Menu
Menu
Menu
Menu
Diphtheria-tetanus-pertussis (DTP) vaccine may be associated with excess female deaths. There are few studies of possible nonspecific effects of the DTP-containing vaccine Penta (DTP-hepatitis B-Haemophilus influenzae type b). We therefore investigated whether Penta vaccinations were associated with excess female deaths in rural Bangladesh. Between June 29, 2011 and April 20, 2016, we examined the mortality rates of 7644 children followed between 6 weeks and 9 months of age. We analyzed mortality using crude mortality rate ratio (MRR) and age-adjusted MRR (aMRR) from a Cox proportional hazards model. Mortality was analyzed according to sex, number of doses of Penta, and the order in which BCG and Penta were administered. During follow-up, 43 children died. For children who were only BCG vaccinated (BCG-only), the adjusted F/M MRR was 0.47 (0.09–2.48). However, among children who had Penta as their most recent vaccination, the adjusted F/M MRR was 9.91 (1.16–84.44). Hence, the adjusted F/M MRR differed significantly for BCG-only and for Penta as the most recent administered vaccination. Although the mortality rate was low in rural Bangladesh, there was a marked difference between adjusted F/M MRR’s for children vaccinated with BCG-only compared with children where Penta was the most recent administered vaccination. Although usually ascribed to differential treatment and access to care, DTP-containing vaccines may be part of the explanation for the excessive female mortality reported in some regions.
International Center for Diarrheal Diseases Research, Bangladesh (icddr,b) runs a Health and Demographic Surveillance System (HDSS) in Chakaria. The HDSS covers a population of 90,000 individuals living in 16,000 households in 49 villages. All households are visited every three months to enquire about marriages, pregnancies, births, migrations, and deaths. The expanded program on immunization (EPI) provides services in the HDSS areas through 95 EPI outreach centers. Vaccines are given as recommended by WHO: BCG at birth, Penta and Oral polio vaccine (OPV) in three doses at 6, 10, and 14 weeks of age, followed by MV at 9 months of age. Information on vital status and vaccinations of children below 3 years of age was collected during household visits every 3 months. Lower respiratory infection (LRI) and diarrhea are the leading causes of morbidity and under-5 mortality for children in the Chakaria HDSS area.18,19 BCG is protective against disseminated TB in childhood and the single shot of Penta provides protection to children from five life-threatening diseases: diphtheria, pertussis (whooping cough), tetanus, hepatitis B, and haemophilus influenzae Type b (Hib). The incidence of these vaccine preventable infections is low among children below five years of age in Chakaria. During 2010–2012, only 2 pertussis, 8 meningitis, and 2 measles infection related deaths were reported among children less than five years of age. Hence, the mortality rates per 100,000 person-years (MR) were 7 for pertussis, 29 for meningitis, and 7 for measles infection among children less than five years of age.19 It should be noted that studies have reported that females are neglected in seeking medical care among under-five children,20,21 even for free immunization22. The study was conducted between June 29, 2011 and April 20, 2016. Children were categorized according to the sequence in which they received BCG and Penta vaccines. We defined 11 mutually exclusive vaccine-sequence categories (Group I–XI) of children’s hitherto administered BCG and Penta vaccinations. Vaccination sequences could be initiated in three ways: (1) the WHO recommended schedule: BCG-first and then three doses of Penta denoted Penta1-3 (Group I–IV); (2) BCG+Penta1 administered simultaneously first and then Penta2-3 (Group V–VII); and (3) Penta1-first followed by Penta2-3 with BCG administered together with Penta2 or Penta3 (Group VIII–XI). In addition, we defined the two categories “documented no vaccination” (unvaccinated) and “have card but not seen” when the family reported possession of vaccination card but was unable to show card to the interviewer. We furthermore focused on three broader, still mutually exclusive groups: BCG-only (Group I), Penta-vaccinated, i.e., Penta as the most recent administered vaccination (Group II–IV and VI–X), and “documented no vaccination.” Children who had only received “BCG and Penta1 simultaneously” (Group V) and the children who received “Penta1 followed by BCG” (Group-XI) were not included in the Penta-vaccinated group. This was because receiving BCG and DTP simultaneously has been associated with decreased mortality.2,12,16 We also compared the mortality between children who started vaccination according to the WHO recommended schedule ((BCG-first-then-Penta) (Group I–IV)) and children who started their vaccination schedule with BCG and Penta administered simultaneously (Group V–VII). The pneumococcal vaccine (PCV) and inactivated polio vaccine (IPV) were introduced on March 21, 2015 in Chakaria and the IPV was administered with PCV3. Therefore, we restricted our study to children who were born before February 9, 2015. Information on vaccinations and deaths was collected retrospectively through trimonthly visits to each household. Vaccinations of dead children are likely to be underreported because parents tend not to keep the vaccine card of a deceased child; even if they do keep the card, they may be reluctant to show the card to the interviewer. In such situations, using retrospectively updated information likely misclassifies some vaccinated deceased children as unvaccinated, resulting in excess deaths in the unvaccinated group. Furthermore, vaccinated children who survive will contribute to a time-to-event analysis with risk-free survival time because they survived up to the next visit. Hence, using the retrospective updating approach to measure the effect of vaccination can introduce survival bias23–25 which can distort the estimated effect of vaccinations. We therefore used a “landmark approach”23–25 to measure the impact of vaccination status on mortality. Depending on the specific analysis reported, we define vaccination status as either the vaccination sequence administered hitherto or the most recent administered vaccination. In the landmark approach, the vaccination status is fixed from the household visit until the date of the next visit when the vaccination card is seen again. Vaccination status changes with visit; hence, children could, e.g., contribute with unvaccinated risk-time after the first visit and vaccinated risk-time after the second visit. We limited the study to children aged 6 weeks (Penta1 is scheduled at 6 weeks) to 9 months of age (MV is scheduled at 9 months of age). Children who were visited only once were excluded from the present study (N = 144; Figure 1) because we could not follow these children for mortality between 6 weeks and 9 months of age. Children with no information about vaccination status at any visit were also excluded (N = 30). Furthermore, we excluded children who were not born in the HDSS area (N = 110) because mothers could not accurately report date of birth and date of vaccinations. Flowchart of study population We used Cox proportional hazards models with age as the underlying time-scale to estimate adjusted MRRs (aMRR) of different hitherto administered vaccination sequences for both boys and girls. Children were followed from 6 weeks of age until migration, death, or 9 months of age, whichever occurred first. We included parity, maternal age, maternal education, season of birth, and village as potential confounders in the adjusted models. To assess the sex-differential effects of different sequences of BCG and Penta vaccinations on mortality, an interaction between sex and hitherto administered sequence of BCG and Penta were included in the adjusted Cox proportional hazards models. To test the sex-differential effects according to whether Penta1, Penta2, or Penta3 were the most recent administered vaccination, we conducted a log-rank test.
N/A
