Impact assessment of malaria vector control using routine surveillance data in Zambia: implications for monitoring and evaluation.

BACKGROUND: Malaria vector control using long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS), with pyrethroids and DDT, to reduce malaria transmission has been expansively implemented in Zambia. The impact of these interventions on malaria morbidity and mortality has not previously been formally assessed at the population level in Zambia. METHODS: The impact of IRS (15 urban districts) and LLINs (15 rural districts) implementation on severe malaria cases, deaths and case fatality rates in children below the age of five years were compared. Zambian national Health Management Information System data from 2007 to 2008 were retrospectively analysed to assess the epidemiological impact of the two interventions using odds ratios to compare the pre-scaling up year 2007 with the scaling-up year 2008. RESULTS: Overall there were marked reductions in morbidity and mortality, with cases, deaths and case fatality rates (CFR) of severe malaria decreasing by 31%, 63% and 62%, respectively between 2007 and 2008. In urban districts with IRS introduction there was a significant reduction in mortality (Odds Ratio [OR] = 0.37, 95% CI = 0.31-0.43, P = 0.015), while the reduction in mortality in rural districts with LLINs implementation was not significant (OR = 0.83, 95% CI = 0.67-1.04, P = 0.666). A similar pattern was observed for case fatality rates with a significant reduction in urban districts implementing IRS (OR = 0.34, 95% CI = 0.33-0.36, P = 0.005), but not in rural districts implementing LLINs (OR = 0.96, 95% CI = 0.91-1.00, P = 0.913). No substantial difference was detected in overall reduction of malaria cases between districts implementing IRS and LLINs (P = 0.933). CONCLUSION: Routine surveillance data proved valuable for determining the temporal effects of malaria control with two strategies, IRS and LLINs on severe malaria disease in different types of Zambian districts. However, this analysis did not take into account the effect of artemisinin-based combination therapy (ACT), which were being scaled up countrywide in both rural and urban districts.