Fewer out-of-sequence vaccinations and reduction of child mortality in Northern Ghana Paul Welaga a,f,⇑ , Abraham Oduro a , Cornelius Debpuur a , Peter Aaby b , Henrik Ravn c , Andreas Andersen c , Fred Binka d , Abraham Hodgson e a Navrongo Health Research Centre, P.O. Box 114, Navrongo, Ghana b Bandim Health Project, Indepth Network, Apartado 861, Bissau, Guinea-Bissau c Bandim Health Project, Statens Serum Institut, 2300 Copenhagen, Denmark d University of Health and Allied Sciences, Ho, Ghana e Research and Development Division, Ghana Health Service, Accra, Ghana f OPEN, Institute of Clinical Research, University of Southern Denmark, Odense, Denmark article info Article history: Received 17 November 2016 Received in revised form 27 February 2017 Accepted 3 March 2017 Available online 22 March 2017 Keywords: DTP Measles vaccine Non-specific effects of vaccines Out of sequence vaccinations Mortality Heterologous immunity abstract Background: Studies suggest that diphtheria-tetanus-pertussis (DTP) vaccine administered simultane- ously with measles vaccine (MV) or DTP administered after MV are associated with higher child mortality than having MV-after-DTP3 as most recent vaccination. We tested this in Northern Ghana where the prevalence of such out-of-sequence vaccinations has declined. Methods: Using annual cohort data of children aged 12–23 months from 1996 to 2012 and Cox propor- tional hazards models, we assessed survival in relation to the most recent vaccination status within the next 12 months and until five years of age. We assessed whether mortality in children aged 12–59 months was higher when the most recent vaccine was non-live (DTP) rather than live (MV or OPV). Results: Out-of-sequence vaccinations with DTP-containing vaccines and MV declined from 86% in 1989 to 24% in 1996 and 0.7% in 2012. Between 1996 and 2012, 38 070 children had their vaccinations status assessed: the adjusted hazard ratio (HR) for out-of-sequence vaccinations (DTP >= MV) compared with the recommended sequence of MV-after-DTP3 was 1.42(1.06–1.90) during the first 12 months after assessment of vaccination status and 1.29(1.03–1.60) with follow-up to five years of age; the HR was 2.58(1.14–5.84) before OPV or MV campaigns and 1.37(1.02–1.85) after the campaigns. Conclusion: Out-of-sequence vaccinations with DTP and MV are associated with higher mortality than MV as most recent vaccination; the effect is unlikely to be due to confounding. Hence, the reduction in out-of-sequence vaccinations may have lowered child mortality. It is recommended not to give DTP with MV or DTP after MV. Ó 2017 Elsevier Ltd. All rights reserved. 1. Introduction WHO’s Expanded Program on Immunization (EPI) recommends whole cell diphtheria–tetanus–pertussis (DTP) to be given at 6, 10 and 14 weeks and measles vaccine (MV) at 9 months of age. How- ever, due to weak health systems, these schedules are often not fol- lowed; many children receive DTP simultaneously with MV or after MV, i.e. out-of-sequence vaccinations. Studies in low-income countries suggest that vaccines may have non-specific effects (NSEs) on child survival which are not explained by the prevention of the targeted infection [1–8]. These NSEs may have sex differential effects on mortality [9]. The NSEs may be due to epigenetic changes reprogramming innate immu- nity, as has been shown for BCG, and may change once the child gets a new vaccination [10,11]. WHO’s Strategic Advisory Group of Experts on Immunization (SAGE) recently reviewed the evidence for potential NSEs of BCG, DTP and MV and recommended further research into these effects [12,13]. Live vaccines reduce mortality more than expected [1–5] but unfortunately inactivated vaccines may enhance susceptibility to unrelated infections [6–8]. Vaccines interact with each other and other immuno-modulating health interventions. Changing the sequence of vaccination may therefore lead to positive or negative NSEs. Consistent with the differential effects of live and inactivated vaccines, African and Asian studies have suggested that out-of-sequence vaccinations may be associ- ated with higher mortality than having MV as the most recent http://dx.doi.org/10.1016/j.vaccine.2017.03.004 0264-410X/Ó 2017 Elsevier Ltd. All rights reserved. ⇑ Corresponding author at: Navrongo Health Research Centre, P.O. Box 114, Navrongo, Ghana. E-mail addresses: pwelaga@gmail.com, pwelaga@yahoo.com (P. Welaga). Vaccine 35 (2017) 2496–2503 Contents lists available at ScienceDirect Vaccine journal homepage: www.elsevier.com/locate/vaccine