The implementation of malaria intermittent preventive trialtreatment with sulphadoxine-pyrimethamine in infants reduced all-cause mortality in the district of Kolokani, Mali: results from a cluster randomized control

It is estimated that about one million of people in sub-Saharan Africa died directly from malaria in 2008, of which 800,000 were children under five years [1]. The numbers are higher if indirect contribution of malaria is taken into account [2, 3]. In the absence of licensed vaccine against malaria, intermittent preventive treatment of malaria in infants (IPTi) consisting of the administration of a curative dose of sulphadoxine-pyrimethamine (SP) at the time of EPI vaccination (DTP2, DTP3 and measles vaccine) regardless of the presence of symptoms or infection was developed to reduce the burden of malaria. Randomized control studies have shown that the strategy reduced the incidence of malaria by 30% and the incidence of anaemia by 20% [4]. The strategy is now recommended as policy for malaria control in infants in areas of moderate or high malaria transmission and where the resistance to SP is low [5]. However, there was no published study on the impact of the strategy on the mortality. The sample size and the duration of follow-up of initial studies, which assessed the efficacy of the strategy on the clinical episodes, did not allow the assessment of its efficacy on mortality. The pilot implementation of IPTi in health services in Mali offered an opportunity to address this question. The objective of the present study was to assess the impact of IPTi implementation in health services in Mali on the mortality in the targeted population.

A total of 5,882 children that reached four months of age between 1 December, 2006 and 1 December, 2008 in the study area were surveyed, including 2,869 from 75 localities in the intervention zone and 3,013 from 75 villages in the non-intervention zone. The gender distribution was similar between the two zones (50.5% in the intervention zone and 51.6% in the non-intervention zone were male, p = 0.43). The exact date of death was unknown for two subjects (a boy in the intervention zone born on 15 February, 2007 and a girl born on 10 January 2008, in the non-intervention zone) who were excluded from the analysis.

Among the 5,880 children, 198 died at the age of four to 18 months and 119 died at the age of four to 12 months. Of the children who died at age of four to 18 months, 86 (2.86%) lived in the intervention and 112 (3.91%) lived in the non-intervention zone (Fischer exact test; two sided p-value p = 0.030). Of those who died at the age of four to 12 months, 51 (1.69%) lived in the intervention zone and 68 (2.37%) lived in the non-intervention zone (Fischer exact test; two sided p-value p = 0.078). The mortality rate, mortality rates ratios and reduction in mortality are presented in Table 1. In the cohort of four to 18 months of age, the mortality rate per 1,000 children was 2.532 in the intervention zone compared to 3.456 in the nonintervention zone, gender-adjusted mortality rates ratio (MRR) = 0.73 (95% CI 0.55-0.97, p = 0.029). In the cohort of the four to 12 months of age, mortality rates per 1,000 children were 2.223 in the intervention zone and 3.126 in the non-intervention zone, adjusted for gender of MRR = 0.71 (95% CI 0.49-1.02, p = 0.64). The curves of the Kaplan-Meier survival estimates of time to death in children from four months shows significant difference in survival between the two zones (p <0.001) (Figure 1)

This study estimated the effect of two years of IPTi implementation on mortality in children and has found a reduction in all-cause mortality by 27% and 29% in children four to 18 months of age and those of four to 12 months of age, respectively. Use of insecticide-impregnated bet nets (ITN) for malaria control in endemic areas was associated with a reduction of 17% in all-cause mortality [8]. The implantation of IPTi along with EPI resulted in a reduction in all-cause mortality greater than the reduction provided by ITN, a current recommended strategy for malaria control.

Demonstrating the effect of a malaria control strategy on malaria specific mortality has always been difficult due to the requirement of large sample size and factors that can confound the relationship between the intervention and the malaria-specific mortality. For example, the impact of the use of the ITN on malaria-specific mortality could not be demonstrated initially in individual, large, randomized control trials [9, 10], but only in the meta-analysis studies [8, 11]. Attributing a death to malaria is difficult as the current indirect methods, such as verbal autopsies, lack sensitivity and specificity [12]. The effect of a malaria intervention on all-cause mortality was reported in previous studies [9, 13, 19]. Although, it requires a larger number of subjects than malaria-specific mortality assessment, there is no issue of sensitivity or the specificity when assessing it. For these reasons, the objective of the present study was limited to the assessment of the impact of the IPTi implementation on all-cause mortality. Additional reason for focusing on this objective was the fact that the data were collected retrospectively and far from the events, which made it difficult to accurately assign the cause of the death (including when verbal autopsy methods were used).

In the present study, the reduction in all-cause mortality may be due to a combination of both the direct and indirect effect of the IPTi on malaria and the indirect effects on other causes of death. Previous evaluations have shown that the implementation of IPTi has resulted in a significant increase in coverage of EPI vaccines, an increase that was more marked in the intervention zone [7]. The higher increase in vaccine coverage in the intervention zone may have contributed to the increase in child survival in these areas. Previous studies have shown that Bacille Calmette-Guerin (BCG) vaccination and measles vaccination have strong beneficial effect on child survival [14‐16]. It is also known that a substantial part of the mortality due to malaria is not directly attributable to malaria (i.e., indirectly attributed to malaria) [17] and results of adequately executed malaria control or elimination programmes could exceed expectations due to decreased indirect malaria mortality [18‐20]. Malaria infections may alter the capacity to survive other affections for example through chronic anaemia and enhancement of the severity of other childhood diseases.

IPTi has now been adopted by the WHO as a policy for malaria control despite the lack of data on the impact of the strategy on mortality [5, 21, 22].

This study has two major strengths. First, it is a randomized controlled study, which ensures the comparability between the two arms. Second, the IPTi was implemented within the health care system, better reflecting the environment in which the strategy will be implemented. Although the data were collected by individuals not involved in the implementation and who were not aware if a locality was in the intervention or non-intervention zone, it is difficult to rule out the possibility of observer bias, since it would be easy to find out if a locality was in the intervention or non intervention zone. The mortality rate observed in the present study was lower than expected (40.8 per 1,000 in the non-intervention zone compared to 106 per 1,000 at the national level according to the national health survey) resulting in relatively lower power and wider confidence intervals. Reassuringly, the mortality rates in this study are consistent with those reported elsewhere in the West African region [14, 23, 24]. Another limitation of the study is the relatively shorter period of the intervention (two years). It is possible that the impact of the strategy on the mortality increases with the duration of the intervention (for example, reduction in the malaria parasite carriage leading to a reduction in the transmission). It is therefore important that the impact of the implementation of the strategy continue to be monitored over longer periods of time during the implementation of the strategy as policy.

The implementation of the IPTi has resulted in a 27% reduction of the all-cause mortality in children aged four to 18 months. The results of this study support the adoption of IPTi as malaria control strategy and its implementation as a tool for malaria control.