Background: Sub-microscopic and asymptomatic infections could be bottlenecks to malaria elimination efforts in Ethiopia. This study determined the prevalence of malaria, and individual and household-level factors associated with Plasmodium infections obtained following detection of index cases in health facilities in Jimma Zone. Methods: Index malaria cases were passively detected and tracked in health facilities from June to November 2016. Moreover, family members of the index houses and neighbours located within approximately 200 m from the index houses were also screened for malaria. Results: A total of 39 index cases initiated the reactive case detection of 726 individuals in 116 households. Overall, the prevalence of malaria using microscopy and PCR was 4.0% and 8.96%, respectively. Seventeen (43.6%) of the index cases were from Doyo Yaya kebele, where parasite prevalence was higher. The majority of the malaria cases (90.74%) were asymptomatic. Fever (AOR = 12.68, 95% CI 3.34-48.18) and history of malaria in the preceding 1 year (AOR = 3.62, 95% CI 1.77-7.38) were significant individual-level factors associated with detection of Plasmodium infection. Moreover, living in index house (AOR = 2.22, 95% CI 1.16-4.27), house with eave (AOR = 2.28, 95% CI 1.14-4.55), area of residence (AOR = 6.81, 95% CI 2.49-18.63) and family size (AOR = 3.35, 95% CI 1.53-7.33) were main household-level predictors for residual malaria transmission. Conclusion: The number of index cases per kebele may enhance RACD efforts to detect additional malaria cases in low transmission settings. Asymptomatic and sub-microscopic infections were high in the study area, which need new or improved surveillance tools for malaria elimination efforts.
The study was conducted in catchment kebeles (smallest government administrative units in Ethiopia) of Kishe and Nada health centres, located in Shebe Sambo and Omo Nada districts of Jimma Zone, respectively (Fig. 1). Shebe Sambo and Omo Nada districts are located at 415 and 285 kms south west of the capital, Addis Ababa, respectively. The geographical coordinates of Shebe Sambo and Omo Nada are approximately 7°30′14″N, 36°30′44″E and 7°38′00″N, 37°15′05″E, respectively. The inhabitants in both areas mainly depend on subsistence farming, cultivating mainly maize and teff. Moreover, in Kishe area rice is cultivated in small scale. Distribution of the index houses and neighbours in the study area Historically, the catchment areas of both health centres have been malarious [23–25]. As in most parts of Ethiopia, the transmission of malaria in these areas is seasonal. The transmission usually peaks from September to October, following the major rains from June to September, and minor transmission occurs in April and May, following the short rains of February to March. According to the information obtained from both health centres, malaria cases detected in the health facilities have remarkably declined in recent years. A total 43 malaria cases have been registered in Kishe Health Centre in 2016. Of these, 62.8% were due to P. falciparum. Kishe Health Centre has been serving a total of 26,843 population in 2016. In the same year, a total of 51 malaria cases were recorded at Nada Health Centre, 49% of which were due to P. falciparum. The health centre has been serving a total of 32,264 population in 2016. A prospective observational study was conducted for 6 months (June to November 2016) in two health centres and their catchment kebeles. Index malaria cases residing within the catchment kebeles of the two health centres who did not travel within 2 weeks prior to presenting to the health centres, and diagnosed with malaria at the health centres during the study period were included in the study. The index cases were identified in the health centres based on the routine blood film microscopy by the laboratory staff at each health facility. Following detection of the index cases, household members of the index houses and neighbours within 200 m radius were included in the study. Febrile patients who sought treatment at Kishe and Nada Health Centres from June 1 to November 31, 2016, were screened for malaria using microscopy by the resident laboratory personnel as a routine practice. Consenting index cases that were microscopy-positive and who agreed that a research team will visit them within 1 week provided their home address. To locate the index household easily, names of three nearby household heads were also recorded. Presence of index cases was communicated to the research team on the same day of presentation to the health centres. The index houses and neighbours were visited within 1 week of detecting the index cases, in most cases, within 3 days. After obtaining consent, family members of the index houses and neighbours within 200 m radius from the index houses were screened for Plasmodium infection using RDT. Moreover, demographic information and some individual and household-level risk factors of malaria were collected using a semi-structured questionnaire. The field data was collected by experienced laboratory technologists. The individual-level factors assessed included demographic characteristics (age, sex, educational status and occupation), recent travel history, history of malaria infection in the last 1 year, LLIN usage the previous night before the survey and axillary body temperature. Fever was defined in this study as having axillary temperature of ≥ 37.5 °C, which was measured during the survey. The household-level factors assessed included housing conditions such as roof structure, presence of visible hole on the wall, presence of eave and presence of window(s), and presence of animals within the house, family size, total number of LLINs owned during the survey and whether the house was sprayed with insecticide during the last 1 year. In Ethiopia, indoor residual spraying is performed once in a year, usually around July to September (before malaria cases peak) as transmission is mainly seasonal. Apart from the household-level characteristics, an approximate distance of each house from the index house was estimated, and coordinates of each house was taken using hand-held global positioning system unit (GPS). All consenting members of the index houses and neighbours were enrolled in this study. Infants less than 6 months of age were not included, with the assumption that they are likely protected from malaria due to passively acquired maternal antibody, and presence of foetal haemoglobin. The household members available during the visit and those found in the nearby farming sites were included in the study. The samples were collected by a team of two laboratory technologists deployed at a time. The data collection team spent, in most cases, all the day around the index houses to maximize coverage of the screened population. However, few individuals who were not available on the day of screening around their houses were not captured, as there was no follow-up to the community members. Finger-prick blood samples were collected from consenting study participants for blood examination by RDT and microscopy. The RDT was done for rapid diagnosis and treatment; hence, RDT-positive individuals were referred to the health centres for confirmation and treatment. Moreover, thick and thin blood smears were prepared for each study participant in the field. After air drying, the thin films were fixed with absolute methanol, and Giemsa stained at Jimma University Medical Parasitology Laboratory on the same day of collection. The slides were examined by two experienced laboratory technologists independently. Approximately 200 high-power fields of the thick blood smears were examined before declaring a microscopy-negative result. The personnel reading the slides were also blinded of the RDT results. Apart from the blood smears, three to four drops of blood were spotted on Whatman 3MM filter paper for further molecular analysis. The blood spots were air-dried and kept individually in air-tight plastic bags and stored at − 20 °C until DNA extraction. DNA was extracted from approximately 20 µL whole blood using the QiaCube DNA extraction system as per standard protocols. DNA was eluted in 100 µL buffer, and 4 µL DNA was screened using a multiplex P. falciparum/P. vivax qPCR with limit of detection of 1 DNA copy per reaction [26]. Thus, the limit of detection of the qPCR was 1–2 parasites/µL blood. Household was defined in this study as group of human subjects residing in the same house as family members. Household access to LLINs was considered “sufficient” when the ratio of the total LLINs owned by a household to the family members is at least 0.5 (assuming that one LLIN covers two individuals), and “not sufficient” when the ratio is less than 0.5 [27]. While a total of 726 individuals participated in this study providing samples for microscopic examination, sufficient DBS sample for PCR was obtained from 603 individuals. Specimens of sufficient quantity could not be obtained from the remaining individuals for PCR following blood film preparation and RDT testing, the results of which were used for immediate care. Hence, the data analysis was based on the PCR-run samples. Asymptomatic malaria infection was considered when an individual who did not experience fever at the time of the survey (axillary body temperature is less than 37.5 °C) and no malaria-related symptoms was positive for Plasmodium species by PCR. The collected data were coded, entered into Excel (Microsoft Office 2010) and cleaned. The data were analysed using a statistical software package STATA 12 (StataCorp., TX, USA). Descriptive statistics including frequency, percentages and median were calculated to summarize demographic profile of the study participants. Univariate and multivariate logistic regressions were employed to determine individual and household-level factors associated with malaria infection. Multi-level regression model (mixed-effects logistic regression) was utilized to determine predictors of malaria infection among individual and household-level variables. Variables with significant association with malaria infection by the univariate analysis and those with p values less than 0.2 were candidates for the multivariate analyses. Odds ratio and the corresponding 95% confidence intervals were calculated to show the strength of the association. Statistical significance was set at p < 0.05 during the analysis.
N/A
DIMA AI Care
