Analysis of malaria surveillance data in Ethiopia: what can be learned from the Integrated Disease Surveillance and Response System?

BACKGROUND: Routine malaria surveillance data is useful for assessing incidence and trends over time, and in stratification for targeting of malaria control. The reporting completeness and potential bias of such data needs assessment. METHODS: Data on 17 malaria indicators were extracted from the Integrated Disease Surveillance and Response System database for July 2004 to June 2009 (Ethiopian calendar reporting years 1997 to 2001). Reporting units were standardized over time with 2007 census populations. The data were analysed to show reporting completeness, variation in risk by reporting unit, and incidence trends for malaria indicators. RESULTS: Reporting completeness, estimated as product of unit-month and health facility reporting, was over 80% until 2009, when it fell to 56% during a period of reorganization in the Ministry of Health. Nationally the average estimated annual incidence of reported total malaria for the calendar years 2005 to 2008 was 23.4 per 1000 persons, and of confirmed malaria was 7.6 per 1,000, with no clear decline in out-patient cases over the time period. Reported malaria in-patient admissions and deaths (averaging 6.4 per 10,000 and 2.3 per 100,000 per year respectively) declined threefold between 2005 and 2009, as did admissions and deaths reported as malaria with severe anaemia. Only 8 of 86 reporting units had average annual estimated incidence of confirmed malaria above 20 per 1,000 persons, while 26 units were consistently below five reported cases per 1,000 persons per year. CONCLUSION: The Integrated Disease Surveillance and Response System functioned well over the time period mid 2004 to the end of 2008. The data suggest that the scale up of interventions has had considerable impact on malaria in-patient cases and mortality, as reported from health centres and hospitals. These trends must be regarded as relative (over space and time) rather than absolute. The data can be used to stratify areas for improved targeting of control efforts to steadily reduce incidence. They also provide a baseline of incidence estimates against which to gauge future progress towards elimination. Inclusion of climate information over this time period and extension of the dataset to more years is needed to clarify the impact of control measures compared to natural cycles on malaria.

Malaria prevalence and mosquito net coverage in Oromia and SNNPR regions of Ethiopia.

BACKGROUND: Malaria transmission in Ethiopia is unstable and seasonal, with the majority of the country's population living in malaria-prone areas. Results from DHS 2005 indicate that the coverage of key malaria interventions was low. The government of Ethiopia has set the national goal of full population coverage with a mean of 2 long-lasting insecticidal nets (LLINs) per household through distribution of about 20 million LLIN by the end of 2007. The aim of this study was to generate baseline information on malaria parasite prevalence and coverage of key malaria control interventions in Oromia and SNNPR and to relate the prevalence survey findings to routine surveillance data just before further mass distribution of LLINs. METHODS: A 64 cluster malaria survey was conducted in January 2007 using a multi-stage cluster random sampling design. Using Malaria Indicator Survey Household Questionnaire modified for the local conditions as well as peripheral blood microscopy and rapid diagnostic tests, the survey assessed net ownership and use and malaria parasite prevalence in Oromia and SNNPR regions of Ethiopia. Routine surveillance data on malaria for the survey time period was obtained for comparison with prevalence survey results. RESULTS: Overall, 47.5% (95% confidence interval (CI) 33.5-61.9%) of households had at least one net, and 35.1% (95% CI 23.1-49.4%) had at least one LLIN. There was no difference in net ownership or net utilization between the regions. Malaria parasite prevalence was 2.4% (95% CI 1.6-3.5%) overall, but differed markedly between the two regions: Oromia, 0.9% (95% CI 0.5-1.6); SNNPR, 5.4% (95% CI 3.4-8.5), p < 0.001. This difference between the two regions was also reflected in the routine surveillance data. CONCLUSION: Household net ownership exhibited nearly ten-fold increase compared to the results of Demographic and Health Survey 2005 when fewer than 5% of households in these two regions owned any nets. The results of the survey as well as the routine surveillance data demonstrated that malaria continues to be a significant public health challenge in these regions-and more prevalent in SNNPR than in Oromia.