Determinants of incomplete vaccination in children at age two in France: results from the nationwide ELFE birth cohort

Abstract

Incomplete vaccination in the pediatric population is a growing public health issue in high-income countries, but its determinants are poorly understood. Their identification is necessary to design target actions that can improve vaccination uptake. Our aim was to assess the determinants of incomplete vaccination in two-year-old children in France. Among the 18,329 children included in the 2011 ELFE French nationwide population-based birth cohort, we selected those for whom vaccination status was available at age two years. Incomplete vaccination was defined as ≥ 1 missing dose of recommended vaccines. Potential determinants of incomplete vaccination were identified by using logistic regression, taking into account attrition and missing data. Of the 5,740 (31.3%) children analyzed, 46.5% (95% confidence interval [CI] 44.7–48.0) were incompletely vaccinated. Factors independently associated with incomplete vaccination were having older siblings (adjusted odds ratio 1.18, 95% CI [1.03–1.34] and 1.28 [1.06–1.54] for one and ≥ 2 siblings, respectively, vs. 0), residing in an isolated area (1.92 [1.36–2.75] vs. an urban area), parents not following health recommendations or using alternative medicines (1.81 [1.41–2.34] and 1.23 [1.04–1.46], respectively, vs. parents confident in institutions and following heath recommendations), not being visited by a maternal and child protection service nurse during the child’s first two months (1.19 [1.03–1.38] vs. ≥ 1 visit), and being followed by a general practitioner (2.87 [2.52–3.26] vs. a pediatrician).

Conclusions: Incomplete vaccination was highly prevalent in the studied pediatric population and was associated with several socio-demographic, parental, and healthcare service characteristics. These findings may help in designing targeted corrective actions.

What is Known: • Incomplete vaccination in the pediatric population is a growing public health issue in high-income countries. • The partial understanding of the determinants of incomplete vaccination precludes the design of effective targeted corrective actions.

What is New: • High prevalence of incomplete vaccination at age two years in France. • Incomplete vaccination was independently associated with several socio-demographic, parental, and healthcare service characteristics.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00431-022-04733-z.

Introduction

Vaccine hesitancy and incomplete vaccination issues were recently highlighted by the COVID-19 pandemic [1] but are a three-decade-old problem in high-income countries (HICs) [2] and primarily affect children. Depending on the age group considered, one-third to one-half of children aged two to six years living in HICs are incompletely vaccinated [3, 4], and 2% to 3% of two-year-old children are completely unvaccinated [5]. Among children living in HICs, vaccination coverage for most vaccines does not meet defined targets [5–7], with harmful consequences. For instance, outbreaks of vaccine-preventable diseases such as measles and pertussis continue to occur sporadically in France [8–10]. Moreover, 25% of childhood deaths and severe sequelae due to community-onset pneumococcal and meningococcal infections were found to be vaccine-preventable [11]. Thus, a better understanding of the main drivers of incomplete vaccination in the pediatric population is pivotal to design corrective actions, notably to identify priority targets and messages for future educational campaigns.

Some determinants of incomplete vaccination in HICs have been identified and are related to (i) child characteristics [3, 5, 12–14]; (ii) household socio-demographic characteristics [3, 5, 12–19]; (iii) parental knowledge, attitude, and practices (KAPs) such as lack of information or perception that natural disease is preferable to vaccination [3, 16]; (iv) healthcare system characteristics such as no usual provider [3, 13, 15, 18]; and (v) local vaccination schedule characteristics such as non-simultaneous vaccination [12, 16, 17] (eTable 1). However, the results of these studies are of limited interest to understand the current main drivers of incomplete vaccination in children because of the timing of data collection [3, 12, 13, 15, 16], their design exposing them to strong selection bias [13, 17], and the paucity of parental KAP [5, 14, 15, 18, 19] data analyzed.

France is a country of particular interest to study the determinants of incomplete vaccination because it hosts a strong and well-established phenomenon of vaccine hesitancy [14, 20] and also hosts one of the most recent European national population-based birth cohorts [21], the ELFE study (Etude Longitudinale Française depuis l’Enfance: French Longitudinal Study from childhood). This cohort allows for studying incomplete vaccination with a very large sample offering good statistical power, a population-based design that limits selection bias, and numerous and complementary data regarding household socio-demographic characteristics and parental KAPs. In this study, we aimed to use this database to assess the determinants of incomplete vaccination in two-year-old children in France.

Materials and methods

General methodology

We used the data from the ELFE prospective nationwide birth cohort, which intends to follow children from birth to adulthood to study the relation between socio-demographic context, behaviors, and overall health [22, 23]. The ELFE cohort enrolled 18,329 children born in 2011 in metropolitan France who were recruited in the 320 participating maternity units among 349 randomly selected hospitals. A stratified sampling according to the size of each maternity unit was used and patient recruitment took place during four inclusion periods of four to eight days spread over the year. The inclusion criteria were infant born alive, singleton or twins, term ≥ 33 weeks of gestation, mother aged 18 years or older, parent(s) able to provide informed consent in one of the established languages (French, English, Arabic, Turkish), and not living temporarily in France. A prospective follow-up of the children was performed by several surveys [22, 23]. The ELFE study was approved by the Advisory Committee for the Treatment of Information on Health Research (no. 13,004), the National Agency Regulating Data Protection (no. 913,074), and the National Statistics Council (visa 2013X719AU). All participating parents provided written consent for their own and their child’s participation.

For the present study, we analyzed all children included in the ELFE cohort whose parents did not withdraw consent and had available data on the main vaccinations recommended before age two years according to the French vaccination schedule in 2011–2012. For twin pairs, only one infant was randomly selected. We used the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines to report this study (eTable 2) [24].

Vaccination status

Vaccination against 11 infectious pathogens was recommended before age two years in France at the time of the study [25]. This immunization program was normally to be achieved by four doses of diphtheria, tetanus, and inactivated poliomyelitis vaccine (DT-IPV); four doses of acellular pertussis vaccine (aP); four doses of Haemophilus influenzae type b conjugate vaccine (Hib); three doses of pneumococcal conjugate vaccine (PCV); three doses of hepatitis B vaccine (HepB); two doses of measles-mumps-rubella vaccine (MMR); and one dose of serogroup C meningococcal vaccine (MenC) (eFigure 1). In France, vaccination for children is available from any primary care practitioner of the parents’ choice. It includes general practitioners (GP) and pediatricians (whereby parents pay for the consultation and the vaccine then are reimbursed in full), as well as physicians working in maternal and child protection services (MCPSs), which are free-of-charge universal services in France offering prevention and health promotion services for pregnant women and children up to age six [26]. Only DT-IPV was mandatory at the time of the study, and MenC was introduced to the vaccination schedule in 2010 [27]. The number of doses received for each vaccine from birth until the visit and reported in the child health booklet was collected in 2013 by the child’s referring physician during the mandatory medical consultation scheduled at the age of two. In France, vaccination during infancy is systematically noted in this booklet. Vaccination was considered incomplete when at least one recommended vaccine dose was missing because the failure to receive a single dose by age two years is considered potentially dangerous for most of the vaccines studied [28].

Potential determinant measurements

Child characteristics, household socio-demographic characteristics, parental KAPs, and healthcare system characteristics were collected by using the successive parent questionnaires of the ELFE study as described elsewhere [22]. The baseline assessment was achieved during the maternity hospital stay with a face-to-face interview with the mother and data collection from the mother’s medical record. During the follow-up, computer-assisted telephone interviews were performed at two months, one and two years of age.

Statistical analyses

We first described the general characteristics of the analyzed children and compared them with those not analyzed (i.e., meeting the ELFE study inclusion criteria but without information on vaccination status at age two) and we estimated the vaccine coverage by vaccine. Then, we calculated the incomplete vaccination rate at age two years and studied crude and multi-variable associations with potential determinants by using a logistic regression model including candidate covariates previously identified as potential risk factors of incomplete vaccination in a literature review [3, 5, 12–19] (eTable 1), significantly associated with incomplete vaccination in the crude analyses or considered of interest by the co-authors.

All descriptive analyses performed on the total sample of children analyzed (rates and confidence intervals [CIs]) were weighted to take into account the inclusion procedure and the selection bias resulting from non-response and to provide representative results of births in 2011 in France. These statistical weights corrected by a calibration on margins were calculated for each child included in the analysis according to numerous covariates (eMethods 1). The parental KAP data were summarized in one latent variable, whose classes were identified by using clustering analysis including a latent class analysis and a selection of relevant variables (eMethods 2) [29, 30]. This parental KAP latent variable was integrated into the final multi-variable logistic regression model as a categorical covariate.

Several sensitivity analyses were performed to (i) repeat the descriptive analyses without weighting, (ii) repeat the multivariable analyses with weighting (eMethods 1), and (iii) modify the outcome (i.e., without MenC, the most recently recommended vaccine [27], or without HepB, the most controversial vaccine in France [31]). The rate of missing values for the potential determinants ranged from 0 to 5%. We performed multiple imputations with chained equations by applying logistic regression for binary variables, polytomous logistic regression for multinomial unordered qualitative variables, and a proportional odds model for ordinal qualitative variables (eMethods 3) [32]. The analyses involved using R v4.0.5 (R foundation for Statistical Computing, Vienna, Austria).

Results

Description of the population

Among the 18,329 newborns included in the initial ELFE study, 5,740 (31.3%) children were included in the present analyses (Fig. 1). Analyzed children differed from non-analyzed ones (n = 11,844) in several characteristics; notably, they were less often born prematurely, more often had a mother born in France, with a university degree, a higher household income, and more often resided in suburban areas (eTable 3).

Fig. 1.

Flowchart of participants in the study. aP: acellular pertussis vaccine, CI: confidence interval, DT-IPV: diphtheria, tetanus, and inactivated poliomyelitis vaccine, HepB: hepatitis B vaccine, Hib: Haemophilus influenzae type b conjugate vaccine, MenC: serogroup C meningococcal vaccine, PCV: pneumococcal conjugate vaccine, WG: weeks of gestation

Incomplete vaccination rate

Among 5,740 children analyzed, 46.5% (95% CI [44.7–48.0]) were incompletely vaccinated at age two years, and 0.4% [0.2–0.7] had never received a single vaccine dose (Table 1). The vaccination coverage was 82.6% [81.2–84.0] for the second dose of MMR, 71.1% [69.4–73.0] for MenC, and 89.6% [87.4–90.0] for HepB. Among the 2,550 incompletely vaccinated children, 46.0% (n = 1,174) were missing one dose, 43.6% (n = 1,111) were missing two to five doses, and 10.4% (n = 265) were missing six or more doses.

Table 1.

Vaccination rate at age 2 years, by vaccine for the ELFE cohort and national 2013 data

	ELFE children (n = 5,740)					National vaccination coverage data 2013 e
Vaccine	Weightedc %	[95% CI]d	Unweighted, n	%	[95% CI]d
DT-IPV
Primo-vaccination a	98.5	[98.0–99.0]	5,666	98.7	[98.4–99.0]	98.5%
Booster b	90.4	[89.3–91.0]	5,202	90.6	[89.8–91.4]	91.0%
aP
Primo-vaccination a	98.3	[97.8–99.0]	5,651	98.4	[98.1–98.7]	98.3%
Booster b	89.3	[88.1–90.0]	5,129	89.4	[88.5–90.1]	90.3%
Hib
Primo-vaccination a	97.4	[96.7–98.0]	5,612	97.8	[97.3–98.1]	97.5%
Booster b	87.6	[86.3–89.0]	5,078	88.5	[87.6–89.3]	88.4%
PCV	94.2	[93.3–95.0]	5,392	93.9	[93.3–94.5]	89.2%
MenC	71.1	[69.4–73.0]	4,214	73.4	[72.2–74.6]	56.4%
MMR
1 dose	97.5	[96.8–98.0]	5,601	97.6	[97.1–98.0]	90.3%
2 doses	82.6	[81.2–84.0]	4,782	83.3	[82.3–84.3]	74.5%
HepB	89.6	[87.4–90.0]	5,072	88.4	[87.5–89.2]	81.5%
Incompletely vaccinated	46.5	[44.7–48.0]	2,550	44.4	[43.1–45.7]	-
Completely unvaccinated	0.4	[0.2–0.7]	16	0.3	[0.2–0.5]	-

aP acellular pertussis vaccine, CI confidence interval, DT-IPV diphtheria, tetanus, and inactivated poliomyelitis vaccine, HepB hepatitis B vaccine, Hib Haemophilus influenzae type b conjugate vaccine, MenC serogroup C meningococcal vaccine, MMR measles-mumps-rubella vaccine, PCV pneumococcal conjugate vaccine

^a“Primo-vaccination” refers to the administration of the first three doses of these vaccines

^b“Booster” refers to the fourth dose

^cWeighting is detailed in eMethods 1

^dCalculated with the Wilson score method with continuity correction.

^eNational vaccination coverage at age 2 years in 2013 calculated by the French national public health agency [45]

Factors associated with incomplete vaccination

We found several significant crude associations between incomplete vaccination and potential determinants (eTable 4), and several independent ones after adjustment on all relevant covariates and on the parental KAP latent variable (eMethods 2). Regarding household socio-demographic characteristics, incomplete vaccination was independently associated with having older siblings (adjusted odds ratio [aOR] = 1.18, 95% CI [1.03–1.34] and 1.28 [1.06–1.54], for one and at least two siblings, respectively, vs. zero) and residing in an isolated area (1.92 [1.36–2.75] vs. an urban area) (Table 2). Regarding parental KAPs, incomplete vaccination was independently associated with having parents not following health recommendations or parents using alternative medicines (1.81 [1.41–2.34] and 1.23 [1.04–1.46], respectively, vs. parents confident in institutions and following heath recommendations). Regarding healthcare system characteristics, incomplete vaccination was independently associated with not receiving any visit from an MCPS nurse during the child’s first two months (1.19 [1.03–1.38] vs. at least one visit) and being medically followed by a GP (2.87 [2.52–3.26] vs. a pediatrician or an MCPS physician).

Table 2.

Multi-variable associations between main characteristics and incomplete vaccination (adjusted logistic regression, reference group: children with full vaccination)

	aOR	[95% CI]	pa
Household socio-demographic characteristics
Number of older siblings			0.02
Zero	1	-
One	1.18	[1.03–1.34]
≥ Two	1.28	[1.06–1.54]
Area of residence b			< 0.001
Urban	1	-
Suburban	1.10	[0.98–1.24]
Isolated	1.92	[1.36–2.75]
Parental KAPs
Parental KAP latent variable c			< 0.001
Compliant with health recommendations and confident in institutions	1	-
Alternative medicine user	1.23	[1.04–1.46]
Low compliant with recommendations during pregnancy	1.15	[0.95–1.39]
Low compliant for child care	1.81	[1.41–2.34]
Healthcare system characteristics
No visit of an MCPS nurse during the first two months (ref =  ≥ 1)	1.19	[1.03–1.38]	0.02
Specialty of child’s physician during the first two years			< 0.001
Pediatrician or MCPS physician alone	1	-
GP and either pediatrician or MCPS physician	1.10	[0.94–1.28]
GP alone	2.87	[2.52–3.26]

No multi-collinearity problem was detected (overall VIF < 1.21)

aOR adjusted odds ratio, CI confidence interval, GP general practitioner, KAP knowledge, attitude, and practice, MCPS maternal and child protection service, ref reference group, VIF variance inflation factor

^aCalculated with likelihood ratio test, adjusted on preterm birth, child’s health, mother’s country of birth, mother’s age, marital status of parents, mother’s level of education, parental difference of education level, mother’s employment, household income, type of mother’s health insurance, type of mother’s complementary insurance, and childcare providers

^bDefined according to the French National Institute of Statistics and Economic Studies (INSEE) 2010 classification [36]

^cClusters within the parental KAP latent variable were identified with latent class analysis (eMethods 2)

Sensitivity analyses

On sensitivity analyses, the same associations as above were found except for (i) the number of siblings when analyses were weighted, (ii) the area of residence and a visit from an MCPS nurse when MenC was removed from the outcome, and (iii) younger maternal age, which became significantly associated with incomplete vaccination (0.98 [0.97–1.00]) when MenC or HepB was removed from the outcome (eTables 5 and 6).

Discussion

Main results and interpretation

In this national-scale population-based prospective study, incomplete vaccination at age two years was frequent, with more than 45% of children incompletely vaccinated. This finding is consistent with the study of Bailly et al., which found 47% of children with at least one delayed vaccination for their age among 443 French children under age two followed by primary-care pediatricians in 2014 [14]. In other HICs, incomplete vaccination rates seemed lower, about 20% to 30% of children aged two to three years [15, 17, 33]. This observation could be explained by differences in the organization of vaccine delivery and administration in different countries, discrepancies in mandatory vaccination, and the strong phenomenon of vaccine hesitancy in France [34]. Incomplete vaccination rates were worrying in view of the resurgence of some vaccine-preventable diseases and their severity in a particularly vulnerable population whose age is close to the peak incidence of many of these diseases [28]. This high rate also emphasizes the need for quantitative evidence-based studies to identify risk factors of incomplete vaccination.

We identified several independent factors associated with incomplete vaccination at age two years, and some were previously identified in the literature. The strength of these associations, from modest (having older siblings, having parents using alternative medicine, and never having received a visit from an MCPS nurse) to more substantial (residing in an isolated area, and being followed by a GP), could be useful in prioritizing targets for corrective actions. Regarding household socio-demographic characteristics, having older siblings was significantly associated with incomplete vaccination, as was found in previous studies [3, 15, 17]. This could be related to an increase in parental self-confidence regarding the absence of visible immediate health consequences of non-vaccination, as found with other preventive attitudes [35], and a restriction of parenting availability associated with the competing needs of each child in the household. Thus, families with a high number of children should probably be targeted by vaccine programs, and vaccines being available in the physician’s office could improve vaccination uptake by allowing immediate vaccination during the consultation.

Consistently with the Cotter et al.’s study in the United States in 2002 [13], we found incomplete vaccination strongly associated with residing in an isolated versus an urban area [36]. Thus, geographic distance to healthcare facilities seems to remain a barrier for families to access vaccination in HICs in the twenty-first century [3]. Suggested solutions to simplify access to vaccination included allowing vaccination in alternative settings such as pharmacies, schools, or by private nurses [37]. Vaccination of children directly in childcare settings during French targeted vaccination campaigns following local outbreaks (e.g., MenC) has been successful [38].

Contrary to previous findings [5, 39], we did not find any association between vaccination status and the socio-economic level of the family, which could be explained by the free access to vaccination for all children under age six in the MCPSs and the French state-funded health coverage for precarious people, which aim to reduce health inequalities.

Regarding the parental socio-cognitive profiles defined according to their KAPs, parents with low compliance for child care and those who used alternative medicine were over-represented among children with incomplete vaccination. These profiles were similar to those associated with the phenomenon of vaccine hesitancy [39–41]. The individual decision-making process of vaccination, more generally health-related decisions, is a complex phenomenon that rests on social, cultural, and historical foundations [40]. In particular, the relationship between a favorable opinion of alternative medicine and skepticism about vaccination can be explained by common attitudes and beliefs, especially magical beliefs about health [42]. Moreover, it was shown in other studies that some parents belonging to an educated environment tend to control their health and have a greater awareness of the risks produced by science and industry, which can lead them to turn to alternative medicine or to refuse some preventive care procedures [39]. We identified a high education level as a risk factor for incomplete vaccination in crude analysis but not adjusted, possibly in relation with the adjustment strategy [5, 15, 17, 18]. So parental health education should probably be adapted to each target profile, in particular, alternative medicine users who accounted for almost half of the parents (eFigure 2). Information should be delivered using messengers who share their worldview to better communicate prevention messages.

Regarding healthcare system characteristics, never having received a visit from an MCPS nurse during the child’s first two months was associated with incomplete vaccination regardless of parental education level, household income, or area of residence and seems particularly relevant for recently recommended vaccines (eTable 6). This result is consistent with the known positive impact of the MCPSs on other aspects of health prevention [26]. Increasing the territorial coverage of preventive health centers such as the MCPSs could be an effective measure to reduce health inequalities in vaccination; indeed, the prevalence of families that did not receive support from an MCPS nurse during the first months was high (over 80%). Moreover, we found a substantially higher incomplete vaccination rate when the child’s medical follow-up was only assured by a GP as compared with a follow-up by a pediatrician or an MCPS physician regardless of the child’s health status at age two or area of residence. Medical follow-up by a GP alone rather than a pediatrician was previously found associated with increased drug prescriptions, lower preventive attitudes, and lower vaccination coverage [18, 43]. The phenomenon of vaccine hesitancy is also present among some GPs and may affect their attitude toward the vaccination advice given to parents [20]. Improving GPs’ awareness of preventive measures for children seems a key measure for improving vaccine coverage.

Strengths and limitations

This population-based study was the first nationwide prospective study to investigate overall vaccination status in HICs. Its unique design allowed for an estimation of the findings to the general national population of two-year-olds. The large sample size allowed for studying many covariates from a variety of research fields [23]. Moreover, the use of latent class analysis to constitute the parental KAP latent variable facilitated the interpretation of parental behaviors [30]. Finally, in this study, missing data for covariates were ≤ 5%, which probably had a limited impact on our results.

However, the main limitation was a selection bias resulting from two phenomena. First, although participation in this study was proposed to almost all eligible mothers at the maternity ward, only 51% agreed to participate [22]. Additionally, the ELFE study did not include children born extremely and very preterm who are more sensitive to infections, including vaccine-preventable ones, and usually benefit from a specific medical follow-up that improves their vaccination status [12]. However, these births represented less than 2% of French live births, and children born with moderate to late prematurity, who account for 5.5% of life births [44], displayed no significant difference in vaccination status at age two in our study (eTable4). Second, the number of patients was further limited by a significant rate of lost to follow-up at age two years and lack of complete information on vaccination status. This attrition bias led to an over-representation of social categories with high income and a high education level (eTable 3) [22]. Such selection bias was expected because the association between high socio-economic level and compliance with follow-up in cohort has been well described in this cohort and others [35]. After a careful weighting, the characteristics of our population were close to those of the reference population of women who gave birth in 2011 known by the National Perinatal Survey [22]. Also, the calculated rates and multivariable analyses were not significantly altered, so the impact of selection bias on the external validity of our study seemed limited. Moreover, vaccination coverages per vaccine were consistent with those calculated by the French national public health agency in 2013 for DT-IPV, aP, and Hib based on the 24-month medical certificates (Table 1) [45]. In practice, these three vaccines are almost always combined in a single injection, so a similar vaccine coverage is expected even though some vaccines are not mandatory. Vaccination coverage for other vaccines was slightly higher overall than the French national public health agency estimates, so the rate of incomplete vaccination may be slightly underestimated in our study.

Moreover, although we investigated numerous potential determinants previously highlighted [3, 5, 12–19], some were not collected, and these included move before or shortly after birth [46] or delay in the first vaccinations [12, 16]. Because age at vaccination was not recorded, we could not assess whether completely vaccinated children experienced potentially dangerous vaccination delays [28]. Moreover, the exact age of data collection was not precisely known, so delayed vaccination of some children may have been reported when they may have been vaccinated shortly after data collection [28], notably for the second dose of MMR scheduled at 24 months of age and the first dose of MenC which could be given up to 24 months of age.

In addition, totally unvaccinated children were included in the incompletely vaccinated group, but these two situations could share different risk factors [5]. A specific analysis of the determinants associated with complete lack of vaccination is planned.

Finally, the findings of this study were limited by the timing of data collection. France has introduced mandatory vaccination for all newborns since 2018, which has improved vaccine coverage per vaccine for children at age two years [7, 8]. This move has probably resulted in a decrease in the current rate of incompletely vaccinated children and potentially a change in the determinants associated with incomplete vaccination in children under two years of age. However, this obligation is not intended to be permanent, and some vaccinations recently recommended by the French National Authority for Health are not mandatory for children or adolescents, such as vaccination against meningococcal B [47], human papillomavirus [48], and SARS-CoV-2 [1]. Thus, our results may guide the implementation strategy for these non-mandatory vaccines and to reach children that are not vaccinated despite the 2018 mandatory vaccination law.

Conclusions

The rate of incompletely vaccinated children at age two was very high in France in the studied population. The risk factors we identified could guide corrective actions such as campaigns promoting vaccination among parents with large families and those with specific socio-cognitive profiles; the strengthening of the vaccination forces by a better territorial coverage of the MCPSs, the removal of logistic and territorial barriers by the prescriber delivering the vaccines and the implementation of alternative vaccination settings; and an active medical education of GPs regarding vaccines and vaccine hesitancy.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

aOR

Adjusted odds ratio

aP

Acellular pertussis vaccine

CI

Confidence interval

DT-IPV

Diphtheria, tetanus, and inactivated poliomyelitis vaccine

ELFE

Etude Longitudinale Française depuis l’Enfance

GP

General practitioner

HepB

Hepatitis B vaccine

Hib

Haemophilus influenzae type b conjugate vaccine

HIC

High-income country

KAP

Knowledge, attitude, and practice

MCPS

Maternal and child protection service

MenC

Serogroup C meningococcal vaccine

MMR

Measles-mumps-rubella vaccine

PCV

Pneumococcal conjugate vaccine

Author contribution

Conception of study: Corinne Bois, Marie-Aline Charles, and Marie-Noëlle Dufourg. Collection of the data: Corinne Bois, Marie-Aline Charles, and Marie-Noëlle Dufourg. Design of study: Corinne Bois, Martin Chalumeau, Marie-Aline Charles, Marie-Noëlle Dufourg, Christèle Gras-Le Guen, Elise Launay, Daniel Lévy-Brühl, Fleur Lorton, and Jocelyn Raude. Data management: Marianne Jacques, Fleur Lorton, Pauline Scherdel, and Thierry Siméon. Data analysis: Marianne Jacques, Fleur Lorton, Pauline Scherdel, and Thierry Siméon. Drafting of the manuscript: Martin Chalumeau, Marianne Jacques, Fleur Lorton, and Pauline Scherdel. Revising of the manuscript: Corinne Bois, Marie-Aline Charles, Marie-Noëlle Dufourg, Christèle Gras-Le Guen, Elise Launay, Daniel Lévy-Brühl, Jocelyn Raude, and Thierry Siméon. Final approval of the version to be published: All.

Funding

The ELFE birth cohort is a joint project between the French Institute for Demographic Studies (INED, Institut national d’études démographiques) and the National Institute of Health and Medical Research (INSERM, Institut national de la santé et de la recherche médicale), in partnership with the French blood transfusion service (EFS, Etablissement français du sang), the French National Public Health Agency (Santé publique France), the National Institute for Statistics and Economic Studies (INSEE, Institut national de la statistique et des études économiques), the French Ministry of Health (DGS, Direction générale de la santé), the Ministry for the Environment (DGPR, Direction générale de la prévention des risques), the Ministry of Health and Social Affairs (DREES, Direction de la recherche, des études, de l’évaluation et des statistiques), the Ministry of Culture (DEPS, Département des études, de la prospective et des statistiques), and the National Family Allowance Fund (CNAF, Caisse nationale des allocations familiales), with the support of the Ministry of Higher Education and Research (INJEP, Institut national de la jeunesse et de l’éducation populaire). Via the RECONAI platform, it receives a government grant managed by the National Research Agency under the investments for the future programs (ANR-11-EQPX-0038).

Availability of data and materials

Access to participant data with identifiers, data dictionary, and code is available only once approval has been obtained through the constituent entities controlling access to the data. The crude data can be requested to the ELFE Data Access Committee (https://pandora-elfe.inserm.fr/public/).

Declarations

Ethics approval and consent to participate

The ELFE study was approved by the Advisory Committee for the Treatment of Information on Health Research (no. 13,004), the National Agency Regulating Data Protection (no. 913,074), and the National Statistics Council (visa 2013X719AU). All participating parents provided written consent for their own and their child’s participation.

Competing interests

The authors declare no competing interests.