Immunization Schedule

The local community of approximately 200 000 people is of low socio-economic status, live in informal housing or crowded conditions and have high levels of unemployment. Infectious diseases including pneumonia, HIV (antenatal prevalence approximately 30%) and tuberculosis (annual reported incidence 293/100 000) are common. There is a high prevalence of tobacco smoke exposure, alcohol misuse, malnutrition and other poverty-related exposures. Pneumonia is the predominant cause of childhood hospitalisation and death, with the estimated incidence similar to the reported LMIC incidence of 0.22 per child-year in early life.2 (link) The population is stable, with little immigration or emigration. More than 90% of the population access healthcare in the public sector including antenatal and child health services. The public health system comprises 23 primary health clinics and one hospital, Paarl Hospital, where all births and hospital care occur. The well-established, free primary healthcare system provides childhood immunisations including 13-valent pneumococcal and H influenzae b vaccines as part of the national immunisation schedule.
Consenting pregnant women are enrolled from two primary health clinics serving different populations—TC Newman (serving a mixed race population) and Mbekweni (serving a black African population). Pregnant women who are not enrolled are included in a control cohort; these mother–infant pairs are followed annually to compare outcomes with the active cohort.

The ZOster ecoNomic Analysis (ZONA) model (Fig. 5) was developed in MS Excel. It is a static multi-cohort Markov model. Cohorts are split into 5 age groups for people aged 50+ years (i.e., 50–59, 60–64, 65–69, 70–79, 80+). If the “50+ years combined” option is selected, the model assumes that all of the subjects in the 50–59, 60–64, 65–69, 70–79 and 80+ age cohorts are vaccinated, as in a one-off ‘catch-up’ campaign, at an age of 50, 60, 65, 70 and 80 y respectively. The model follows all subjects within a cohort over their remaining life-time from the year of vaccination with annual cycle lengths. As such all subjects remain in their initial cohort and all subsequent events are counted in that cohort only. Three different HZ vaccination strategies are compared; no vaccination (control), vaccination with ZVL, and vaccination with HZ/su. Within each vaccine arm/strategy, individuals can be fully compliant with the vaccine dosing schedule, partially compliant or not vaccinated at all, depending on the corresponding vaccine coverage and compliance rates assumed. An overview of the model structure is presented in Fig. 5. Transition probabilities between the health states HZ, natural death, HZ related deaths, recover, recurrent HZ,.. occur using an annual time step. PHN and NON-PHN complications are health states which occur within a HZ episode and as such occur within this annual time step. Probabilities of moving between health states are derived from Germany specific literature and are age-group specific.
Figure 5.

Schematic overview of Markov structure – ZONA model. HZ: herpes zoster; PHN: postherpetic neuralgia.

In this analysis 3 age cohorts were considered, i.e., 50–59, 60–69 and 70+ years, i.e., combining results from the 60–64 and 65–69 cohorts and the 70–79 and 80+ cohorts, respectively, for presentation purposes. The age cohorts were selected to capture age-dependent differences in disease incidence, complications, outcomes, costs and potential public health decision making.

The 2013–2014 DHS, a nationally representative survey based on a stratified 2-stage cluster design, occurred from November 2013 to February 2014. The first stage of the survey consisted of Enumeration Area formation in which a stratified sample of geographic locations, or clusters (n = 540), was selected with proportional probability according to size. The second stage involved sampling households from each Enumeration Area; complete listings of households were created within each cluster, and households (n = 9000) were selected with equal probability.^{14 –17} Within the selected households, individuals interviewed included 18,827 women 15–49 years old (all selected households) and 8656 men 15–59 years old (50% of selected households). The DHS collected biomarker data only on children from households in which men were interviewed. In our dataset, there were 812 children 9–59 months old with written (and dated) record of receiving the full 3-dose tetanus vaccine series and who also had anthropometric, measles and tetanus IgG serology data available, and of these, 713 children had data on measles that occurred following DTwPHibHepB vaccination as well as all other covariates of interest. Four of these children who met measles case criteria had had measles less than 2 months before dried blood spot (DBS) collection and so were removed from analyses to prevent associations due to acute measles-induced immune suppression in this study, bringing the final total to 711 children (this total incorporates complex survey weighting methods utilized by the 2013–2014 DRC DHS).
Information was collected on weight, health outcomes, vaccination history and vaccine-preventable disease serology. After parental consent, DBSs were collected from participants to assess immunity to vaccine-preventable diseases and processed at the University of California, Los Angeles (UCLA)—DRC laboratory at the National Laboratory for Biomedical Research in Kinshasa. All survey data were transferred from paper questionnaires to an electronic format using the Census and Survey Processing System (U.S. Census Bureau, ICF Macro). Data were double entered and verified by comparison of both datasets.
At the time of interview, if mothers possessed a health care worker-provided vaccination card indicating the date the child was vaccinated, this was considered a “dated card” report. Since “dated card” reports are considered most reliable, we utilized only this type of report when categorizing a child as vaccinated, whether the vaccination was against tetanus or measles. Children classified as “unvaccinated” in this study were those reported as such in the 2013–2014 DHS. Tetanus vaccination status was obtained via dated vaccination card and limited to children receiving the complete 3-dose vaccination series (all other children were removed from analyses). Due to concern of live versus killed vaccines resulting in nonspecific effects on vaccinees,^{18 (link)} we examined whether the DTwPHibHepB versus measles vaccine was given last within the sample of participating children. Of the 649 children who received both DTwPHibHepB and measles vaccines, 5 (0.1%) children received DTwPHibHepB after the measles vaccine (indicating an incorrect vaccination schedule), 22 (3.3%) children received both vaccines within the same month and 622 (96%) received the measles vaccine after DTwPHibHepB. We did not examine other vaccines a child may have received.