Integrated maternal and child health services in Mozambique: structural health system limitations overshadow its effect on follow-up of HIV-exposed infants

Background: The follow-up of HIV-exposed infants remains a public health challenge in many Sub-Saharan countries. Just as integrated antenatal and maternity services have contributed to improved care for HIV-positive pregnant women, so too could integrated care for mother and infant after birth improve follow-up of HIV-exposed infants. We present results of a study testing the viability of such integrated care, and its effects on follow-up of HIV-exposed infants, in Tete Province, Mozambique. Methods: Between April 2009 and September 2010, we conducted a mixed-method, intervention-control study in six rural public primary healthcare facilities, selected purposively for size and accessibility, with random allocation of three facilities each for intervention and control groups. The intervention consisted of a reorganization of services to provide one-stop, integrated care for mothers and their children under five years of age. We collected monthly routine facility statistics on prevention of mother-to-child HIV transmission (PMTCT), follow-up of HIV-exposed infants, and other mother and child health (MCH) activities for the six months before (January-June 2009) and 13 months after starting the intervention (July 2009-July 2010). Staff were interviewed at the start, after six months, and at the end of the study. Quantitative data were analysed using quasi-Poisson models for significant differences between the periods before and after intervention, between healthcare facilities in intervention and control groups, and for time trends. The coefficients for the effect of the period and the interaction effect of the intervention were calculated with their p-values. Thematic analysis of qualitative data was done manually. Results: One-stop, integrated care for mother and child was feasible in all participating healthcare facilities, and staff evaluated this service organisation positively. We observed in both study groups an improvement in follow-up of HIV-exposed infants (registration, follow-up visits, serological testing), but frequent absenteeism of staff and irregular supply of consumables interfered with healthcare facility performance for both intervention and control groups. Conclusions: Despite improvement in various aspects of the follow-up of HIV-exposed infants, we observed no improvement attributable to one-stop, integrated MCH care. Structural healthcare system limitations, such as staff absences and irregular supply of essential commodities, appear to overshadow its potential effects. Regular technical support and adequate basic working conditions are essential for improved performance in the follow-up of HIV-exposed infants in peripheral public healthcare facilities in Mozambique.