Sc

Ka

b

c

a

ARR2AA

KVASG

1

1

waiit

lm[rlacp

if

0d

Vaccine 28 (2010) 1861–1869

Contents lists available at ScienceDirect

Vaccine

journa l homepage: www.e lsev ier .com/ locate /vacc ine

ocioeconomic factors play a more important role in childhood vaccinationoverage than parental perceptions: a cross-sectional study in Greece

. Danisa,∗, T. Georgakopouloub, T. Stavrouc, D. Laggasa, T. Panagiotopoulosa

Department of Child Health, National School of Public Health, 196 Alexandras Av., 11521 Athens, GreeceDepartment of Surveillance and Intervention, Hellenic Centre for Disease Control and Prevention, GreeceDepartment of Public Health, Ministry of Health, Greece

r t i c l e i n f o

rticle history:eceived 28 February 2009eceived in revised form9 November 2009ccepted 30 November 2009vailable online 19 January 2010

a b s t r a c t

To identify predictive factors of complete and age-appropriate vaccination status in Greece, we con-ducted a cross-sectional study, using stratified cluster sampling, among children attending the firstyear of the Greek Grammar school (about 6 years of age) and their parents/guardians. Almost 88%(N = 3878) of pupils in the selected clusters (school classrooms) provided their vaccination booklet andtheir parents/guardians completed a questionnaire regarding beliefs and attitudes towards immuniza-tion. Belonging to a minority group, having other siblings and perceiving long distance to immunizationsite as a barrier were independent predictors of both incomplete and delayed vaccination status in the

eywords:accination coveragettitudesocioeconomicreece

final logistic regression model. Maternal age ≥ 30 years and the perception that natural disease is prefer-able to vaccination were associated with complete vaccination, whereas paternal education of high schoolor higher was the other independent determinant of age-appropriate immunization. Socioeconomic fac-tors rather than parental beliefs and attitudes towards immunization explained underimmunization.Further interventions are warranted to enhance vaccine coverage in high-risk groups identified in this

study.

. Introduction

.1. Current knowledge on factors for underimmunization

Vaccine-preventable diseases are still a public health burdenorldwide [1–3], largely due to suboptimal vaccination rates [4,5],

nd constitute an important component of the health care debaten many countries. To improve vaccination coverage and reach highmmunization levels, the reasons for inadequate vaccinations needo be identified and addressed.

A group of possible factors have been found to be associated withow immunization rates, including different socioeconomic deter-

inants such as low family income [6–10], lack of health insurance10–12], low levels of parental education [6,7,13,14], non-whiteace [6,8,15], young age of parents [14], the presence of other sib-

ings in the household [8,11]; and structural barriers such as lack ofroutine source of health care [9] or decreased availability of physi-ian services [10–12,14–21]. Several studies suggest that vaccinerovider-associated factors including physicians’ behavior have an

∗ Corresponding author at: National School of Public Health, Tomeas Ygeias Paid-ou, 196 Alexandras Av., 11521 Athens, Greece. Tel.: +30 6976133793;ax: +30 2132010155.

E-mail address: daniscostas@yahoo.com (K. Danis).

264-410X/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2009.11.078

© 2009 Elsevier Ltd. All rights reserved.

important impact on vaccination coverage [16,22–26]. However,most of these studies were conducted in geographically confinedareas with distinct cultural factors and differences in health caresystems, limiting the generalizability of findings.

Other studies have identified perceived barriers, parental healthbeliefs and attitudes towards childhood immunization as riskfactors for undervaccination [23,27–33]. However, some studiesconcluded that parental attitudes and beliefs could not explainunderimmunization, with a few suggesting that socioeconomicfactors play a more important role [34–38]. A recent studyhas indicated that socio-demographic characteristics might affectbeliefs, attitudes and perceived control, which eventually deter-mine immunization [39].

1.2. Vaccination in Greece

In Greece, vaccines included in the National Vaccination Pro-gramme (NVP) (Box 1) [40] are provided free of charge to allthe residents (including immigrants) in primary health centresor health insurance clinics. Childhood vaccination coverage is not

monitored routinely. Several ad hoc studies were conducted at locallevels occasionally using non-representative samples [41]. How-ever, most of the current knowledge on immunization uptake isbased on two national population-based surveys conducted in 1997and in 2001 using representative samples of children attending the

1862 K. Danis et al. / Vaccine 28 (2010) 1861–1869

Box 1: Recommended childhood vaccination schedule according to the National Vaccination Programme (NVP) at thetime of the study in Greece [40].

Recommended age DTaP* Hib* IPV* MMR* HepB* MCV*† PCV7*† Var*† BCG* dT*

2 months Yes Yes Yes Yes Yes Yes4 months Yes Yes Yes Yes Yes Yes6 months Yes Yes Yes6–18 months Yes Yes12–15 months Yes Yes Yes12–18 months Yes15–18 months Yes Yes4–6 years Yes Yes Yes6 years Yes11–18 years Yes13–18 years Yes>18 years Yes

*DTaP = diphtheria-acellular tetanus-pertusis (DTP); Hib = Haemophilus influenzae type b (HiB); IPV = inactivated polio vaccine;

MMR = measles–mumps–rubella; HepB = hepatitis B; MCV = meningococcal conjugate vaccine; PCV7 = Pneumococcal heptavalent;etanu

fi2s3a(et5r

1

dbtacatc2icbi

2

pi

•

•

•

BCG = Bacillus Calmette-Guérin; dT = reduced diphtheria toxoid-t†Introduced in the NVP in 2006.

rst year of Greek Grammar school (children born in 1991) and of-year olds (children born in 1997), respectively. Results of thoseurveys showed high vaccination coverage [93.5% and 98.9% for therd dose of diphtheria-tetanus toxoid and pertusis (DTP) vaccine;nd 89.1% and 93.7% for the first dose of measles–mumps–rubellaMMR) vaccine in 1997 and 2001, respectively] [42,43,47]. How-ver, an outbreak of measles in 2006 revealed low coverage amonghe Roma minority and immigrants [44–46]; groups accounting for5% and 15% of cases, but only 1% and 8% of the general population,espectively.

.3. Rationale for the study

Most large studies have only assessed the effect of socio-emographic factors on vaccination, whereas the role of parentaleliefs and attitudes or the combination of a wide range of fac-ors were less commonly examined. In Greece, there have not beenny published studies conducted to identify determinants of vac-ination uptake. By identifying potential parental beliefs, barriersnd social factors associated with underimmunization, modifica-ions may be made in strategies designed to ensure that morehildren receive all recommended vaccines. Therefore, as part of the006 national vaccination coverage survey among children attend-

ng the first year of the Greek Grammar school, we conducted aross-sectional study to identify the potential effects of parentaleliefs, perceived barriers and socioeconomic characteristics on

mmunization status of children.

. Aim and objectives

The aim of the study was to identify predictive factors of com-lete and age-appropriate vaccination status of 6-year-old children

n Greece. The specific objectives were to:

To assess the potential effect of parental beliefs and atti-tudes towards immunization on compliance with recommended

immunization.To evaluate the relationship between socioeconomic character-istics and underimmunization.To determine if perceived barriers to vaccination affect the vac-cination status of children.

s toxoid.

3. Methods

3.1. Sampling strategy

The country was stratified into 6 regions, which, apart from thecapital region (Attica), were further subdivided into urban and ruralareas, to create 11 strata. School classrooms (clusters) were ran-domly selected with probability proportional to the total numberof first year Grammar school pupils in each region. The samplingframe of all school classrooms in the country for the academic year2004–2005, was obtained from the Greek Ministry of Education.The list contained 108,538 pupils from both public and privateschools. All pupils of the selected school classrooms and their par-ents/guardians were asked to participate in the study. Residentsin the rural areas of the North-Western region of Thrace, a largeproportion of whom belong to the Greek Muslim minority, wereoversampled, to achieve more accurate estimates for this group.

3.2. Sample size calculation

Assuming a vaccination coverage of 90% and a design effect of 2,a sample size of approximately 4000 children was required to esti-mate this with a precision of ±1.2% points. These sample size targetswere expected to allow the estimation of region-specific estimatesof vaccine coverage with sufficient precision (±3% points). Giventhe above figures, this study had a power of 85% to detect a sig-nificant effect of factors predicting vaccination, at the 5% level ofsignificance, assuming a 55% vaccination coverage in the referencegroup, a Relative Risk (RR) of 1.10 and a ratio of subjects in thedifferent groups of 3.

3.3. Data collection instruments

The following data collection instruments were used:

• School registry to obtain basic information about the school,including a list of all selected classrooms and pupils and basicdemographic characteristics of the children.

• Child vaccination booklet to gather information regarding vacci-nation status.

• Self-administered questionnaire completed by parents/guardiansregarding beliefs and attitudes towards immunization, perceivedbarriers to vaccination and parental socioeconomic character-istics. The questionnaire was validated during a pilot studythat involved 20 parents/guardians belonging to both minor-

K. Danis et al. / Vaccine 28 (2010) 1861–1869 1863

Box 2: Definitions of complete and age-appropriate vac-cination statusComplete vaccination statusChildren were considered fully vaccinated if they had receivedall of the following vaccinations, according to the NVP [40]: (i) 5doses of DTP vaccine, (ii) 5 doses of poliomyelitis vaccine, (iii) 2doses of MMR vaccine, (iv) 3 doses of hepatitis B (HBV) vaccineand (v) full vaccination for Haemophilus Influenza type b (Hib).Hib full immunization was considered as having received oneof the following: (i) 4 doses of Hib, if the first dose was admin-istered at the age of 0–6 months and the 4th at the age of >12months, or (ii) 3 doses, if Hib vaccination began at the age of7–11 months and the 3rd dose was administered at the age of>12 months, or (iii) 2 doses, if the first dose was administeredat the age of 12–14 months, or (iv) 1 dose, if Hib vaccinationstarted at the age of ≥15 months.Age-appropriate vaccinationChildren were considered age-appropriately vaccinated if theywere both completely and timely immunized, i.e. if they hadreceived all of the following, according to the NVP [40]: (i)the 3rd, 4th and 5th doses of DTP vaccine by the age of 12,24 and 72 months, respectively, (ii) the 3rd, 4th and 5th dosesof poliomyelitis vaccine by the age of 12, 24 and 72 months,respectively, (iii) the first and second doses of MMR vaccineby the age of 2 and 6 years, respectively, (iv) 3 doses ofHBV vaccine and (v) primary vaccination for Hib. Hib primaryvaccination was considered as having received one of thefollowing: (i) 3 doses of Hib, if the first dose was administeredat the age of 0–6 months, or (ii) 2 doses if Hib vaccinationbegan at the age of 7–14 months or (iii) 1 dose if the first dosewas administered at the age of ≥15 months. Time componentwas not included for Hib and HepB vaccines, because (i) Hibwas only introduced in the NVP after the cohort of the studychildren had been born and (ii) the large majority of children(>80%) received the first dose of HepB vaccine later than

3

3

noAtiw“

pvmh

3

oPa

D

Box 3: Definition of positive and uncritical positiveopinion towards immunization

• Positive opinionA positive opinion towards immunization was defined as

agreement with the following two positively worded state-ments: (i) “Vaccinations are necessary for my child” and(ii) “I keep my child’s immunization up-to-date, accordingto scientific advice”, and disagreement with the followingtwo negatively worded statements: (i) “I fear that vaccinesmay harm my child” and (ii) “Natural childhood disease ispreferable to vaccination”. The scores from each of thesestatements were summed up, with higher scores denot-ing more positive beliefs regarding immunization. Based onthese scores, positive opinion was classified (on a 10-pointscale) as strong (>7.5), moderate (6.0–7.5) and not positive(<6.0).

• Uncritical positive opinionUncritical positive opinion towards immunization was

defined as agreement with the following two statements:(i) “Children should be immunized immediately with everynewly licensed vaccine” and (ii) “Vaccines are completely

recommended (in the second semester of their life) possiblydue to vaccine providers’ beliefs.

ity groups and the non-minority population from non-selectedclusters.

.3.1. Main outcomesThe definitions of the main outcomes are presented in Box 2.

.3.2. Parental beliefs and perceived barriersParents/guardians completing the self-administered question-

aire were asked to indicate their agreement with statementsn perceived benefits or harms of and barriers to immunization.nswers were transferred to numeric scores between 0 and 4, with

he maximum score corresponding to “completely agree” for pos-tively worded statements or “completely disagree” for negatively

orded statements and the middle score (2) corresponding to theI don’t know” category.

Statements addressed issues regarding positive and uncriticalositive opinions towards immunization (Box 3), new licensedaccines, perceived safety of immunizations, accuracy of recom-endations, financial profit and perceived financial, structural or

ealth care service barriers to vaccination.

.4. Study procedures

The survey team consisted of 234 health professionals from all

ver the country who received special training for the fieldwork.articipation of both children and parents/guardians was voluntarynd anonymous.

Investigators visited the selected schools on two occasions.uring the first visit, members of the survey team handed an

safe and never harm a child’s health”.

explanatory letter to the school headmaster providing detailedinformation regarding the study and asking for his/her cooper-ation. Subsequently, they handed an envelope-package for theparents/guardians containing: (i) a letter explaining the studyand asking them to provide their child’s vaccination booklet and(ii) an anonymous self-administered questionnaire asking theirattitudes towards immunization. In case of twin siblings, the par-ents/guardians were asked to complete only one questionnaire.Parents/guardians with literacy problems or poor knowledge ofGreek, completed the questionnaire with the help of the teachers.During the second visit, the investigators received the completedquestionnaires and photocopied the child’s vaccination booklets.Subsequently, they provided parents/guardians with an informa-tion note indicating the child’s vaccination status and providingrelevant recommendations.

The study received ethical approval from the Ethics Committeeof the Institute of Child Health, Athens, Greece.

3.5. Statistical analysis

Data were entered in a database designed using Epidata soft-ware (Epidata association, Denmark, version 3.1). Consistency andrange checks were performed for data validation. The analysis wascarried out using STATA software (Stata Corporation, TX, USA, ver-sion 10).

The estimated proportions of school children vaccinated by theage of 6 were weighted to adjust for the effect of the samplingdesign. Sample weights were calculated for each stratum, accordingto the stratum size of 6-year-old pupils, which was derived from thesampling frame. RR was calculated from the weighted proportionsin the different groups to identify factors associated with completeand age-appropriate vaccination status.

Multiple logistic regression models were constructed to iden-tify independent factors associated with vaccination status. Initialregression models were first constructed including all variables forwhich (i) the p-value (for the RR) was less than 0.05 or (ii) the RR

was more than 1.1 or less than 0.90 in the univariate analysis (cor-responding to a 10% change in risk). If two or more factors werehighly correlated, only the most plausible one or the one with theleast missing values was included in the model. To simplify the

1 ine 28

msaabe

4

4

Ope31oiptctt

ypnip

pOr

4

(9rbh(

4

vGspn3o

sv7p4wc1ar

864 K. Danis et al. / Vacc

odel, variables were removed one at a time depending on theignificance testing (p < 0.05) by the likelihood ratio (LR) test. Thedjusted RR derived from binomial regression using all the vari-bles of the final logistic regression model. Potential interactionsetween parental attitudes and socioeconomic factors were alsoxamined.

. Results

.1. Response rate and demographic characteristics

In total, 342 (99.2%) school classes participated in the study.f the 4390 pupils attending the selected classes, 3878 (88.3%)rovided their vaccination booklet and 3880 (88.4%) eligible par-nts/guardians returned a completed questionnaire, resulting in609 successfully linked parent/guardian-child pairs. Among those,75 were excluded from further analysis, as some of their answersn beliefs and attitudes were found inconsistent, leaving 3434 pairsn the final analysis. Answers were considered inconsistent if (i)arents feared that “vaccines may harm their children”, althoughhey also agreed that “vaccines are completely safe and never harmhild’s health” or (ii) parents felt that “vaccination is necessary forheir children”, but they reported that they would not “try to keepheir child’s immunization up-to-date”.

The age distribution of participant children (mean age 6.76ears) was similar to that of the total first year Grammar schoolupils in Greece (mean age 6.68; p = 0.16). Participant children didot differ significantly from all first year Grammar school children

n terms of gender distribution (p = 0.39) and urban-to-rural pro-ortion (p = 0.86).

The demographic characteristics of sampled children and theirarents/guardians included in the analysis are presented in Table 1.f the respondents, 80% were mothers, 19% fathers, with the

emainder including grandparents, aunts, uncles and older sisters.

.2. Vaccination coverage of recommended vaccines

Of 3434 children that were included in the final analysis, 91.2%95%CI 90.0–92.5) had received 5 doses of DTP, 98.3% (95%CI8.2–99.1) and 77.1% (95%CI 74.6–79.5) 1 and 2 doses of MMR,espectively, and 84.6% (95%CI 82.9–86.3) 3 doses of Hib vacciney the time of school entry. Furthermore, 63.9% (95%CI 61.4–66.4)ad received all the recommended doses and vaccines and 52.1%95%CI 49.5–54.8) had an age-appropriate vaccination status.

.3. Factors affecting vaccination status

Children were less likely to be completely and age-appropriatelyaccinated if they belonged to a minority group, were born outsidereece, were uninsured, were born to a younger mother, had manyiblings or their parents were less educated (Table 1). The weightedroportions of children with complete and age-appropriate vacci-ation among households with 3 or more children were 26% and8% lower, respectively, compared with those of households withnly one child.

Parents/guardians indicated a high degree of perceived neces-ity of immunization with the vast majority agreeing thataccination is beneficial for their children (Table 2). In addition,1.7% (95%CI 69.9–73.9) considered that vaccines are com-letely safe and never harm child health and 41.4% (95%CI1.3–45.6) felt that children should be immediately immunized

ith every newly introduced vaccine, reflecting an uncriti-

al positive opinion on vaccination. Nevertheless, 19.4% (95%CI6.2–22.4) reported that natural childhood disease may be prefer-ble to vaccination (Table 2). More skepticism was observedegarding new vaccines and accuracy of recommendations with

(2010) 1861–1869

approximately 30.4% (95%CI 28.2–32.5) feeling confused due toconflicting physicians’ opinions, 43.8% (95%CI 41.9–45.7) statingthe need to cross-check doctors’ recommendations and 60.3% (95%59.3–63.3) believing that new vaccines are developed for financialprofit.

Children whose parents had a positive opinion regarding immu-nization had 23% and 16% higher complete and age-appropriatevaccination coverage, respectively (Table 2). However, theseassociations were not statistically significant. Complete and age-appropriate vaccination coverage was 11% lower in children ofparents/guardians reporting that natural childhood disease may bepreferable to vaccination. After adjusting for age, gender and type(mother/father/other) of respondents, the effects of attitudes andbeliefs on both complete and age-appropriate vaccination statuswere not altered (data not shown).

Among all respondents, cost (including the cost of the visit toan immunization provider and the cost of the vaccine itself) wasthe barrier most commonly identified (Table 3). However, thisfactor was not associated with either being completely or age-appropriately vaccinated.

Both complete and age-appropriate vaccination coverage waslower in children of parents/guardians who perceived long distanceto immunization site, as a major barrier to immunization. Childrenwhose parents/guardians reported inconvenient opening hours orunfriendly behavior of immunization staff as a barrier to vaccina-tion were less likely to be completely vaccinated. The effects didnot change significantly, when adjusted for age, gender and type ofrespondents (data not shown).

4.4. Multiple regression analysis

Of the 16 variables that met the criteria for entry in the initiallogistic regression models (Tables 1, 2), country of birth, father’s ageand maternal educational level were excluded from further analy-sis, as these variables were highly correlated with minority groupstatus (�2 = 1127; p < 0.0001), mother’s age (Spearman’s correla-tion coefficient, r = 0.70; p < 0.0001) and father’s educational level(r = 0.63; p < 0.0001), respectively. In the final logistic regressionmodel, factors remaining as independent predictors of completevaccination were maternal age, belief that natural childhood dis-ease is preferable to vaccination, perception of long distance toimmunization site as an important barrier, presence of other sib-lings in the family and belonging to a minority group (Table 4). Thelatter three variables were also independently associated with age-appropriate vaccination status, along with the father’s educationallevel. Similar results were found when the analysis was restrictedto the non-minority population (data not shown).

5. Discussion

5.1. Main determinants of vaccination status

This large study identified several independent determinants ofcomplete and age-appropriate immunization status among 6-year-old school children in Greece. Both complete and age-appropriatevaccination coverage was significantly lower among children whobelonged to a minority group, who had siblings in the householdand whose parents/guardians perceived long distance to the placeof immunization as a barrier to vaccination. In addition, maternalage <25 years and the perception that natural disease is prefer-able to vaccination were significantly associated with lower rates

of complete vaccination. Furthermore, low paternal education wasan additional predictor of delayed vaccination.

Consistent with other studies, belonging to a minority groupwas the strongest predictor of incomplete and delayed immuniza-tion [5–7,13,17,18]. In addition, paternal education, which has been

K. Danis et al. / Vaccine 28 (2010) 1861–1869 1865

Table 1Demographic characteristics of children and parents/guardians and their associations with complete and age-appropriate vaccination status.

Characteristic Total N (%a) (N = 3434) Completely vaccinatedchildren N (%a)

RRb (95%CIc) Age-appropriately vaccinatedchildren N (%a)

RRb (95%CIc)

Child’s genderFemale 1919 (49.6) 1036 (63.8) Ref. 827 (50.5) Ref.Male 1949 (50.4) 1083 (63.9) 1.00 (0.94–1.06) 887 (53.8) 1.06 (0.98–1.15)

Urban/rural place of residenceRural 898 (23.2) 448 (58.7) Ref. 360 (49.3) Ref.Urban 2980 (76.8) 1676 (65.1) 1.11 (0.99–1.24) 1357 (52.8) 1.07 (0.91–1.26)

Minority groupNon-minority population 3221 (83.3) 1904 (67.1) Ref. 1593 (55.7) Ref.Immigrants 303 (7.8) 77 (29.9) 0.45 (0.37–0.53) 50 (19.2) 0.35 (0.26–0.46)Greek muslimsd 290 (7.5) 126 (52.8) 0.79 (0.63–0.99) 67 (31.6) 0.57 (0.42–0.76)Roma 52 (1.3) 9 (40.1) 0.59 (0.36–0.98) 1 (3.8) 0.07 (0.01–0.51)

Country of birthGreece 3644 (94.4) 2074 (65.7) Ref. 1686 (54.2) Ref.Other than Greece 216 (5.6) 50 (31.5) 0.48 (0.37–0.62) 31 (15.6) 0.29 (0.19–0.45)

Child’s health insurancePublic (exclusively) 2882 (84.0) 1786 (63.8) Ref. 1448 (52.7) Ref.Private or mixede 405 (12.9) 271 (69.4) 1.09 (1.01–1.17) 217 (54.6) 1.04 (0.92–1.16)Uninsured 95 (3.1) 38 (41.9) 0.66 (0.50–0.85) 29 (31.4) 0.60 (0.42–0.85)

Mother’s age at child’s birth (years)Mean age [SDf] 28.1 [5.1]<25 967 (26.5) 529 (56.6) Ref. 409 (44.9) Ref.25–29 1203 (36.3) 766 (65.1) 1.15 (1.06–1.25) 645 (54.8) 1.22 (1.08–1.37)≥30 1181 (37.2) 784 (68.1) 1.20 (1.11–1.31) 635 (55.4) 1.23 (1.10–1.37)

Father’s age at child’s birth (years)Mean age [SDf] 32.4 [5.8]<30 1071 (30.0) 613 (60.6) Ref. 487 (49.9) Ref.30–34 1166 (36.1) 760 (66.8) 1.14 (1.07–1.22) 617 (53.7) 1.08 (0.98–1.18)≥35 1085 (33.9) 689 (64.4) 1.12 (1.05–1.19) 570 (53.4) 1.07 (0.97–1.17)

Mother’s education<9 years 536 (11.8) 278 (53.5) Ref. 185 (41.2) Ref.9–11 years 429 (14.1) 240 (57.1) 1.07 (0.92–1.25) 197 (45.4) 1.10 (0.91–1.33)12 years (high school) 1336 (42.7) 859 (65.4) 1.22 (1.07–1.39) 708 (52.3) 1.27 (1.08–1.49)College/university graduate 985 (31.4) 670 (69.2) 1.29 (1.14–1.46) 576 (60.2) 1.46 (1.23–1.73)

Father’s education<9 years (Grammar school) 555 (14.2) 300 (55.9) Ref. 212 (41.1) Ref.9–11 years 534 (17.3) 302 (58.5) 1.05 (0.92–1.19) 241 (47.7) 1.16 (0.98–1.37)12 years (high school) 1132 (37.7) 706 (64.6) 1.16 (1.04–1.28) 609 (54.9) 1.34 (1.16–1.54)College/university graduate 949 (30.8) 666 (70.5) 1.26 (1.13–1.40) 553 (59.2) 1.44 (1.24–1.66)

Highest level of education attained by anyone in householdLess than high school 640 (17.1) 334 (53.9) Ref. 234 (40.4) Ref.High school graduate 1234 (41.3) 759 (63.3) 1.17 (1.07–1.29) 634 (52.0) 1.29 (1.13–1.47)More than high school 1249 (41.6) 853 (69.0) 1.28 (1.16–1.41) 725 (59.1) 1.47 (1.28–1.68)

Number of other siblingsMean number [SDd] 1.21 [0.9]Only child 254 (7.2) 336 (65.4) Ref. 288 (57.4) Ref.1–2 other children 2570 (77.1) 1630 (65.3) 0.99 (0.91–1.08) 1313 (52.8) 0.92 (0.83–1.02)≥3 other children 535 (15.7) 124 (48.6) 0.74 (0.63–0.88) 86 (35.5) 0.62 (0.48–0.79)

a Weighted proportion that allows for sample weights and clustering within school classes.b Relative Risk.c 95% Confidence interval.

e.

uaTltan

hir

d Greek Muslims were oversampled.e Private insurance also includes those reporting both private and social insurancf Standard deviation.

sed a proxy measure for socioeconomic status in Greece [48]ppeared to be inversely associated with delayed immunization.he use of paediatric preventive health services have been reportedower among socioeconomically disadvantaged children, despiteheir poorer health status [2]. It is possible that for immigrantsnd other socially disadvantaged population groups, meeting daily

eeds may be more important than obtaining preventive services.

The presence of three or more other siblings in the householdad a strong negative impact on both complete and age-appropriate

mmunization regardless of other factors. Other studies haveeported this association [6,11,23,25,51,52]. Larger numbers of chil-

dren seem to place competing demands on mothers, while timeand resources available to provide for each child become morelimited. Furthermore, this study has indicated that the risk ofundervaccination was associated with lower maternal age. It hasbeen previously shown that children born to younger mothers,those of higher birth order and those from larger families receive

fewer health care services in general and preventive services inparticular [13,16,23,51,52].

Several perceived barriers to vaccination have been identifiedin this study, including inconvenient opening hours, unfriendlyprovider behavior and bad organization of preventive services.

1866K

.Danis

etal./V

accine28 (2010) 1861–1869

Table 2Parental beliefs/attitudes towards immunization and their associations with complete and age-appropriate vaccination status.

Parents’ beliefs/attitudes Total parents N (%a) Completely vaccinatedchildren N (%a)

RRb (95%CIc) Age-appropriately vaccinatedchildren N (%a)

RRb (95%CIc)

Positive opinion on vaccinationd Yes 3184 (98.5) 1993 (64.5) 1.23 (0.91–1.66) 1618 (52.9) 1.16 (0.83–1.63)No 58 (1.5) 24 (52.3) Ref. 24 (45.5) Ref.

Strongd 2610 (81.2) 1663 (65.5) 1.25 (0.93–1.69) 1352 (53.9) 1.19 (0.85–1.66)Moderate 574 (17.3) 330 (59.6) 1.14 (0.84–1.54) 266 (48.3) 1.06 (0.74–1.52)Not positive 58 (1.5) 24 (52.3) Ref. 24 (45.5) Ref.

Uncritical positive opinion on vaccinationd Yes 1620 (48.2) 988 (63.1) 0.97 (0.91–1.02) 800 (51.4) 0.95 (0.88–1.02)No 1691 (51.8) 1068 (65.2) Ref. 877 (54.0) Ref.

Belief that vaccination is necessary for child Yes 3376 (99.8) 2093 (64.0) 1.25 (0.41–3.78) 1699 (52.4) NAf

No 5 (0.2) 2 (51.1) Ref. 0 (0) Ref.

Belief that vaccines are completely safe and never harm child health Yes 2307 (71.7) 1640 (76.2) 0.98 (0.94–1.03) 1114 (52.7) 1.07 (0.99–1.16)No 484 (28.9) 496 (77.5) Ref. 378 (49.2) Ref

Belief that children should be immunized immediately with every newly licensed vaccine Yes 1420 (41.4) 779 (60.8) 0.96 (0.91–1.02) 642 (49.9) 841 0.95 (0.89–1.03)No 1386 (58.6) 954 (63.2) Ref. (52.3) Ref.

Fear that vaccines may be harmful Yes 133 (3.3) 64 (54.8) 0.85 (0.70–1.03) 59 (48.2) 0.92 (0.74–1.15)No 2991 (96.7) 1882 (64.4) Ref. 1515 (52.6) Ref.

Belief that natural childhood disease preferable to vaccination Yes 631 (19.4) 359 (58.5) 0.89 (0.82–0.96) 298 (48.1) 0.89 (0.80–0.99)No 2580 (80.6) 1637 (65.6) Ref. 1326 (53.9) Ref.

Feeling confused regarding vaccination due to conflicting physicians’ opinions Yes 1001 (30.4) 490 (60.2) 0.93 (0.87–0.99) 443 (49.7) 0.92 (0.85–0.99)No 2103 (69.6) 1357 (64.9) Ref. 1139 (54.2) Ref.

Need to cross-check physician’s recommendations Yes 1644 (43.8) 874 (60.9) 0.95 (0.90–1.00) 741 (49.3) 982 0.93 (0.87–0.99)No 1624 (56.2) 1162 (64.0) Ref. (53.2) Ref.

Belief that new vaccines are developed for financial profit Yes 1587 (60.3) 882 (65.4) 1.07 (1.01–1.14) 758 (52.3) 521 1.01 (0.94–1.09)No 791 (39.7) 615 (60.7) Ref. (51.7) Ref.

Parental fear for vaccines and injections Yes 268 (7.5) 137 (54.2) 0.83 (0.72–0.97) 105 (43.6) 0.82 (0.68–0.98)No 3026 (92.5) 1913 (65.0) Ref. 1567 (53.4) Ref.

Trying to keep child’s immunization up-to-date, according to scientific advice Yes 3309 (99.8) 2057 (64.1) 1.16 (0.46–2.92) 1675 (52.6) NAf

No 7 (0.2) 4 (55.2) Ref. 0 (0) Ref.

a Weighted proportion that allows for sample weights and clustering within school classes.b Relative Risk.c 95% Confidence interval.d For definitions, please refer to Section 3.3.2 of this report.f Non-applicable.

K. Danis et al. / Vaccine 28

Tab

le3

Perc

eive

dba

rrie

rsto

vacc

inat

ion

and

thei

ras

soci

atio

ns

wit

hco

mp

lete

and

age-

app

rop

riat

eva

ccin

atio

nst

atu

s.

Perc

eive

dba

rrie

rsto

vacc

inat

ion

Tota

lpar

ents

N(%

a)

Com

ple

tely

vacc

inat

edch

ild

ren

N(%

a)

RR

b(9

5%C

Ic )A

ge-a

pp

rop

riat

ely

vacc

inat

edch

ild

ren

N(%

a)

RR

b(9

5%C

Ic )

Lon

gd

ista

nce

toim

mu

niz

atio

nsi

teY

es66

5(2

0.1)

356

(56.

0)0.

85(0

.79–

0.92

)26

3(4

2.6)

0.78

(0.6

7–0.

89)

No

2646

(79.

9)17

00(6

6.0)

Ref

.14

03(5

4.9)

Ref

.

Inco

nve

nie

nt

open

ing

hou

rsof

imm

un

izat

ion

site

Yes

675

(20.

9)38

3(5

9.6)

0.91

(0.8

4–0.

99)

289

(45.

1)0.

82(0

.75–

0.92

)N

o25

87(7

9.1)

1646

(65.

3)R

ef.

1359

(54.

5)R

ef.

Lack

ofac

cura

tein

form

atio

nY

es11

25(3

5.5)

666

(61.

8)0.

95(0

.89–

1.01

)55

9(5

1.6)

No

2134

(64.

5)13

59(6

5.2)

Ref

.10

86(5

3.0)

Cos

tof

vacc

ines

Yes

1579

(49.

0)99

2(6

3.9)

0.99

(0.9

3–1.

05)

815

(52.

3)0.

97(0

.89–

1.07

)N

o16

93(5

1.0)

1047

(64.

6)R

ef.

835

(52.

5)R

ef.

Cos

tof

visi

tto

the

doc

tor

Yes

1265

(39.

4)79

5(6

3.8)

0.99

(0.9

3–1.

06)

634

(52.

0)1.

00(0

.92–

1.08

)N

o20

16(6

0.6)

1244

(64.

4)R

ef.

1019

(52.

7)R

ef.

Un

frie

nd

lybe

hav

ior

ofim

mu

niz

atio

nst

affo

rba

dor

gan

izat

ion

ofse

rvic

esY

es28

0(8

.4)

146

(55.

3)0.

85(0

.75–

0.97

)11

7(4

6.2)

0.99

(0.9

1–1.

07)

No

2960

(91.

6)18

63(6

4.9)

Ref

.15

12(5

2.8)

Ref

.0.

87(0

.73–

1.04

)R

ef

aW

eigh

ted

pro

por

tion

that

allo

ws

for

sam

ple

wei

ghts

and

clu

ster

ing

wit

hin

sch

oolc

lass

es.

bR

elat

ive

Ris

k.c

95%

Con

fid

ence

inte

rval

.

(2010) 1861–1869 1867

However, long distance to the place of immunization was the onlyperceived obstacle that consistently remained a significant predic-tive factor for both under- and delayed vaccination. This may partlyreflect lack of transportation, long travel times, lack of time avail-able for preventive services or general inaccessibility to preventivehealth care. Structural and health care system factors have beenshown to have an important effect in achieving desired immuniza-tion rates [32].

Despite previous reports that immunizations are no longer ahigh priority for many parents of children seen by physicians [24],this study has shown that parents strongly believe that vaccinesare essential to their children’s health, with most respondentsholding beliefs that would seem to favor immunization. Neverthe-less, in our study, overall positive opinions regarding vaccinationshave not been shown to significantly affect receipt of childhoodimmunization. Our study suggests that socioeconomic factors playa more important role in immunization coverage than parentalperceptions. Alternatively, the limited variability of responsesregarding beliefs, with parents/guardians holding uniformly favor-able attitudes towards vaccination, made it difficult to examine thepotential associations between those and vaccination status.

Despite the general agreement that vaccinations are benefi-cial, one fifth of parents/guardians thought that natural childhooddisease may be preferable to vaccination, possibly resulting fromperceptions of natural lifestyle or from a widespread lack of aware-ness of the hazards of vaccine-preventable diseases. Since mostparents/guardians grew up when vaccines were available, theymay have little or no firsthand knowledge about the seriousnessof the diseases prevented by these vaccines, which is consis-tent with several survey findings [49]. In addition, this was theonly perception that remained a significant independent predictorof incomplete but not of delayed vaccination. Alternative healthbeliefs ware reported as a risk factor for undervaccination in arecent German study [23]. Nevertheless, other studies have notfound such an association. It is possible, however, that different cul-tural factors in other countries and ethnic groups may explain thisdiscrepancy.

Contrary to expectations, the majority of parents did not seemconcerned about the side effects and contraindications of immu-nization. Concerns about vaccine safety have been cited as one ofthe most important obstacles to immunization [39,41–46,49–52].Perhaps concerns about adverse events following vaccinationswere low because of little coverage on vaccine safety in theGreek media. Alternatively, it could reflect the usually inadequateinformation provided to parents by physicians during vaccineadministration.

5.2. Strengths and limitations

This was a large study with adequate power to detect smalldifferences in immunization rates between different groups. Itassessed the effect of a wide range of factors including parentalbeliefs and attitudes, perceived barriers and socioeconomic factorson vaccination. The analysis was based on a nationally representa-tive sample of school children living in both urban and rural areas,with response rates high enough to avoid selection bias. Vacci-nation status data were derived from the children’s vaccinationbooklets, previously shown to be a very reliable source of suchinformation [53], and thus recall bias was not likely.

One limitation of the study is that Roma children attendingschool might not be representative of all children belonging to this

group, as school attendance rate in this group is rather low (about50% according to some estimates) [54]. The same is true for theirparents/guardians, as those who send their children to school maybe more likely to comply with immunization recommendations.However, we do not think that the above issue applies to children

1868 K. Danis et al. / Vaccine 28 (2010) 1861–1869

Table 4Predictive factors for being fully and age-appropriately vaccinated from logistic regression analysis (final model).

Variable Adjusted RRa for complete vaccination finalmodel (95%CIb) N = 3007

Adjusted RRa for age-appropriate vaccinationfinal model (95%CIb) N = 2950

Minority groupNon-minority population Ref. Ref.Immigrants 0.50 (0.36–0.93) 0.38 (0.28–0.51)Greek muslims 0.84 (0.70–1.01) 0.66 (0.50–0.88)Roma 0.45 (0.36–0.57) 0.13 (0.06–0.52)

Number of other siblingsOnly child Ref. Ref.1–2 other children 0.97 (0.89–1.06) 0.89 (0.81–0.98)≥3 other children 0.71 (0.59–0.85) 0.67 (0.53–0.84)

Long distance to immunization siteNo Ref. Ref.Yes 0.93 (0.86–1.01) 0.87 (0.76–0.99)

Mother’s age at child’s birth (years)<25 Ref. NAc

≥25 1.26 (1.03–1.54)

Perception that Natural childhood disease is preferable to vaccinationNo Ref. NAc

Yes 0.92 (0.85–0.99)

Father’s education<9 years (Grammar school) Ref.9–11 years NAc 1.10 (0.94–1.30)12 years (high school) 1.18 (1.02–1.35)College/university graduate 1.23 (1.07–1.42)

oa

bbhpwie5t

ircm

fnacti

5

oriad6si

a Relative Risk adjusted for all variables in the table (final model).b 95% Confidence interval.c Non-applicable.

f the other minority groups, as Grammar school attendance ratesmong 6-year olds have been consistently exceeding 98.5% [54].

A second limitation of our study is that we compared currenteliefs and attitudes with previous immunizations. Because theeliefs were measured after vaccination, some of the attitudes mayave been influenced by the vaccination experiences. Therefore,art of our findings may have been different if parental beliefsere measured prospectively (i.e. prior to the beginning of the

mmunization). Furthermore, we estimated the vaccination cov-rage among 6-year olds, who had mostly been vaccinated almostyears earlier. This limits the timeliness of the information needed

o guide interventions.Information on barriers to vaccination, such as distance to the

mmunization site, was gathered from the parental questionnaireather than being measured. In addition, the use of parental edu-ation might not be the best measure of socioeconomic status, as itay also reflect parental motivation, awareness and coping skills.Finally, although we examined a wide range of potential risk

actors for delayed immunization or undervaccination, data haveot been gathered on some previously identified predictors such asdequate use of prenatal care, mother’s locus of control towards thehild’s health and paediatricians’ immunization beliefs and prac-ices. It is possible that physicians’ behavior might also have anmpact on immunization status of children.

.3. Conclusions/recommendations

The data in this large study offer insight into the relevant effectf socioeconomic factors and parental attitudes on immunizationates in Greece. The results suggest that belonging to a minor-ty group, presence of other siblings, father’s education, maternal

ge, distant immunization locations and the perceived severity ofisease are important determinants of vaccination status among-year-old school children in Greece. These findings indicate thatocioeconomic factors may be more important determinants ofmmunization coverage than parental perceptions.

Based on the findings of this study, the following identifiedgroups merit increased attention in future interventions designedto improve immunization delivery in Greece: (i) minority groups,especially Roma and immigrants, (ii) families with many children,(iii) young mothers, and (iv) households headed by fathers withlow educational level, possible reflecting low socioeconomic sta-tus of the families. Interventions aimed at those high-risk families,although difficult to deliver, may have the greatest effects on com-munity immunization rates. In addition, parents/guardians mustbe educated about the hazards of diseases that vaccinations pre-vent and the likely susceptibility of their children to these diseases.There is a need for policies to overcome structural and health caresystem barriers to immunization, and identify effective and com-prehensive approaches for improving the immunization levels ofchildren in high-risk groups.

Acknowledgements

We would like to thank Dedoukou Xanthi, Iliodromiti Zoi,Kavaka Niki, Menegas Damianos, Milona Fani, OikonomakosDimotrios, Benos Alexis, Spala Georgia, Antoniadou Ioanna, VovaMagda, Milona Eleni, Gavana Magda, Giannakopoulos Stathis,Koutis Antonis for their valuable help in coordinating the fieldworkof the study; Katerelos Panos for helping with the sampling design;Pandeli Niki and Sirmopoulou Ioanna for data entry; GeorgiadouLina for her valuable administrative support. We would also like tothank all the 234 health professionals for their participation in thefieldwork and data collection. Finally, we would like to thank TomNichols from the Health Protection Agency, London for statisticaladvice and Hynek Pikhart, from University College London for hiscomments.

References

[1] WHO. Expanded programme on immunization. In: The EPI Coverage Surveytraining for Mid-level managers. WHO; 1991 [WHO/EPI/MLM/91.10].

ine 28

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[V, et al. Validity of “Child’s Vaccination Booklet” in Greek childhood population.

K. Danis et al. / Vacc

[2] Impact of vaccines universally recommended for children—United States,1990–1998. MMWR 1999;48:243–8.

[3] Bedford H, Elliman D. Concerns about immunisation. BMJ 2000;320:240–3.

[4] GIVS: Global Immunization Vision and Strategy 2006–20151. Geneva:WHO/UNICEF; 2005. Available from: http://whqlibdoc.who.int/hq/2005/WHO IVB 05.05.pdf.

[5] Rodewald LE, Szilagyi PG, Shiuh T, Humiston SG, LeBaron C, Hall CB. Is under-immunization a marker for insufficient utilization of preventive and primarycare? Arch Pediartr Adolesc Med 1995;149:393–7.

[6] Luman ET, McCauley MM, Shefer A, Chu SY. Maternal characteristics associatedwith vaccination of young children. Pediatrics 2003;111(5):1215–8.

[7] Kim SS, Frimpong JA, Rivers PA, Kronenfeld JJ. Effects of maternal and providercharacteristics on up-to-date immunization status of children aged 19 to 35months. Am J Public Health 2007;97(February (2)):259–66 [Epub 2006 Decem-ber 28].

[8] Steyer TE, Mainous 3rd AG, Geesey ME. The effect of race and residence on thereceipt of childhood immunizations: 1993–2001. Vaccine 2005;23(February(12)):1464–70.

[9] The National Vaccine Advisory Committee. Strategies to sustain success inchildhood immunizations. JAMA 1999;282:363–70.

10] Santoli JM, Huet NJ, Smith PJ, Barker LE, Rodewald LE, Inkelas M, et al. Insur-ance status and vaccination coverage among US preschool children. Pediatrics2004;113(Suppl. (June)):1959–64.

11] Smith PJ, Santoli JM, Chu SY, Ochoa DQ, Rodewald LE. The association betweenhaving a medical home and vaccination coverage among children eligible forthe vaccines for children program. Pediatrics 2005;116(1):130–9.

12] Allred NJ, Wooten KG, Kong Y. The association of health insurance and con-tinuous primary care in the medical home on vaccination coverage for 19- to35-month-old children. Pediatrics 2007;119(Suppl. February (1)):S4–11.

13] Theeten H, Hens N, Vandermeulen C, Depoorter AM, Roelants M, Aerts M, etal. Infant vaccination coverage in 2005 and predictive factors for complete orvalid vaccination in Flanders, Belgium: an EPI-survey. Vaccine 2007;25(June(26)):4940–8 [Epub 2007 April 4].

14] Bobo JK, Gale JL, Thapa PB, Wassilak SG. Risk factors for delayed immunizationin a random sample of 1163 children from Oregon and Washington. Pediatrics1993;91(February (2)):308–14.

15] Fiks AG, Alessandrini EA, Luberti AA, Ostapenko S, Zhang X, Silber JH. Iden-tifying factors predicting immunization delay for children followed in anurban primary care network using an electronic health record. Pediatrics2006;118(December (6)):e1680–1686 [Epub 2006 November 6].

16] Smith PJ, Chu SY, Barker LE. Children who have received no vaccines: who arethey and where do they live? Pediatrics 2004;114(1):187–95.

17] Bates AS, Wolinsky FD. Personal, financial, and structural barriers toimmunization in socioeconomically disadvantaged urban children. Pediatrics1998;101:591–6.

18] Wood D, Donald-Sherbourne C, Halfon N, Tucker MB, Ortiz V, Hamlin JS, et al.Factors related to immunization status among inner-city Latino and African-American preschoolers. Pediatrics 1995;96(August (2 Pt 1)):295–301.

19] Smith PJ, Stevenson J, Chu SY. Associations between childhood vaccinationcoverage, insurance type, and breaks in health insurance coverage. Pediatrics2006;117(June (6)):1972–8.

20] Rosenthal J, Rodewald L, McCauley M, Bermans S, Irigoyen M, Sawyer M,et al. Immunization coverage levels among 19- to 35-month old childrenin 4 diverse, medically underserved areas of the United States. Pediatrics2004;113(4):E296–302.

21] O’Connor KS, Bramlett MD. Vaccination coverage by special health care needsstatus in young children. Pediatrics 2008;121(April (4)):e768–74.

22] Swennen B, Van Damme P, Vellinga A, Coppieters Y, Depoorter AM. Analy-sis of factors influencing vaccine uptake: perspectives from Belgium. Vaccine2001;20:S5–7.

23] Zucs AP, Crispin A, Eckl E, Weitkunat R, Schlipkoter U. Risk factors for under-vaccination against measles in a large sample of preschool children from ruralBavaria. Infection 2004;32(3):127–33.

24] Taylor JA, Darden PM, Slora E, Hasemeier CM, Asmussen L, Wasserman R.The influence of provider behavior, parental characteristics, and a public pol-icy initiative on the immunization status of children followed by privatepediatricians: a study from Pediatric Research in Office Settings. Pediatrics1997;99(February (2)):209–15 [Erratum in: Pediatrics 1997 September;100(3Pt 1):333.].

25] Bonanni P, Bergamini M. Factors influencing vaccine uptake in Italy. Vaccine

2001;20:S8–12.

26] Schmitt HJ. Factors influencing vaccine uptake in Germany. Vaccine 2001;20:S2–4.

27] Taylor JA, Darden PM, Brooks DA, Hendricks JW, Wasserman RC, Bocian AB.Pediatric Research in Office Settings; National Medical Association. Associationbetween parents’ preferences and perceptions of barriers to vaccination and the

[

(2010) 1861–1869 1869

immunization status of their children: a study from Pediatric Research in OfficeSettings and the National Medical Association. Pediatrics 2002;110(December(6)):1110–6.

28] Gust DA, Strine TW, Maurice E, Smith P, Yusuf H, Wilkinson M, et al.Underimmunization among children: effects of vaccine safety concerns onimmunization status. Pediatrics 2004;114(July):e16–22.

29] Jheeta M, Newell J. Childhood vaccination in Africa and Asia: the effects ofparents’ knowledge and attitudes. Bull WHO 2008;86(June (6)):419.

30] Bardenheier B, González IM, Washington ML, Bell BP, Averhoff F, Massoudi MS,et al. Parental knowledge, attitudes, and practices associated with not receiv-ing hepatitis A vaccine in a demonstration project in Butte County, California.Pediatrics 2003;112(October (4)):e269.

31] Santoli JM, Szilagyi PG, Rodewald LE. Barriers to immunization and missedopportunities. Pediatr Ann 1998;27:366–74.

32] Kimmel SR, Madlon-Kay DJ, Burns IT, Admire JB. Breaking the barriers to child-hood immunization. Am Fam Physician 1996;53:1648–56.

33] Brenner RA, Simons-Morton BG, Bhaskar B, Das A, Clemens JD, NIH-D.C. Initia-tive Immunization Working Group. Prevalence and predictors of immunizationamong inner-city infants: a birth cohort study. Pediatrics 2001;108:661–70.

34] Strobino D, Keane V, Holt E, Hughart N, Guyer B. Parental attitudes do notexplain underimmunization. Pediatrics 1996;98:1076–83.

35] Taylor JA, Cufley D. The association between parental health beliefs and immu-nization status among children followed by private pediatricians. Clin Pediatr1996;35:18–22.

36] Miller LA, Hoffman RE, Barón AE, Marine WM, Melinkovich P. Risk factors fordelayed immunization against measles, mumps, and rubella in Colorado twoyear-olds. Pediatrics 1994;94:213–9.

37] Wright PF. Global immunization—a medical perspective. Soc Sci Med 1995;41(September (5)):609 16.

38] Bates AS, Fitzgerald JF, Dittus RS, Wolinsky FD. Risk factors for underimmu-nization in poor urban infants. JAMA 1994;272(October (14)):1105–10.

39] Prislin R, Dyer JA, Blakely CH, Johnson CD. Immunization status and sociodemo-graphic characteristics: the mediating role of beliefs, attitudes, and perceivedcontrol. Am J Public Health 1998;88(December (12)):1821–6.

40] Greek Ministry of Health and Social Solidarity. New national vaccination pro-gramme. Circular: Y1/�.�. 4986/13-1-2006; Athens, Greece.

41] Panagiotopoulos T. Methodological problems in research on immunizationcoverage in Greece. A systematic review of studies 1970–1995. Paediatriki1999;62:116–30 [In Greek].

42] Panagiotopoulos T, Valassi-Adam E, Sarafidou E, Mandeki A, Stratiki Z, Benos A,et al. National vaccination coverage survey. Arch Greek Med 1999;16:154–62[In Greek].

43] Panagiotopoulos T, Valassi-Adam E, Sarafidou E, Benos A, Livaditi V, Papadim-itropoulou A, et al. National vaccination coverage survey 2001: coverage for DTPand MMR. In: Proceedings of the 4th panhellenic conference of public healthand health services. 2002.

44] Institute of Education of foreign students in Greece (IPODE). Report on foreignand minority school children; 2004. Athens, Greece.

45] Panagiotopoulos T, Antoniadou I, Valassi-Adam E. Incidence of congenitalrubella in Greece. BMJ 2000;321(November (7271)):1287.

46] Georgakopoulou T, Grylli C, Kalamara E, Katerelos P, Spala G, PanagiotopoulosT. Current measles outbreak in Greece. Euro Surveill 2006;11(February (2))[E060223.2].

47] Panagiotopoulos T, Georgakopoulou T. Epidemiology of rubella and congenitalrubella syndrome in Greece, 1994–2003. Euro Surveill 2004;9(April (4)):17–9.

48] Kyrozis A, Psaltopoulou T, Stathopoulos P, Trichopoulos D, Vassilopoulos D, Tri-chopoulou A. Dietary lipids and geriatric depression scale score among elders:The EPIC-Greece cohort. J Psychiatr Res 2008;(October).

49] Brody JE. Skipping vaccinations puts children at risk, vol. 11. New York Times(August); 1993 [Sect C].

50] Ramsay ME, Yarwood J, Lewis D, Campbell H, White JM. Parental confidencein measles, mumps and rubella vaccine: evidence from vaccine coverage andattitudinal surveys. Br J Gen Pract 2002;52(November (484)):912–6.

51] Samad L, Tate AR, Dezateux C, Peckham C, Butler N, Bedford H. Differences inrisk factors for partial and no immunisation in the first year of life: prospectivecohort study. BMJ 2006;332(7553):1312–3.

52] Matsumura T, Nakayama T, Okamoto S, Ito H. Measles vaccine coverage andfactors related to uncompleted vaccination among 18-month-old and 36-month-old children in Kyoto, Japan. BMC Public Health 2005;5.

53] Perifanou D, Panagiotopoulos T, Valassi-Adam E, Sarafidou E, Benos A, Livaditi

In: Proceedings of the 4th panhellenic conference of public health and healthservices. 2002.

54] Maratou-Alibrabti L, Bagavos C, Papadakis, Paliakou B. Evolution of popula-tion and education in Greece. Athens: National Centre of Social Research; 2002[In Greek].