study. He also thanks the Regional Health Authority of Tuscany and all health professionals of the Tuscan health care system, who actively promoted participation in the systematic survey. Without their work, the collection of data used for this study would not have been possible. Open Access Funding provided by Scuola Superiore Sant’Anna within the CRUI-CARE Agreement.

    FUNDING INFORMATION

    The data source and the publication charges were funded as part of the research activities of the MeS Lab of Sant'Anna School of Advanced Studies. Such activities were funded by the Tuscany Region Health Authority under the collaboration agreement with Sant'Anna School. Tuscany Region had no role in the study design, data collection and analysis, or writing and submitting of the article for publication.

    CONFLICT OF INTEREST STATEMENT

    No conflicts of interest declared by the authors.

    DATA AVAILABILITY STATEMENT

    Research data are not shared.