|Home | About | Journals | Submit | Contact Us | Français|
Vaccination coverage studies are important in determining a population's vaccination status and strategically adjusting national immunization programs. This study assessed full and timely vaccination coverage of preschool children aged 2–3 y attending nurseries-kindergartens (N-K) nationwide at the socioeconomic crisis onset. Geographically stratified cluster sampling was implemented considering prefectures as strata and N-K as clusters. The N-K were selected by simple random sampling from the sampling frame while their number was proportional to the stratum size. In total, 185 N-K (response rate 93.9%) and 2539 children (response rate 81.5%) participated. Coverage with traditional vaccines for diphtheria-tetanus-pertussis, polio and measles-mumps-rubella was very high (>95%), followed by Haemophilus influenzae type b and varicella vaccines. Despite very high final coverage, delayed vaccination was observed for hepatitis B (48.3% completed by 12 months). Significant delay was observed for the booster dose of pneumococcal conjugate vaccines (PCV) and meningococcal C conjugate vaccines (MCC). Of the total population studied, 82.3% received 3 PCV doses by 12 months, while 62.3% received the fourth dose by 24 months and 76.2% by 30 months. However, 89.6% received at least one MCC dose over 12 months. Timely vaccinated for hepatitis A with 2 doses by 24 months were 6.1%. Coverage was significantly low for Rotavirus (<20%) and influenza (23.1% one dose). High vaccination coverage is maintained for most vaccines at the beginning of the crisis in Greece. Coverage and timeliness show an increasing trend compared to previous studies. Sustained efforts are needed to support the preventive medicine system as socioeconomic instability continues.
Vaccines save millions of lives and are among the most successful and cost-effective health interventions ever developed.1,2 However, vaccine-preventable diseases still cause significant public health burden worldwide, mainly because of suboptimal vaccination rates,3 while several goals set by the WHO (World Health Organization) failed to be achieved.4,5
The Greek National Immunization Program (NIP) (Table 1) is based on the recommendations of the National Vaccination Committee in agreement with the Ministry of Health.6 Vaccinations are administered to Greek citizens including immigrants in primary health centers and health insurance clinics free of charge or in the private sector and are recorded in the child health booklet (CHB). There is neither a national registry nor an official reminder system.
Systematic monitoring of child vaccination status is performed in many developed countries.7-11 Several studies have been conducted in Greece, in many cases without using representative samples or presenting various methodological problems.12 The most robust data derive from 4 national population-based surveys conducted in 1996–97, 2001, 2006 and 2012. All used representative samples of children in the first grade of primary school or as in the 2001 survey of 2-year olds.13-14 Despite high coverage, outbreaks during the last 15 y highlighted gaps in coverage of certain population groups and in timely administration of specific vaccine doses.15-18 In addition, socioeconomic factors, parental perceptions and concerns regarding vaccine safety along with the growing complexity of immunization schedules undermine adherence to the NIP.19-21
The study aimed to assess full and timely vaccination coverage among preschool children attending nurseries-kindergartens (N-K) nationwide and the socioeconomic crisis impact on it.
In the cross-sectional study participated 185/197 randomly selected N-K (participation rate 93.9%). During the period January-June 2013, 2539/3114 children 2–3 y old provided their CHB to record vaccination history (response rate 81.5%). The mean age of the children was 33.1 months (median 33 months; IQR= 30–36 months, range 24–40 months). Basic characteristics of the study population are presented in Table 2.
Vaccination coverage of the total study population exceeded 95% for most doses of traditional vaccines included in the NIP for years (diphtheria-tetanus-pertussis/acellular pertussis (DTP/DTaP), poliomyelitis (polio) and measles-mumps-rubella (MMR) vaccines). The weighted final coverage of the study population by dose administered at any point of time and population group is presented in Table 3 while proportions of complete vaccination by population group in Table 4. Comparably high was the final coverage with Haemophilus influenzae type B vaccine (Hib) ranging from 94.7% to 100% depending on the dose while newer ones like the varicella vaccine (VAR) reached 92.0%. Despite very high final coverage, delayed vaccination was observed for several vaccines with most notable hepatitis B vaccine (HepB). HepB final coverage ranged from 99.4% to 96.2% according to the dose, nevertheless only 48.0% had completed the first 2 doses by 6 months age (Fig. 1A) and 48.3% the 3-dose schedule by 12 months age. A significant delay was recorded for the fourth dose of pneumococcal conjugate vaccines (PCV) (Fig. 1B) and the booster dose of meningococcal C conjugate vaccines (MCC). Although final coverage with 3 PCV doses surpassed 95%, a significant delay was observed in both timely administration of the fourth dose (13.9% by month 15) and completion of vaccination; 82.3% of the total population received 3 doses by month 12, while 62.3% received the fourth dose by 24 months and 76.2% by the age of 30 months. Coverage with MCC was very high for the primary doses (99.7% and 92.0% for the first and second dose respectively), but considerably lower for the booster dose (77.6%). Nonetheless, 89.6% of the total population was covered with at least one dose of MCC over 12 months. Other remarkable examples of moderate or low coverage were hepatitis A (HepA), rotavirus (RV) and inactivated influenza (IIV) vaccines. A 79.5% of the total population had received one dose of HepA while the rate dropped to 41.7% for the second dose. Timely vaccinated with 2 doses of HepA by 24 months were 6.1%. The rates for RV and IIV were similarly low as fully vaccinated with either the monovalent (RV1) or the pentavalent (RV5) were under 20% of the total study population and only 23.1% had received one dose of IIV.
Children from immigrant families had analogously high coverage for all traditional vaccines with the general population. Among newer vaccines, significant differences in coverage were more apparent in cases of fourth doses as with Hib (p = 0.028) and PCV (p = 0.023) while for MMC the difference was marginally non-significant (p = 0.052). Still, the biggest variation was recorded in the coverage with RV (p < 0,001) and one dose of IIV (p = 0.021). Vaccination timeliness was comparable regarding the traditional vaccines but a significant difference was clearly demonstrated mainly for PCV (by 12 and 24 months) and RV and to a lesser extent for Hib and MMC (by 24 months). Analytical final vaccination rates at any point of time and timely coverage can be found in Table 3 and Table 4 respectively. Differences in coverage were also recorded between groups of development regions (NUTS-1) in several cases but without a particular region constantly presenting reduced coverage. Most vaccinations (73%, range: 67.2–87.8% for various vaccines) were administered through the private sector.
During the onset of the socioeconomic crisis, despite general austerity measures and drastic reduction of public expenditure for health, vaccines have been excluded from rebate imposition and aggressive pricing due to their key role in public health. Until 2014, the socioeconomic crisis seems not to have negatively impacted coverage or timely administration of most vaccines. Comparisons were made with a previous national study in 2012 in 6 y old children, which reflected coverage before the onset of the crisis. However, it should be noted that only timely vaccination coverage can be adequately compared between studies due to cohort age differences.
Vaccination coverage with DTaP, serves as a benchmark as the most established vaccine in the Greek NIP. Rates remained very high for all 4 doses, both for the general population and immigrant children.13,14,20,21 Coverage with 3 DTP doses by 12 months (95.7%), a key WHO indicator, was similar to the estimated coverage for the European region of WHO in 2013.22 Delay in completing 4 doses by 24 months, although lower than in 2012, is possibly due to the increased number of vaccines in the second year of life.
Vaccination with IPV, in the context of multivalent vaccines' use, was equally high in the general population and immigrant children while timeliness rates with 2 and 3 doses by 12 and 24 months respectively (99% level) improved compared to the previous study. It is a prerequisite to maintain polio coverage >95% to stop virus transmission and retain Greece and the European region “polio free”.5 The geopolitical location of Greece accentuates this importance, having neighboring countries where health systems have collapsed and recent outbreaks revive the danger of wild-type poliovirus resurgence.23,24 Furthermore, Greece is facing an increased influx of refugees and illegal immigrants exacerbating the pressure on the public health system.
Measles and rubella elimination remain a top WHO priority for the European Region.4,25 Coverage of the total study population with one MMR dose (97.3%) was well over the required rate (≥95%) to interrupt measles virus transmission, while timeliness by 24 months fell just short (94.5%), with no statistically significant difference between population groups. MMR coverage was slightly lower than before the crisis, but timeliness has improved. In addition, coverage in NUTS-1 regions was well above the required level of ≥80%. However, high and timely coverage with the second dose of MMR (out of scope of the study) remains of utmost importance in achieving adequate measles protection and preventing epidemics locally (last epidemic in 2010–2011)17,18 and at European level.26,27
HepB and Hib were included in the NIP in 1998 and 2002 respectively. Coverage with the first 3 doses of Hib was comparatively high to DTP due to multivalent vaccines. Significant improvement was observed in timely coverage with 3 doses by 12 months (94.9% vs. 76.8% in 2012) together with an >15% increase of coverage with the fourth dose.14
Timely HepB administration is particularly important, as infection at an early age increases the likelihood of chronic carriers. Final coverage with 3 doses was satisfactory, but slightly lower than in 2012 (96.2% vs. 97.8%, p < 0.05). Although delayed administration of the first 2 doses by 6 months (48.0% vs. 42.4% in 2012) remained,13,14 full coverage with 3 doses by 12 and 24 months presented a slow longitudinal improvement. Vaccination was postponed for the second semester in a large proportion of the cohort (Fig. 1A), in concordance with local data and suboptimal coverage rates in other countries.13,14,20,21,28,29 The delay is attributable to the overcrowded vaccination schedule in the first 6 months as well as the non-universal use of hexavalent vaccines. The use of combined vaccines has potential benefits, including reduced number of patient visits, fewer complications associated with multiple injections, decreased stocking and visit costs, and lowering the risk of delayed or missed vaccinations.30,31
Although PCVs (regardless serotype number) were included in the NIP in 2006 with a 3 + 1 schedule, vaccination with 3 doses exceeded 95% in the total study population. The time of implementation in the NIP and the fact that full reimbursement was granted in 2008, so not enabling all children to complete 4 doses, makes comparison with 2012 rates impossible. However, vaccination rates are much higher than those in 2 recent local studies in the Athens area and a prospective one in Athens and Viotia.20,21,32 Apart from the overcrowding of the NIP, reduced coverage with the fourth dose and delayed vaccination completion in the general population and especially immigrant children may be linked to an increasing number of uninsured children during the socioeconomic crisis and the relatively high vaccine cost. Similar observations of non adherence to recommended PCV schedules have been made in other European countries and may impact protection against pneumococcal disease.33 High coverage in the first year of life is important for protection against invasive pneumococcal disease presenting higher incidence in this age group, nasopharyngeal carriage and also the development of a herd effect which is crucial for unvaccinated children and adults. In addition, timely administration of the booster dose has a substantial role in inducing immunologic memory and conferring long-term protection against pneumococcal diseases.
Coverage of the total study population with MCC has greatly improved during the last years. Compared to the national study in 2012, significantly more children received 2 (92.0% vs. 31.3%) and 3 doses (77.6% vs. 17.3%) of MCC.14 Analogous improvement was observed in timely coverage with 2 doses by 12 months (77.2% vs.18.4% in 2012) and 3 doses by 24 months (66.7% vs. 14.5% in 2012). Coverage with one MCC dose in children aged ≥12 months was maintained at satisfactory levels in both studies (89.6% vs. 86.4%), in line with the adoption of just one dose after the first year of life by several European countries.34 The time lapse from its NIP introduction and a price reduction might have contributed to the improved coverage.
Vaccination with the first dose of VAR was very high, despite the few years of NIP inclusion (2006) but below the respective concurrently administered dose of MMR. Lower coverage by 24 months may be attributed to the relatively scarce use of quadrivalent vaccines for fear of potential adverse reactions and also the belief that chickenpox is less serious.
Our study indicates that certain vaccines, as HepA, IIV and RV, might be presumed of lower priority. HepA, was introduced in the NIP and fully reimbursed in 2008 although available in the market since 1999. Coverage with HepA was significantly lower than any other NIP vaccine characterized as “necessary” and showed a potential downward trend compared to the levels of 81.7% recorded in first graders in 2012. Usually, HepA is implemented after the second year as opposed to the NIP. Hence, only a very small number of children have completed the 2-dose schedule by 24 months while final coverage has reached higher levels. Physicians seem to give lower priority to HepA vaccination during the second year of life except in high risk groups and postpone it for later in childhood.14,20 The ongoing crisis might have contributed in making parents more reluctant to pay the visit cost in the private sector for less established vaccines. Greece remains a country of low endemicity based on 2004–2012 data where 29.6% of reported cases were Roma.35 During the recent years small outbreaks have been confined within the Roma population with only geographical correlation with community cases in the general population.35 In addition, hepatitis A hospitalization rate has significantly (p = 0.005) decreased between pre-vaccination (1999–2008) and post-vaccination era (2009–2013) as shown in a recent study from a tertiary pediatric center in Athens possibly due to the implementation of universal vaccination as well as the improvement of sanitation and living conditions.36 However, low coverage and significant delay in HepA vaccination of the general population and children of Roma and immigrants might prove to be insufficient in preventing small outbreaks. These data highlight the need for more efficient vaccination of targeted population groups (Roma) in conjunction with improved education and living conditions.
In Greece, only 2 vaccines were considered as “recommended” and not “necessary” by the NIP: RV and IIV. Although rotavirus is a frequent cause of acute gastroenteritis,37 the low coverage might be attributed to low assessment of vaccine necessity by physicians and parents that reflect a common impression of limited disease burden. Additionally, the relatively high cost of vaccination and partial compensation from major funds further aggravate perception of unfavorable cost-effectiveness. Similar barriers have been identified in many European countries.38 The ongoing crisis might have further negatively affected the uptake. Likewise, perceived need for vaccination against influenza seems to be low. Possible reasons include the fact that according to the NIP the vaccine is indicated for children belonging to high risk groups only and that influenza is not considered as a severe childhood infection by many physicians.
Vaccination coverage studies are an important tool in evaluating vaccination programs and planning appropriate strategic adjustments. In many developed countries studies are based on electronic registers11,39 or executed through telephone interviews and e-mails using structured questionnaires. Greece currently lacks a national electronic register and vaccinations are not administered exclusively in vaccination centers allowing a representative sample to be drawn. Meanwhile, doctors in the public and private sectors do not keep detailed vaccination records permitting information collection through telephone interviews. Thus, Greece relies on periodical vaccination coverage surveys to gather necessary data. Interestingly, in a Belgian study, coverage rates from registrations in the Flemish immunization register (Vaccinnet) were found systematically lower than the corresponding estimates obtained through surveys.40
The sample was large enough to offer sufficient power in detecting small differences in coverage. Response rates were high enough to assume participation bias was averted, while collection of vaccination data from the CHB avoided potential recall bias.
A number of limitations need to be taken into account in the interpretation of the results. The study illustrates the situation 1–2 y ago, a period that coincides with the start of the ongoing Greek socioeconomic crisis. Control of vaccinations from CHB is among the necessary documents for registration/annual renewal in N-K but non-compliance with the current NIP does not constitute a reason for registration denial. However, children this age are not obliged to attend N-K, and possibly children from special population groups have reduced access. The percentage of children from immigrant families was lower than expected (5.5% vs. 8–10% from past studies) while the minimal number of Greek Roma children (0.4% vs. anticipated 1.6% in the total population) did not allow conclusions to be drawn regarding this particular population group, characterized by significantly lower coverage than the general population.13,14 Among factors potentially affecting access to N-K are the dispersion of the pediatric population and parents' economic status. N-K are absent in remote areas and small villages, but the number of children is small and assumed not to significantly affect the representativeness of the sample. Also, the existing agreement, giving priority to children from lower income families to enter public N-K, as well as their percentage in the final sample (including OAED N-K) greatly reduces a potential systematic error in selecting children from upper social classes with expected higher coverage.
This study offers, for the first time, data for children in the important age group of 2–3 y attending N-K on a national level in Greece. Findings should be considered optimistic as high vaccination coverage seems to be maintained for nearly all vaccines. Vaccination rates and timely completion of vaccination, following NIP recommendations, show an increasing trend compared to recent studies. Identified delays in administering primary or booster doses and completing the schedule should be monitored continuously and efforts should aim to avoid missed vaccination opportunities. In addition, population groups with suboptimal vaccination coverage should be targeted for intervention both in terms of education as well as vaccination campaigns. The economic crisis in Greece seems not to have particularly impacted childhood vaccinations during the years of the study; however, this does not permit complacency as a constant effort to strengthen the preventive medicine system is required in years of continuing socioeconomic instability.
A cross-sectional nationwide vaccination coverage study of children born between January–December 2010 was conducted in 2013. The target population included children aged 24–42 months (age was calculated in months taking into consideration the completion of the indicated month of life) i.e. 2–3 full years, living in Greece. Children attending N-K were the sampling population while the sampling frame comprised all public and private N-K operating nationally in the 2012–2013 school-year provided by the Ministry of Interior, the Hellenic Agency for Local Development and Local Government and the Panhellenic Association of Private Nurseries-Kindergartens.
Geographically stratified cluster sampling was implemented considering country prefectures (Eurostat's NUTS-3) as strata and N-K as clusters (primary sampling unit). Within each selected cluster all children 2–3 y old were included. The N-K were selected by simple random sampling from the sampling frame within each stratum while their number was proportional to the stratum size.
The initial target was a representative sample size of around 2400 children, assuming full vaccination coverage of 80% with ±2.25% precision for a 95% CI and a design effect of 2. Accounting for an 80% response rate the final adjusted sample size was 3000 children distributed proportionally to births in 2010 within each stratum. Dividing the needed strata sample size by the estimated strata average number of children aged 2–3 y old per N-K (considering an average of 24, 18 and 12 children per N-K for the prefectures of Attica, Thessaloniki and the rest of Greece respectively) resulted in selecting 197 N-K across the country.
A letter was sent to the headteachers of selected N-K illustrating study details and asking for their voluntary collaboration. Parents received an informational letter and a consent form for their child's participation and the provision of the CHB. Demographics derived from the N-K registry while vaccination status from the CHB. The study received official approval from the Ministry of Interior.
A database was created with EpiData software (Epidata Association, Odense, Denmark, Version 3.1) while data validity was checked by consistency and range checks. Quantitative variables are presented as mean with standard deviation (SD) or range and qualitative variables as frequencies with percentages or with 95% CI. Survey data analysis was conducted to obtain population estimates and their 95% CI in accordance with the sampling plan. Pearson's Chi-square test was used to identify any association between demographic factors and vaccination coverage of children. All statistical analyses were performed using the complex samples module of SPSS 21 (IBM, SPSS Statistics). Results were considered statistically significant when the p-value was less than 0.05.
The authors declare no potential conflicts of interest in preparing this article.
The authors would like to thank the parents of all children that participated in the study and provided their CHB as well as the headteachers and teachers of participating N-K for their contribution in collecting the data. We would also like to address special thanks to Alexandra Vernardaki for the administrative support of the study and data entry, Efthimia Paza, Filippos Markotsis, Panagiotis Katsaounos and Pandelis Mavraganis for the data entry.
TG and JK conceived the study and all authors contributed to the design. TG and CH coordinated the study and all authors contributed to the analysis and interpretation of the data.
TG, DM and AK drafted the manuscript and MT, JK and CH critically edited it. All authors read and approved the final manuscript.