Introduction: Reliable information on rates of up-to-date coverage and timely administration of routine childhood immunizations is critical for guiding public health efforts worldwide, yet prospective observation of vaccination programs within individual communities is rare. Here we provide a longitudinal analysis of the directly-observed administration of a 3-dose primary vaccination series to infants in a low-resource community in Lusaka, Zambia.

Methods: Throughout 2015, we recruited a longitudinal birth cohort of mother/infant pairs (initial enrollment, 1,981 pairs; attending, 1,497 pairs) from the peri-urban informal settlement of Chawama compound, located in Lusaka, Zambia. We prospectively monitored the administration of scheduled Diphtheria-Tetanus-Pertussis (DTP) vaccinations across the first 14-18 weeks of life. We analyzed study attendance and vaccine coverage, both overall and stratified by age group. We employed Kaplan-Meier analyses to estimate delays in age-appropriate administration of vaccine doses. We also assessed schedule timing violations, including early and compressed dose administration.

Results: At study completion, first dose (DTP1) rates were high (92.9% of attending), whereas third dose completion (DTP3) rates were far lower (61.9%). Missed vaccinations and study dropout both contributed to the low DTP3 completion rates. DTP1 was administered very late (at or after 10 weeks) to 61 infants (4.1%). DPT1 was administered too early to 64 infants (4.3%), and 77 (5.1%) received consecutive doses below the minimum recommended spacing of 28 days.

Conclusions: We observe substantial individual variation in the timing of early childhood DTP doses, though following this birth cohort proved challenging. Our results indicate that timely administration of both DTP1 and DPT3 remains a challenge in this community. These directly-observed, individual-based results provide an important counterpoint to more course-grained, survey-based national and province estimates of up-to-date vaccine coverage. This study also highlights the challenges of vaccine hesitancy and sub-optimal utilization of (no-cost) healthcare services in a low-resource urban setting.

See manuscipt Methods for enrollment and cohort details.

Each individual was assigned a random offset of +/- 3 days, and all study dates were adjusted by this offset so that the dataset contains no personally identifiable information.