Two decades of regional trends in vaccination completion and coverage among children aged 12-23 months: an analysis of the Uganda Demographic Health Survey data from 1995 to 2016
BMC Health Services Research volume 22, Article number: 40 (2022)
Childhood vaccination is an important public health intervention but there is limited information on coverage, trends, and determinants of vaccination completion in Uganda at the regional level. We examined trends in regional vaccination coverage and established the determinants of vaccination completion among children aged 12-23 months in Uganda.
We analyzed data from the women’s questionnaire for the 1995-2016 Uganda Demographic Health Survey (UDHS). Vaccine completion was defined as having received a dose of Bacillus-Calmette Guerin (BCG) vaccine; three doses of diphtheria, pertussis, and tetanus (DPT) vaccine; three doses of oral polio vaccine (OPV) (excluding OPV given at birth); and one dose of measles vaccine. We performed Chi-square tests to compare vaccination completion by socio-demographic factors stratified by 10 sub-regions: Eastern, East Central, Central 1, Central 2, Kampala, Karamoja, North, Western, West Nile, and Southwest. We performed logistic regression analysis for each of the regions to identify factors associated with vaccination completion at 5% level of statistical significance.
Overall vaccination completion was 48.6% (95%CI, 47.2, 50.1) and ranged from 17.3% in Central 1 to 65.9% in Southwest. Vaccination completion rates declined significantly by 10.4% (95% confidence interval (CI), − 16.1, − 4.6) between 1995 and 2000, and increased significantly by 10.0% (95% CI, 4.6, 15.4) between 2000 and 2006, and by 5.4% (95% CI, 0.2, 10.6) between 2006 and 2011. Maternal education (secondary or higher level), receipt of tetanus toxoid (TT) during pregnancy, and possession of a child health card were associated with vaccination completion across all the sub-regions. Other factors like place of residence, religious affiliation, household wealth, maternal age, childbirth order, size of child at birth, and place of delivery were associated with vaccination completion but differed between the 10 sub-regions.
Besides considerable regional variations, the vaccination completion rate among children aged 12-23 months in Uganda remains suboptimal despite the availability of vaccines. Maternal education, receipt of TT, and possession of a child health card are associated with a higher likelihood of vaccination completion among children aged 12-23 months in all the regions of Uganda. Interventions to improve the utilization of vaccination services in Uganda should consider these factors.
Childhood vaccination is a cost-effective public health intervention outlined in the public health armamentarium and a key pillar for achieving Universal Health Coverage (UHC) and the Sustainable Development Goals (SDGs) . Despite interventions to improve vaccination coverage and the subsequent achievements gained in the last 5 years (2008-2014), the third dose coverage of the diphtheria, pertussis, and tetanus (DPT3) vaccine in sub-Saharan Africa (SSA) has plateaued at 72% . For example in 2017, about 10 million children were neither vaccinated or under-vaccinated in the sub-Saharan African region  and this potentially exposes them to vaccine-preventable diseases. Consequently, the burden of vaccine-preventable diseases in the region remains high, with an estimated 31 million children under the age of five suffering from vaccine-preventable diseases annually and at least half a million deaths occur due to a lack of access to vaccines . Sub-Saharan Africa, therefore, contributes to between 45 and 50% of severe morbidity and fatalities from leading vaccine-preventable diseases and this estimate is projected to reach 60% in 2030 if no improvements in the deployment of vaccines occur .
In Uganda, the Uganda National Expanded Programme for Immunization (UNEPI) established in 1983 is mandated to ensure that all children under the age of 1 year, 10-year-old girls, and all women of childbearing age receive full vaccination of high quality and effective vaccines. Ever since UNEPI has focused on optimal deployment and delivery of vaccines that target childhood preventable diseases namely tuberculosis, polio, whooping cough, diphtheria, tetanus, measles, hepatitis B virus infection, cancer of the cervix, Haemophilus influenzae Type B, and pneumococcal infections. To further improve access, coverage, and vaccination completion in Uganda, UNEPI has instituted several vaccination strategies such as community-based outreaches, home-based vaccination, child day plus, and mass immunization campaigns that have been employed over the years. Despite these strategies, vaccination coverage in Uganda has remained relatively low over the past years. For example, DPT-3 and the first dose of measles vaccine coverage have stagnated at 79 and 80%, respectively .
National data further show that merely 55% of children aged 12-23 months have received all the basic vaccines. Moreover, vaccination is mandatory for children aged 12-23 months and the expectation is that by the end of 12 months, the children should receive all the vaccine doses. In addition, there are wide regional disparities in the completion of vaccination, with the East Central (Busoga) sub-region standing at 45% and the Karamoja sub-region at 73% , leaving the majority of the children either under or not vaccinated. This might increase their risk of morbidity and mortality from vaccine preventable diseases.
Although the determinants of vaccination completion such as maternal education status, antenatal care visits, institutional delivery, maternal age, and household wealth index [4,5,6,7,8] have been well documented, there is limited information on the determinants of vaccination completion at the regional level despite substantial differences in the regional vaccination coverage. Furthermore, previous studies conducted in Uganda [9,10,11,12] have focused on factors at the national level while others have looked at specific regions and/or districts besides using data for one period only. Therefore, the studies do not provide a clear picture of the immunization program in the country and have not identified regional variations including the determinants of vaccination completion for targeted action. Furthermore, although vaccination coverage trends have been reported in the Uganda Demographic Health Surveys (UDHS), an analysis of the regional trends in vaccination coverage is limited. Informed by these gaps, we aim to describe the regional trends in vaccination coverage and to identify the determinants of vaccination completion at the regional level among children aged 12-23 months in Uganda using data from Demographic Health Surveys datasets from 1995 to 2016.
Methods and materials
Uganda is a landlocked country in East Africa, with a total population of 34,634,650 people according to the 2014 census, which is the most recent survey and 45% (15,585,593) of the total population are ≥18 years. There are 134 districts and 6937 health facilities in Uganda, from the latter, 45.2% (3133) are public health facilities. The health facilities are distributed as follows: five national referral hospitals, 14 regional referral hospitals, 169 general hospitals, 194 health center IVs or county-level health facilities, and the rest are health center IIIs or sub-county-level health facilities and health center IIs or parish level health facilities. Uganda has five super-specialized hospitals and two specialized institutes, the Uganda Heart Institute and Uganda Cancer Institute . The country has 10 sub-regions: Central 1, Central 2, East Central, Eastern, Kampala, Karamoja, Northern, Western, West Nile, and Southwest.
Vaccination schedules in Uganda
According to the UNEPI schedule, a child must receive vaccination 5 times before the first birthday: at birth (BCG, Oral Polio Vaccine (OPV) 0), at 6 weeks (DPT-HEPB-HIB 1, PCV1, Rotavirus vaccine 1, and OPV 1); at 10 weeks (OPV 2, DPT-HEPB-HIB 2, PCV2, Rotavirus vaccine 2); at 14 weeks (OPV 3, DPT-HEPB-HIB 3, IPV, PCV3) and 9 months (measles vaccine). The DPT vaccine is part of a pentavalent vaccine that protects a child from 5 life-threatening diseases namely diphtheria, pertussis, tetanus, hepatitis B, and Haemophilus Influenzae type B. Despite the introduction of new vaccines such as the pentavalent DPT-HepB-Hib vaccine, 10-valent pneumococcal conjugate vaccine (PCV), rotavirus vaccine and inactivated polio vaccine (IPV) between 1995 and 2016 to routine vaccination, the schedule for the basic vaccines has remained the same since 1993 [14,15,16,17,18].
Data source, study design, and population
We conducted a secondary analysis of data from five consecutive UDHS, which are national cross-sectional surveys conducted from 1995 to 2016. The surveys are conducted every 5 years to provide an up-to-date estimate of basic demographic and health indicators. To access the UDHS data, we developed a research concept note and registered the request on the Demographic and Health Surveys (DHS) program website at http://www.DHSprogram.com, stating clearly the research question, the variables needed, time period, country, and the proposed analysis plan. Following the review of the request, we received approval from the DHS Program to retrieve and analyze the dataset. Our study population consisted of 8036 children aged 12-23 months who were alive at the time of the data collection. In the UDHS, a stratified multi-stage sampling technique was applied to select the enumeration areas, both in urban and rural settings and households, using a list of households compiled for all the selected enumeration areas. All women aged 15-49 years in the selected households were eligible to be interviewed using the women’s health questionnaire.
Eligible women were asked questions on background characteristics such as age, level of education, media exposure, birth history, and childhood mortality, knowledge and use of family planning methods, fertility preferences, antenatal care, delivery, and postnatal care, breastfeeding and infant feeding practices, vaccination and childhood illnesses, marriage and sexual activity, women’s work and husband’s background characteristics, awareness and behaviour regarding human immunodeficiency virus (HIV) and Acquired Immunodeficiency Syndrome (AIDS) and other sexually transmitted infections, adult mortality including maternal mortality, and knowledge of tuberculosis and other health issues. In the present study, we only extracted information on background characteristics, and maternal and child health.
The study outcome was vaccination completion, defined as receipt of all of the following antigens: 1) a dose of Bacillus-Calmette Guerin (BCG) vaccine at between birth and 2 weeks; 2) three doses of diphtheria, pertussis, and tetanus (DPT) with the first, second, and third doses at 6, 10, and 14 weeks, respectively; 3) three doses of OPV (excluding OPV given at birth), with the first, second, and third doses at 6, 10, and 14 weeks, respectively; 4) one dose of measles vaccine at 9 months. Any child who has missed any one or more of these vaccines/doses was regarded to had not completed their vaccination schedule.
The independent variables were categorized into four: 1) child related characteristics namely, sex (male or female), birth order (1, 2-3, 4-5, and 6), and size of child at birth (very large, above average, average, below average, and small); 2) maternal related factors such as age in absolute years, education (no education, primary, secondary, and higher), occupation (not working versus working), marital status (never married, married, or separated /divorced/widowed); 3) household related characteristics included wealth index (lowest, second, middle, fourth and highest quantiles), religion (Catholics, Protestants, Muslims, and others), and place of residence (rural versus urban); and 4) the health services-related factors included antenatal care visits (none, 1-3, and ≥ 4), tetanus toxoid injections (none versus yes/ever received), and place of delivery (home, public health facility, private health facility, and others). In the UDHS, data about vaccination coverage were retrieved from the child health card and maternal self-reporting if the child health card was not available.
To determine trends in the regional vaccination coverage, we performed descriptive analyses using the chi-square test to examine regional differences in vaccination coverage, vaccination completion, and the five-year percentage change in vaccination completion rate. To assess the relationship between independent variables and vaccination completion at the regional level, we selected four sub-regions, namely East central, Central 1, Northern, and Southwest, based on their vaccination completion status.
In particular, we included sub-regions with a sample size of at least 50 children aged 12-23 months in a given year. We compared categorical variables with the outcome variable using the Chi-square test and considered variables with probability values (p-values) less than 0.05 as statistically significant. All variables that were statistically significant in the bivariate analysis and those that we deemed are biologically plausible to explain the outcome were considered for the multivariate logistic regression modeling. We fitted a stepwise regression model using a backward selection approach and factors with p-value < 0.05 were retained in the final model. We further grouped predictors into four categories of household, maternal, child, and health utilization predictors. Crude odds ratio (OR) and adjusted odds ratios (AORs) along with the 95% confidence intervals (CIs) were calculated as a measure of association between the independent variables and vaccination completion. Weighting was performed at every level of analysis to account for the sampling design and the overall analysis was performed in Stata version 14.0  and R version 3.5.1 .
Trends in vaccination completion across sub-regions and socio-demographic characteristics by year
The proportion of fully vaccinated children aged 12-23 months varied at national and sub-regional levels in Uganda. The proportions ranged from 28.2% (95% CI, 21.0, 36.6) in East-Central sub-region to 64.2% (95% CI, 55.2, 71.9) in Southwest sub-region in 1995; from 17.3% (95% CI, 10.6, 27.1) in Central 1 sub-region to 47.6% (95% CI, 40.1, 55.3) in Southwest sub-region in 2000; from 38.9% (95%CI, 30.7, 47.9) in East-Central to 52.7% (95%CI, 41.6, 63.4) in Western sub-region in 2006; from 38.6% (95%CI, 29.4, 48.7) in East-Central to 67.5%, (95%CI, 54.4, 78.3) in Kampala in 2011; and from 45.0% (95%CI, 35.8, 54.5) in East-Central to 65.9% (95%CI, 59.1, 72.1) in Southwest in 2016 (Table 1), Karamoja sub-region was excluded due to small sample size. The proportion of fully vaccinated children generally improved in all the sub-regions from 2000 to 2016 after a dramatic decline in most of the sub-regions between 1995 and 2000.
The proportion was slightly above the national level in the Western and Southwest sub-regions and increased gradually in Eastern and West Nile sub-regions, and then remained below the national level in East Central, Central 1, and Central 2 sub-regions throughout the period under review (Fig. 1). Overall, the proportion of fully vaccinated children declined significantly from 1995 to 2000 by 10.4% (95%CI, − 16.1, − 4.6) and consecutively increased significantly from 2000 to 2006 and 2006-2011 by 10.0% (95%CI, 4.6, 15.4) and 5.4% (95%CI, 0.2, 10.6) respectively. By sub-region, the proportion declined significantly in Kampala by 20.7% (95%CI, − 38.8, − 2.6), Central 1 by 38.2% (95%CI, − 51.1, − 25.3), Central 2 by 18.1% (95%CI, − 32.5, − 3.7), Western by 25.0% (95%CI, − 45.0, − 5.1) and Southwest by 16.6% (95%CI, − 27.8, − 5.4) from 1995 to 2000. However, between 2000 and 2006, the proportion increased significantly in Central 1 by 25.8% (95%CI, 13.2, 38.5), Central 2 by 18.8% (95%CI, 2.5, 35.1), and Northern by 19.4% (95%CI, 7.1, 31.5) and similarly in Kampala by 21.9% (95%CI, 6.9, 36.9) and Southwest by 15.7%, (95%CI, 0.6, 30.8) in the period between 2006 and 2011 (Table 2).
The proportion of fully vaccinated children aged 12-23 months was also found to vary by key socio-demographic characteristics. Although the proportion of fully vaccinated children declined significantly across all the four-key socio-demographic characteristics from 1995 to 2000, the proportion generally increased from 2000 to 2016 as shown in Fig. 2. Among those with no education, primary and secondary education, the proportion of fully vaccinated children declined significantly between 1995 and 2000 by 8.7% (95%CI, − 17.3, − 0.1), 10.7% (95%CI, − 18.0, − 3.5) and 21.7% (95%CI, − 34.7, − 8.7) respectively, and increased significantly between 2000 and 2006 by 10.8% (95%CI, 2.2, 19.5) and 9.4% (95%CI, 2.6, 16.1) among those with no education and primary education respectively and by 15.0% (95%CI, 4.7, 27.2) among uneducated in the period 2011-2016 (Table 2).
Among Catholics, Protestants and Muslims, the proportion declined significantly by 9.9% (95%CI, − 17.6, − 2.2), 12.8% (95%CI, − 20.9, − 4.7) and 12.4% (95%CI, − 24.2, − 0.7) respectively between 1995 and 2000, and increased significantly in the same groups by 11.2% (95%CI, 4.1, 18.3), 8.5% (95%CI, 0.3, 16.7) and 13.6% (95%CI, 1.1, 26.0) from 2000 to 2006 respectively; however, the proportion only increased significantly among Catholics and other religious groups (Pentecostal, Seventh-day Adventists, Orthodox and Jehovah witness) by 9.3% (95%CI, 2.0, 16.5) and 13.4% (95%CI, 3.5, 23.3) between 2006 and 2011 and 2011 and 2016, respectively (Table 2). Among the rural and urban residents, the proportion of fully vaccinated children declined significantly between 1995 and 2000 by 9.9% (95% CI, − 16.2, − 3.6) and 15.0% (95% CI, − 25.3, − 4.7) respectively, however, the proportion increased significantly among rural residents by 10.0% (95%CI, 4.2, 15.9) and 5.8% (95%CI, 0.6, 10.9) in the period 2000-2006 and 2011-2016 respectively and only increased significantly among the urban residents by 11.7% (95%CI, 0.9, 22.5) in the period 2006-2011 (Table 2). Finally, the proportion declined significantly between 1995 and 2000 by 14.5% (95% CI, − 26.7, − 2.3) and 30.5% (95% CI, − 40.0, − 20.9) among those in the fourth and highest wealth quantile, respectively. During the period 2000-2006, the proportion increased significantly in the same groups by 19.3% (95% CI, 7.1, 31.5) and 14.6% (95% CI, 4.9, 24.3), respectively (Table 2).
Vaccination coverage for basic vaccines (BCG, DPT3, OPV3, and measles)
The vaccination coverage for the basic vaccines (BCG, DPT3, OPV3, and measles) among children aged 12-23 months varied between sub-regions between 1995 and 2016, with the coverage declining in 2000 and then increasing in the subsequent years (Fig. 3). The proportion of children who received BCG ranged from 55.6% (95% CI, 43.9, 66.7) in Central 1 sub-region in 2000 to 99.2% (95% CI, 94.9, 99.9) in Kampala sub-region in 2016. The majority of the sub-regions had high BCG coverage of more than 70%, however, some sub-regions had low coverage for instance the Central 1 at 55.6% in 1995 and Northern at 69.0% in 2000 (Table 1).
The proportion of children who received all the three doses of DPT differed between sub-regions ranging from 23.5% (95% CI, 15.9, 33.2) in the Central 1 sub-region in 2000 to 87.0% (95% CI, 81.2, 91.1) in the Southwest sub-region in 2016. However, the coverage of DPT3 was generally lower than 50% in most of the regions in the years 1995 and 2000. For example in East Central (41.2%), West Nile (47.2%) and Northern (49.1%) in 1995 and Central1 (23.5%), East-Central (36.4%), Central 2 (39.9%) and Northern (41.3%) in 2000 (Table 1).
The proportion of children who received all the three doses of OPV excluding OPV0 was also found to vary greatly between regions ranging from 26.1% (95% CI, 17.2, 37.3) in the Central 1 sub-region in 2000 to 80.3% (95%CI, 72.4, 86.4) in Southwest sub-region in 2011. However, most of the sub-regions had low OPV3 coverage of less than 50% in the years 1995 and 2000, for example, East Central (36.8%), Northern (41.6%) and West Nile (45.6%) in 1995, and Central 1 (26.1%), Central 2 (43.2%), East Central (45.6%) and Northern (49.7%) in 2000 (Table 1). Similarly, the proportion of children who received the measles vaccine was found to vary between sub-regions ranging from 39.7% (95% CI, 29.4, 50.9) in the Central 1 sub-region in 2000 to 86.3% (95% CI, 80.4, 90.4) in Southwest sub-region in 2016. However, in the years 1995 and 2000, the measles vaccination coverage was less than 50% in most of the sub-regions, namely East Central (39.3%) and West Nile (47.1%) in 1995 and Central 1 (39.7%), East Central (43.5%), and Northern (47.2%) in 2000 (Table 1). Karamoja sub-region was excluded from the comparison.
Predictors of vaccination completion
We grouped the predictors of vaccination completion among children aged 12-23 months into four: household, maternal, child, and health services utilization characteristics. At the bivariate analysis, place of residence, maternal level of education, possession of health card, use of tetanus toxoid were statistically significant predictors of vaccination completion in all the four sub-regions. Religious affiliation was borderline while wealth index was a statistically significant predictor of vaccination completion in the Northern sub-region.
Birth order and place of delivery were statistically significant predictors of vaccination completion in Central 1, Northern, and Southwest sub-regions. Lastly, antenatal care visit was a statistically significant predictor in East Central, Northern, and Southwest sub-regions (Table 3).
After controlling for other variables in the multivariate analysis, the predictors of vaccination completion among children aged 12-23 months varried between the sub-regions. For household characteristics in Central 1, East Central, and Southwest, children aged 12-23 months residing in the urban setting were found to be more likely to complete vaccination compared to those residing in the rural setting AOR (95% CI) 1.85 (1.18-2.89), 1.85 (1.03-3.33) and 1.58 (1.03-2.45). In the Northern sub-region, children aged 12-23 months affiliated to the protestant religion and from middle wealth index households were less likely to complete vaccination compared to those affiliated to the Catholic religion and from the poorest wealth index households AOR (95%CI) 0.67 (0.48-0.93) and 0.52 (0.35-0.79) respectively, while children from highest wealth index households were more likely to complete vaccination compared to those from the lowest wealth index households AOR (95%CI) 3.16 (1.34-7.48) (Table 4). For maternal characteristics, in the Central 1 sub-region, children aged 12-23 months born to mothers aged 25-34 years were more likely to complete vaccination compared to those born to young mothers (15-24 years) AOR (95%CI) 1.51(1.02-2.24). In three sub-regions selected for analysis (Central 1, East Central, and Northern) children aged 12-23 months born to mothers with secondary and higher education were more likely to complete vaccination compared to those born to mothers with no education at all AOR (95%CI) 2.75 (1.47-5.17) and 3.21 (1.43-7.17) in Central 1, 2.94 (1.59-5.45) and 9.50 (2.45-36.85) in East Central, and 4.17(1.69-10.32) and 2.71(1.00-7.32) in Northern, while in Southwest sub-region, only children born to mothers with secondary education were more likely to complete vaccination compared to those born to mothers with no education at all AOR (95%CI) 2.56 (1.51-4.35) (Table 4).
Child characteristics were found to influence vaccination completion differently in the sub-regions. In the Central 1 sub-region, children in the 2-3 birth order were more likely to complete vaccination AOR (95%CI) 1.73 (1.04-2.89) and less likely to complete in Southwest sub-region AOR (95%CI) 0.60 (0.40-0.90) compared to those in the 1st birth order, while in the Northern and Southwest sub-regions, children in the 6+ birth order were less likely to complete vaccination AOR (95%CI) 0.60 (0.41-0.88) and 0.61 (0.41-0.91) respectively compared those in the 1st birth order. We also found that in the Southwest sub-region, children with a birth weight below average (2.5 kg) were less likely to complete vaccination AOR (95%CI) 0.47 (0.23-0.96) compared to those with very large birth weight (Table 4). Finally, the health utilization characteristics were also found to influence vaccination completion differently in the sub-regions. In all four sub-regions children who had a child health card were more likely to complete vaccination AOR (95CI) 13.14 (8.51-20.26) in Central 1, 7.30 (4.96-10.75) in East Central, 10.56 (6.18-18.06) in Northern, and 6.93 (4.93-9.73) in Southwest compared to children who did not have the health card. Similarly, children whose mothers received a tetanus toxoid injection during the pregnancy were more likely to complete vaccination across the four sub-regions AOR (95%CI) 1.87 (1.08-3.26) in Central 1, 1.91 (1.14-3.22) in East Central, 2.63 (1.56-4.42) in Northern and 1.48 (1.01-2.17) in Southwest compared to those whose mothers did not receive the tetanus toxoid injection during pregnancy. Antenatal care visits did not have a statistically significant influence on vaccination completion across the four regions, similarly, place of delivery did not have a statistically significant influence on vaccination completion in three sub-regions (East Central, Northern, and Southwest). However, in Central 1, children delivered at private health facilities and others (roadside) were more likely to complete vaccination for AOR (95%CI) 1.71 (1.05-2.78) and 6.38 (1.28-31.87) respectively compared to those born at home (Table 4).
Our study examined regional trends in vaccination coverage and identified the determinants of vaccination completion at the regional level among children aged 12-23 months using five Demographic Health Surveys in Uganda from 1995 to 2016. We found substantial regional disparities in vaccination completion and vaccination coverage of basic vaccines. We found that the vaccination completion and vaccination coverage rates declined between 1995 and 2000 across the majority of the sub-regions, but with notable improvements across almost all the sub-region between 2000 and 2016, except for Kampala and Western sub-regions where we observed a decline between 2011 and 2016, and between 2006 and 2011 for Central 2 region.
Our data show that vaccination completion and vaccination coverage rates are slightly above the national rates in Western and Southwest sub-regions, although we observed lower rates in East Central, Central 1, and Central 2. The rest of the sub-regions had rates within the range of the national rates. The dramatic decline in vaccination completion and vaccination coverage between 1995 and 2000 could be attributed to a disruption in the healthcare system which was characterized by the adoption of a decentralization approach of health system delivery, poor injection safety measures, and limited funding according to an adhoc report. For instance during the decentralization process in Uganda, the focus of the Uganda Ministry of Health shifted from preventive and primary health care services to private curative services. Furthermore, the decentralization process resulted in reduced funding to services traditionally addressed by vertical programs, an example was immunization program hence reduced number of vaccinators to deliver routine and outreach immunization services . The notable progress in vaccination completion and coverage was a result of the revitalization plan of Uganda expanded immunization program which started around 1999 when there was the remarkable strengthening of infrastructure for delivery of vaccination services, rehabilitation of the cold chain and communication services, capacity building, improving injection safety practices, introducing supplementary immunization activities, renewing high-level advocacy, improving surveillance and adding more vaccines .
Besides, investments into the program from GAVI or the Vaccine Alliance, sustainable outreach services, and the implementation of reaching every district (RED) approach  also contributed to improvements in vaccination completion and coverage between 2000 and 2016.
Furthermore, vaccination completion declined between 1995 and 2000 by key socio-demographic characteristics, namely education, residence, religious affiliation, and wealth index and increase from 2000 to 2016. By 2016, there were almost no disparities in vaccination completion rates by education level, wealth index, and religious affiliation. This could also be due to the above-mentioned reasons; however, a striking result was noted, the vaccination completion rate declined in the urban setting below the rural setting between 2011 and 2016. This decline could be attributed to much more focus on increasing access to health services in the rural settings by constructing more health facilities and implementation of sustainable outreach services in the rural settings.
The study also revealed variations in the predictors of vaccination completion in the four selected sub-regions. We found the maternal level of education, possession of child health card, and receipt of TT injection during pregnancy are associated with vaccination completion in all four sub-regions. Maternal education implies that the mother is able to comprehend information received, which in turn may lead to increased awareness, change of attitude, autonomy and control and decision making. Possession of the child health card enables the parents to remember the child’s immunization schedule, therefore the card acts as a reminder to parents. Mothers who received TT injection during pregnancy are aware of the benefits of immunization and therefore will ensure that their children complete their immunization. These findings are consistent with studies conducted in Togo, Ethiopia, disaster-affected communities in Pakistan, Malawi, and Zimbabwe [24,25,26,27,28]. The findings are also consistent with a study conducted in a Somali national regional state in Ethiopia that found receiving TT injection predicts vaccination completion . The other predictors of vaccination completion were the place of residence in the three sub-regions of Central 1, East, central, and Southwest where children from urban settings were more likely to complete vaccination.
This may be attributed to the higher concentration of both public and private health facilities in the three sub-regions, with the private facilities providing free immunization services thus reducing the distance to the immunization centers. This finding is consistent with a study conducted in Ethiopia  but differs from a study in Bolivia that found children from urban settings were less likely to complete vaccination . The childbirth order was a predictor in the three sub-regions of Central 1, Northern, and Southwest and this is consistent with a study conducted in Zimbabwe . Maternal age and place of delivery were only statistically significant in the Central1 sub-region. This finding is similar to the findings from studies conducted in East Africa [11, 28, 29]. Elsewhere in Greece, a study also found that older mothers tended to complete vaccination compared to younger mothers . Our data show that a child coming from a household affiliated to the Protestant religion and a household with middle and richest wealth status is more likely to complete vaccination in the northern sub-region only. This finding is consistent with studies conducted in Togo, Pakistan, Malawi, and Zimbabwe [24, 26,27,28]. Interestingly, in our study, antenatal care visits are not associated with vaccination completion contrary to studies conducted in Ethiopia , Eastern Uganda , and Zimbabwe .
Study strengths and limitations
This study has several strengths and limitations. By combining datasets from five demographic health surveys, the study had a large sample size to produce unbiased estimates and trend analyses. The data were collected using a standardized data collection procedure and pre-tested tool with good psychometric properties which has been validated at the international level for adoption at the national level. However, there are limitations. We noted that for certain sub-regions such as the Karamoja region, the sample size was too small to detect a statistically significant difference. We, therefore, excluded data from the region in the descriptive analysis of vaccination completion and coverage to minimize biased results. The data analyzed was from cross-sectional surveys so our findings lack temporal association in that they cannot demonstrate a cause-effect relationship. The possibility of reporting biases during the primary survey and its effects on the current study is another important limitation to consider.
We did not have qualitative data to support or explain the findings and this should be a subject for further research.
This study shows that the vaccination completion rate among children aged 12-23 months in Uganda has slightly improved across all the sub-regions. Although vaccination services are readily available, the vaccination completion rate is suboptimal and differs across the majority of the sub-regions. Maternal level of education, possession of a child health card, and receipt of TT injection during pregnancy are associated with vaccination completion among children 12-23 months across all the sub-regions.. Other factors like place of residence, religious affiliation, household wealth, maternal age, childbirth order, size of child at birth, and place of delivery were associated with vaccination completion but differed between the 10 sub-regions. We concluded that there is a need for the Ministry of Health and other implementing agencies to implement targeted interventions to improve the vaccination completion rates in Uganda. Besides, there is a need to conduct additional studies to explore reasons at individual, community, and health care levels that account for failure to complete vaccination despite the availability of vaccination services.
Availability of data and materials
The datasets generated and/or analysed during the current study are available in the www.dhsprogram.com repository.
Adjusted odds ratio
Diptheria, Purtusis, and Tetanus
Inactivated Polio Vacine
Oral Polio Vaccine
Pneunonia Conugate Vaccine
Universal Health Coverage
Uganda National Expanded Program for Immunization
Sustainable Development Goals
Uganda Demographic Health Surveys
Madhi SA, Rees H. Special focus on challenges and opportunities for the development and use of vaccines in Africa. Hum Vaccines Immunother. 2018;14(10):2335–9.
Experts caution against stagnation of immunization coverage in Africa. World Health Organization, Africa Region; 2019. Available from: https://www.afro.who.int/news/experts-caution-against-stagnation-immunization-coverage-africa. Cited 17/09/2019.
UBOS. Uganda demographic and health survey 2016 (UDHS2016). Kampala: Uganda Bureau of Statistics Kampala, Uganda, The DHS Program ICF Rockville, Maryland, USA; 2018.
Girmay A, Dadi AF. Full immunization coverage and associated factors among children aged 12-23 months in a hard-to-reach areas of Ethiopia. Int J Pediatr. 2019;2019:8.
Acharya P, Kismul H, Mapatano MA, Hatløy A. Individual- and community-level determinants of child immunization in the Democratic Republic of Congo: a multilevel analysis. PLoS One. 2018;13(8):e0202742.
Holipah, Maharani A, Kuroda Y. Determinants of immunization status among 12- to 23-month-old children in Indonesia (2008-2013): a multilevel analysis. BMC Public Health. 2018;18(1):288.
Nozaki I, Hachiya M, Kitamura T. Factors influencing basic vaccination coverage in Myanmar: secondary analysis of 2015 Myanmar demographic and health survey data. BMC Public Health. 2019;19(1):242.
Sarker RA, Akram R, Ali N, Chowdhury IZ, Sultana M. Coverage and determinants of full immunization: vaccination coverage among Senegalese children. Medicina. 2019;55(8):480.
Bbaale E. Factors influencing childhood immunization in Uganda. J Health Popul Nutr. 2013;31(1):118–29.
Malande OO, Munube D, Afaayo RN, Annet K, Bodo B, Bakainaga A, et al. Barriers to effective uptake and provision of immunization in a rural district in Uganda. PLoS One. 2019;14(2):e0212270.
Sematimba D. Exploring the predictors of vaccination completeness in the first year of life. Analysis of a community-based household survey in Eastern Uganda: Department of Public Health Sciences Master’s Program in Global Health; 2016.
Ecweu R. Factors influencing coverage of the expended program of immunization in Iganga district, Eastern Uganda. University of Rwanda, East African Centre for Vaccines, Immunization, and Health Supplies Chain Management; 2019.
Javira S. Massive! Uganda’s health sector delivering service with confidence as citizens reap from infrastructure investiment (part II); 2020.
UDHS I. Uganda demographic and health survey. Kampala: Uganda Bureau of Statistics; 1995.
UDHS I. Uganda demographic and health survey. Kampala: Uganda Bureau of Statistics; 2001.
UDHS I. Uganda demographic and health survey. Kampala: Uganda Bureau of Statistics; 2006.
UDHS I. Uganda demographic and health survey. Kampala: Uganda Bureau of Statistics; 2011.
UBOS I. Uganda demographic and health survey. Kampala: Uganda Bureau of Statistics; 2016.
StataCorp. Stata Statistical Software: Release 14. College Station: StataCorp LP; 2015.
Team RC, inventorR: a language and environment for statistical computing. 2020.
Hutchinson P, Akin JS, Ssengooba F. The impacts of decentralization on health behaviors in Uganda. St Chapel Hill: MEASURE Evaluation, Carolina Population Center, University of North Carolina; 2003.
MOH. Expanded programme on immunisation financial sustainability plan. Kampala: Ministry of Health; 2003.
MOH. Uganda national expanded programme on immunization multi year plan 2012-2016. Kampala: Ministry of Health; 2012.
Ekouevi DK, Gbeasor-Komlanvi FA, Yaya I, Zida-Compaore WI, Boko A, Sewu E, et al. Incomplete immunization among children aged 12–23 months in Togo: a multilevel analysis of individual and contextual factors. BMC Public Health. 2018;18(1):952.
Kinfe Y, Gebre H, Bekele A. Factors associated with full immunization of children 12–23 months of age in Ethiopia: a multilevel analysis using 2016 Ethiopia demographic and health survey. PLoS One. 2019;14(11):e0225639.
Rehman SU, Siddiqui AR, Ahmed J, Fatmi Z, Shah SM, Rahman A, et al. Coverage and predictors of routine immunization among 12-23 months old children in disaster affected communities in Pakistan. Int J Health Sci (Qassim) 2017;11(1):1-6.
Ntenda PAM. Factors associated with non- and under-vaccination among children aged 12–23 months in Malawi. A multinomial analysis of the population-based sample. Pediatr Neonatol. 2019;60(6):623–33.
Mukungwa T. Factors associated with full immunization coverage amongst children aged 12–23 months in Zimbabwe. Afr Popul Stud. 2015;29(2):1761–74.
Mohamud AN, Feleke A, Worku W, Kifle M, Sharma HR. Immunization coverage of 12-23 months old children and associated factors in Jigjiga District, Somali National Regional State, Ethiopia. BMC Public Health. 2014;14:865.
Osetinsky B, Gaydos LM, Leon JS. Predictors of completed childhood vaccination in Bolivia. Int J Child Adolesc Health. 2015;8(4):413–23.
Danis K, Georgakopoulou T, Stavrou T, Laggas D, Panagiotopoulos T. Predictors of childhood vaccination uptake: a cross-sectional study in Greece. Procedia Vaccinol. 2010;2(1):86–91.
Tefera YA, Wagner AL, Mekonen EB, Carlson BF, Boulton ML. Predictors and barriers to full vaccination among children in Ethiopia. Vaccines. 2018;6(2):22.
I would like to thank my family for supporting me during my studies. I thank my academic Advisers, Promotors, and Mentors at the University of Antwerp. I thank the Uganda Ministry of Health for technical support, the VLIRUOS for funding my studies, and the DHS for providing data access.
Ethics approval and consent to participate
Data files for the Uganda demographic health surveys (UDHS) from 1995 to 2016 were requested and received from the DHS program (www.dhsprogram.com). The surveys are normally conducted every 5 years and the protocols were approved by the institutional review board and National Council of Science and Technology in Uganda and ICF international. No additional ethical approval was required for the secondary analysis of the datasets.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Okello, G., Izudi, J., Ampeire, I. et al. Two decades of regional trends in vaccination completion and coverage among children aged 12-23 months: an analysis of the Uganda Demographic Health Survey data from 1995 to 2016. BMC Health Serv Res 22, 40 (2022). https://doi.org/10.1186/s12913-021-07443-8