Risk Factors for Incomplete Immunization in Children with HIV Infection
- First Online:
- Cite this article as:
- Bhattacharya, S.D., Bhattacharyya, S., Chatterjee, D. et al. Indian J Pediatr (2014) 81: 850. doi:10.1007/s12098-013-1049-0
To document the immunization rates, factors associated with incomplete immunization, and missed opportunities for immunizations in children affected by HIV presenting for routine outpatient follow-up.
A cross-sectional study of immunization status of children affected by HIV presenting for routine outpatient care was conducted.
Two hundred and six HIV affected children were enrolled. The median age of children in this cohort was 6 y. One hundred ninety seven of 206 children were HIV infected, nine were HIV exposed, but indeterminate. Fifty (25 %) children had incomplete immunizations per the Universal Immunization Program (UIP) of India. Hundred percent of children had received OPV. Ninety three percent of children got their UIP vaccines from a government clinic. Children with incomplete immunization were older, median age of 8 compared to 5 (p = 0.003). Each year of maternal education increased the odds of having a child with complete UIP immunizations by 1.18 (p = 0.008)-children of mothers with 6 y of education compared to those with no education were seven times more likely to have complete UIP vaccine status. The average number of visits to the clinic by an individual child in a year was 4. This represents 200 missed opportunities for immunizations.
HIV infected children are at risk for incomplete immunization coverage though they regularly access medical care. Including routine immunizations, particularly catch-up immunizations in programs for HIV infected children maybe an effective way of protecting these children from vaccine preventable disease.
KeywordsImmunizationsVaccinationPublic health practiceHIVPediatricMaternal education statusUniversal immunization program
HIV infection significantly increases risk for severe disease and death from vaccine preventable pathogens in infected children [1–3]. Immunizations, amongst the most cost-effective interventions in public health, are essential tools in the management of infectious diseases in children with HIV infection [4–6].
HIV infected children however may be at an increased risk for decreased rates of immunization coverage. Studies from sub-Saharan Africa have shown that HIV infected children and the children of HIV infected mothers are at an increased risk for incomplete childhood immunizations. Zambian children with HIV infection have been shown to have almost twice the risk of incomplete immunization compared to uninfected children . Studies from South Africa and Uganda show that children of HIV infected mothers have 30–70 % chance of being incompletely immunized compared to the children of HIV uninfected mothers [8, 9].
Official estimates suggest that 70,000 children are living with HIV in India and annually, an additional 21,000 become infected [10, 11]. Hospitalized HIV infected children in India have significantly increased rates of mortality, much of this due to pneumonia and gastroenteritis which may be vaccine preventable . Immunization coverage in HIV infected children from India needs to be studied. With the introduction and scale up of National AIDS Control Program-III (NACP-III) significant numbers of HIV infected children have been reached and have had access to care and Antiretroviral Therapy (ART). Immunizations, particularly catch-up immunizations are not yet included in these programs.
As part of a nasopharyngeal swab survey in a cohort of HIV infected children from West Bengal, the authors systematically collected data on immunizations . Previously the authors have reported data from this study, in this paper they present data and analysis on immunization rates, factors associated with incomplete immunization, and missed opportunities for immunizations in children affected by HIV presenting for routine outpatient follow-up .
Material and Methods
HIV infected or exposed but indeterminate children between the ages of 0–18 y, presenting for routine outpatient care at the pediatric HIV clinic at Medical College Kolkata, were enrolled in a cross-sectional nasopharyngeal swab survey from March 2008 through April 2009. Hospitalized children and those requiring an indoor admission were excluded. This clinic serves as a referral center for children with HIV infection from the region. Written informed consent was obtained from caregivers and assent from all children over five.
The study was approved by the Institutional Review Boards of three institutions Indian Institute of Technology Kharagpur, the Medical College Kolkata, and the National Institute of Cholera and Enteric Diseases (NICED).
The methods for data collection have been described previously in detail . A questionnaire was developed to capture information on demographics in Bengali. This instrument captured information from the parent or guardian on a child’s age, gender, number of occupants in the house, school attendance, parental status, tuberculosis history, maternal and paternal education and recent antibiotic use. Weight, CD4 count, and antibiotic history were obtained from the medical record.
Immunization history was taken from the immunization card when available or through recall from the parent or guardian by asking preset questions. BCG scar was noted. Data collection was carried out by a physician/translator team at Medical College Kolkata. Clinic registries were reviewed to track the number of visits by patients during the study period.
Children were classified as completely vaccinated if they had received all doses of the vaccines for the appropriate age included in the Universal Immunization Program of India at the time of the survey: BCG vaccine at birth, four doses of (OPV) oral polio vaccine, DTP four doses, Measles at age 9 mo, and DT at age 5 and 10 y. In addition, families were asked about Hepatitis b immunization, other non-UIP vaccine administration, and participation in the oral polio vaccine campaign.
Data was entered and edited in Epi info 3.5 and analyzed with STATA 9. Weight for- age z-scores (waz) were calculated using Epi info 3.5 and z scores −2 to −3 were categorized as moderate and below −3 as severe malnutrition, as defined by WHO . Children were classified into immunologic categories based on CD4 count or CD4% by the 1994 CDC revised classification guidelines .
Missed opportunities were calculated by looking at the median number of visits a child had to the clinic in a year and multiplying that by the number of children with incomplete immunizations . Univariate logistic regression was used to analyze associations with a child’s immunization status, the outcome variable. Predictor variables included, maternal education, maternal HIV status, maternal death, paternal death, paternal education, paternal HIV status, child’s age, and presence of other children in house, child’s nutritional status, and child’s immunologic category. Factors that were found to be significant in univariate analysis with p < 0.05 were included in a multivariate logistic regression model.
Two hundred and seven interviews were conducted with individual children and their caretakers affected by HIV. One child was excluded because the child was HIV negative. Of the 206 remaining children, 197/206 were HIV positive, nine were HIV exposed but indeterminate.
The median age of children was 6 y, 113/206 were male, 35 (17 %) were on ART. Information on CDC immune status was available for 187 children. Eighty four were categorized as normal, 83 had moderate immune suppression and 20 were severely immune-suppressed. Malnutrition was significant. The median weight for age z-score was −2.07.
Information on maternal education was available for 144 mothers. The median years of schooling for mothers was 6 y, (inter-quartile range 5–8 y: range 0–15 y). Ninety two percent of the mothers were HIV infected. Thirty eight children did not have mothers and seventy did not have fathers. Nineteen children lived in an institution. Fifty (25 %) children had incomplete immunizations. There were ten children between the ages of 1–2 y and three of them had incomplete immunizations. Ninety three percent of children had gotten immunizations from a government clinic. Ninety eight percent of children had received BCG, three had not. All children had gotten oral polio vaccine. Thirty four of 206 children were missing DTP3 and 34/206 children were missing measles immunization. Thirty of 104 eligible children between the age of 5–10 y had missed the DT booster at age 5. Fourteen of 34 eligible children had not received the 10 y tetanus booster. The median number of visits to the clinic by children in a one year period was four. This represents 200 missed opportunities for immunizations.
Risk factors for incomplete immunization in children with HIV infection
% incompletely immunized
Unadjusted OR for incompletely immunized (95 % CI, p)
Adjusted OR for incompletely immunized (95 % CI, p)
Median (IQR; range)
1.16 (1.05–1.28, p = 0.003)
1.11(0.96–1.28, p = 0.15)
1.12(0.59–2.12, p = 0.72)
Other children in house
1.08(1.04–1.11, p < 0.003)
0.98(0.82–1.18, p = 0.86)
No other children in house
CDC Immune Class
0.82(0.50–1.35, p = 0.432)
Malnutrition (Median waz)
1.03(0.66–1.59, p = 0.89)
2.98(1.38–6.44, p = 0.005)
4(1.19–13.4, p = 0.025)
Mom School (IQR, median; range)
0.84(0.75–0.96, p = 0.008)
0.86(0.76–0.97, p = 0.016)
0.26(0.13–0.56, p < 003)
0.45(0.12–1.64, p = 0.22)
Father School (IQR, median; range)
0.85(0.76–0.95, p = 0.005)
0.6(0.30–1.22, p = 0.158)
Paternal education was found to be statistically significant in univariate logistic regression but not in the multivariate model when included along with maternal education and ART. Paternal education and maternal education were found to be highly correlated (Spearman’s correlation coefficient = 0.42, p value = 0.000) – and hence paternal education was not included in the model to avoid issues of multi-collinearity.
Maternal education significantly protected a child from being incompletely immunized, adjusted OR = 0.86 (95%CI--.76–.97, p = 0.016). Compared to women with six years of education, women who had zero years of education were three times as likely to have a child incompletely immunized or children with mothers with six years of education compared to those with no education were seven times more likely to have complete UIP vaccine status.
To control for whether the analysis was biased by children whose parents were deceased and living in an institution, and thus may have caretakers who are less aware of their immunization history, the analysis was repeated after eliminating these children (n = 19). There was no change in the analysis. Maternal education remained a strong protective factor in keeping a child from being incompletely immunized OR = 0.86 (0.76–0.98, p = 0.027) and those on ART remained at risk for incomplete immunization OR = 3.9 (1.15–13.3, p = 0.029).
Information on the impact of HIV infection on routine childhood immunization from India is required. This study is the first to look at the issue of immunization coverage in a cohort of HIV infected children from West Bengal regularly accessing care. The authors found that the overwhelming numbers of children were dependent on government services for immunizations and 25 % had incomplete immunizations. Including immunizations in programs for HIV infected children could potentially have prevented 200 missed opportunities for immunizations in this cohort.
The issue of the impact of maternal HIV and pediatric HIV status on completion of childhood immunizations has been studied in sub-Saharan Africa specifically in Uganda, Zambia, and South Africa. Setse et al. looked at 473 children hospitalized with measles in South Africa from 1998 to 2000 and looked at the issue of whether HIV status affected immunization completion. In this cohort of South African children between the ages of 4–60 mo they found that 32 % of HIV infected children were incompletely immunized compared to 21 % of HIV uninfected children . In a community based study from the Rakai district of Uganda of 6–35 mo old children the authors found that immunization rates among children of HIV infected mothers was lower 21 % compared to 27 % among children of HIV uninfected mothers . In South Africa the authors looked at over 2,400 children 12–23 mo of age from 2005 to 2006 and found that children of mothers with HIV infection had up to a 40 % less chance of being completely immunized .
The present study has limitations. The authors did not have an HIV negative cohort to compare, investigate whether HIV infection itself is an independent risk factor for incomplete childhood immunizations.
A recent analysis of district level immunization data in West Bengal, by Som et al., shows that 54 % of children 12–35 mo of age are fully immunized with rates in Kolkata at about 47.4 % . The immunization rates in this chronically ill, disadvantaged pediatric population is higher than what has been reported for the region. This is encouraging. Perhaps what it really shows is that vaccines if made available are valued by these families who may be accessing health care more frequently than the average family.
The major determinant of complete immunization found by Som et al. was maternal education. Similarly, in this HIV positive cohort, maternal education remains the major risk factor for incomplete immunization. Others have shown that globally maternal education may in fact be a proxy for socioeconomic status and is more strongly associated with complete immunization status of a child than the household income .
Maternal education has been shown as protective against incomplete immunizations in children globally [18–20]. Studies from Uganda showed that infants of mothers with some secondary education were 50 % less likely to have missed immunizations  and a recent study from Nigeria found that mothers with some post-secondary education were more than twice as likely to have children with complete immunizations . Analysis of data from the National Family Health Survey from India shows that mothers, particularly adolescent mothers with some middle school education are 3.2 times more likely to have completely immunized their children . Studies from different regions of India, looking at different socioeconomic strata have shown that maternal education is a key determinant for complete childhood immunizations in the country [23, 24].
Interventions discussing the importance of immunization targeted at women with low formal education can be very effective. In a recent randomized controlled trial, communities in Pakistan that were randomized to an intervention that involved a 3 min conversation in simple Urdu targeting mothers, about the importance of immunization showed complete immunization rates of over 60 % compared to control communities rates of 39 % .
The combination of maternal education and HIV status on childhood immunization has not been studied extensively in India. In families affected by HIV maternal HIV status has been shown to be significantly linked to a child’s immunization status . Targeting information about the importance of childhood immunization for HIV affected women accessing Volunteer Counseling and Testing Center (VCTC) counseling in Uganda has been suggested, a similar approach may be effective in the Indian context . The information from this cohort suggests that interventions and resources for childhood immunizations need to target HIV infected women with low formal education.
The authors found that children on ART had high rates of incomplete immunizations and there seemed to be a higher risk in those children on ART whose mothers had only a primary education (Fig. 1). These children were accessing care regularly. This suggests that there may be a gap in understanding. Perhaps women who are bringing their children in for ART are not aware of the need for immunizations as well, or perhaps having to access different health services such as immunizations and HIV care at multiple different locations is logistically overwhelming for mothers who may themselves be suffering from a chronic debilitating medical condition. These issues warrant exploration.
Developing comprehensive programs for HIV infected children that includes immunizations is important in the Indian context . Given the increased burden of vaccine preventable disease in this group, many including the Indian Academy of Pediatrics have called for including additional vaccines such as the Hib, Pneumococcal conjugate vaccines and Hepatitis b vaccines into programs for HIV infected children . Also including provisions for revaccinating children on highly active antiretroviral therapy (HAART) who have immune-reconstitution is an important area of discussion . Studies from Zambia arguing for the revaccination of HIV infected children on ART have suggested that HIV infected children may represent a cohort of “measles susceptible” children that could affect the UN millennium development goal of a 90 % reduction in measles mortality by 2015 .
Unlike in HIV infected adults, immune reconstitution in HIV infected children involves naïve T cells and not memory T cells . As a child’s immune system is affected by the virus they may lose immunity to vaccine preventable pathogens. Once ART is initiated they develop immune reconstitution, however ART by itself may not by restore immunity to pathogens for which vaccines were administered prior to the initiation of ART. This may leave HIV infected children with immune reconstitution at risk for vaccine preventable pathogens.
Recent studies looking at the measles vaccination as a model to understand this issue show that children on highly active antiretroviral therapy (HAART) who achieve normal CD4 counts are able to seroconvert after revaccination with the measles vaccine at the same rate as is expected in uninfected 9-mo-old infants on first vaccination with measles . Developing comprehensive programs for HIV infected children that includes immunizations, is important to address both the HIV infected child and to prevent mortality from vaccine preventable diseases at the population level.
Integrating maternal and child health services into immunization programs is part of the WHO and UNICEF global vision strategy 2006–2015. HIV counseling and treatment services have been integrated with immunization services in Africa . With the scale-up of NACP-III significant numbers of HIV infected children are benefiting from ART. Including immunizations into these programs for HIV infected children are an important next step.
HIV infected children are at risk for incomplete immunization coverage though they regularly access medical care. Including routine immunizations, particularly catch-up immunizations and revaccination for children with immune reconstitution in programs for HIV infected children maybe an effective way of protecting these children from vaccine preventable disease.
Conflict of Interest
Role of Funding Source
An institutional grant from IIT Kharagpur was received.