Malaria management in children with fever in rural Sierra Leone

Sierra Leone is one of the highest malaria burden countries Background: in the world and was severely affected by the 2014-15 Ebola outbreak. As fever is a common symptom of both malaria and Ebola, it might have affected the management of fever in children. Among under-fives in Koinadugu district, Sierra Leone, we determined fever cases that had malaria diagnostic testing and treated with Artemisinin-based Combination Therapy (ACT) during pre-Ebola, intra-Ebola and post-Ebola periods. The study population included all children under five with fever Methods: who presented to 68 primary healthcare facilities in Koinadugu district. Malaria management was in line with national guidelines. All individuals presenting with fever should be subjected to a malaria diagnostic test, which may involve a Rapid Diagnostic Test (RDT) or microscopy. Only confirmed malaria cases should receive ACTs. The study spanned pre-Ebola (June 1, 2013 – April 30, 2014), intra-Ebola (June 1, 2014 – April 30, 2015) and post-Ebola (June 1, 2016 – April 30 ,2017) periods. Data were sourced directly from routine morbidity registers available at each health facility. In the 68 health facilities, fever cases increased from 43,245 Results: pre-Ebola to 74,367 post-Ebola (1.7-fold increase). Diagnosed malaria ranged between 66% and 75%. Only 47% of malaria cases were treated during Ebola. ACT use was 95% pre-Ebola, 99% intra-Ebola and dropped to 71% post-Ebola. Post-Ebola, an average of 40 (59%) facilities had monthly stock-outs of ACT (range 28-45). What has changed since the Ebola outbreak is the increased Conclusion: utilisation of services for malaria. However, ACT stockouts are of concern, and this requires attention in order to ensure compliance with national malaria treatment guidelines. 1 2 3 4

Creative Commons Attribution IGO , which permits copying, adaptation and distribution in any medium or format for any purpose, provided the original work is properly cited, a License link is provided to the license, and any changes made are indicated. Any such copying, adaptation and distribution must not in any way suggest that World Health Organisation endorses you or your use.

Introduction
A cross-sectional study conducted in 2017 1 involving 68 primary health facilities in Koinadugu district of rural Sierra Leone compared the management of children with fever for malaria for a period before, during and after the Ebola outbreak. There were two key findings. First, less than half of all confirmed malaria cases were treated for malaria during the Ebola outbreak. As fever is a common symptom of both malaria and Ebola, health workers may have "played safe" by simply referring such children to Ebola management sites 2 . Second, monthly utilization of malaria diagnostics closely matched the number of reported fever cases, implying that fever cases were being routinely subjected to malaria testing 3 .
Although the post-Ebola period was included in this evaluation, it was for a relatively short period (six months) which was probably too soon to gauge health system recovery. At the primary healthcare level in the same district and among children under five, we thus performed a new analysis with a longer post-Ebola period and compared these data for similar periods before and during the Ebola outbreak. Our specific objectives were to determine: a) numbers of reported fever cases and malaria tests done, and b) numbers treated for malaria with artemisinin combination treatment (ACT) and within 24 hours of fever onset

Methods
This was a cross-sectional study using routine program data. The setting has been previously described 1 . The study population included all children under five with fever who presented to 68 primary healthcare units. Malaria management was in line with national guidelines and has been described previously 4 . In brief, all individuals presenting with fever to any given health facility were subjected to a malaria diagnostic test, which may involve a Rapid Diagnostic Test (RDT) or microscopy. Only confirmed malaria cases should receive ACT.
The study spanned a pre-Ebola ( Data were sourced directly from routine morbidity registers available at each health facility. A data manager was responsible for data quality, and a dedicated data entry clerk performed data entry. We used EpiData software for data entry and analysis (version 4.1 for entry and version 2.2.2.182 for analysis; EpiData Association, Odense, Denmark).
Ethics approval was obtained from the Sierra Leone Ethics and Scientific Review Committee (dated 18 December 2018) and the Union Ethics Advisory Group (International Union against Tuberculosis and Lung Disease, Paris, France; EAG number 70/18). As we used aggregated data, the need for informed consent was waived by the ethics committees. Table 1 shows the numbers of fever cases reported and malaria tests done during the pre-, intra-and post-Ebola periods (see underlying data 5 ). Fever cases increased from 43,245 pre-Ebola to 74,367 post-Ebola (1.7-fold increase). Diagnosed malaria among fever cases ranged between 66% and 75%. During Ebola, all diagnosed malaria cases had malaria tests, while in the pre-and post-Ebola periods, 99% received testing and 1% was diagnosed on clinical grounds.

Numbers of reported fever cases and malaria tests done
Numbers treated for malaria in relation to ACT and fever onsets Table 2 shows numbers treated for malaria in relation to ACT and timing of fever onset. While all diagnosed malaria cases received treatment in the pre-and post-Ebola periods, only 47% were treated during Ebola. ACT use was 95% pre-Ebola, 99% intra-Ebola and dropped to 71% post-Ebola. In the post-Ebola period, out of 68 health facilities, an average of 7.5 facilities per month had 7 days stockouts of ACT.

Discussion
This study shows that despite the Ebola outbreak, the number of reported fever cases progressively increased and was 1.7 times higher post-Ebola compared to the pre-Ebola period.
Reassuringly, while less than half of malaria cases received treatment with ACT drugs during Ebola, in the post-Ebola period, all cases received treatment (which included ACT and other anti-malarial drugs). ACT use declined from 99% during Ebola to 71% post-Ebola, and this was accompanied with ACT stock-outs. This could be explained by the increase in numbers needed to be treated for malaria in the post-Ebola period (2.8-times that during Ebola), which might have caused pressure on available ACT stocks and supply chains.
As part of Sierra Leone's post-Ebola health recovery strategy 6 , more health personnel were trained in surveillance and reporting, incentives were introduced, and community involvement was promoted. Operationally, this seems to have increased health service utilisation for malaria, but the health system seemed unable to cope with the increased ACT demand. This needs to be addressed in order to ensure compliance with malaria treatment guidelines and the rational use of antimalarial drugs.
The main study strength is that we used data from all health facilities in the district and compared similar periods of time before, during and after the Ebola outbreak. A study limitation was the lack of data on complicated and uncomplicated malaria cases. Availability of such information would help to justify (or not) the use of other antimalarial drugs apart from ACT which was seen in the post-Ebola period. A short-coming in the district health information system (DHIS2) is that it does not capture severity of malaria (complicated and uncomplicated malaria). Adding this variable would improve monitoring of rational use of antimalarials.
In conclusion, what has changed since the Ebola outbreak is the increased utilisation of services for malaria. However, ACT stockouts are of concern, and this requires attention in order to ensure compliance with national malaria treatment guidelines.  For the analysis, in Table 1, the percentages of total cases diagnosed as malaria either by RDT or microscopy out of all fever cases are to be included and interpreted carefully. "Diagnosed malaria among fever cases ranged between 66% and 75%" is not enough. In fact, in post Ebola period, diagnosed malaria among fever cases was lower than pre and intra Ebola outbreak period (66%). This indicates less need for diagnostics despite increased utilization of health services for children with fever.

Data availability
In Table 2, during Ebola period, cases treated for malaria dropped to 47% which should be included in the discussion part.
Conclusion part is weak and needs to rewrite and add what needs to be done further.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Partly
No competing interests were disclosed. Competing Interests: I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative Your article is indexed in PubMed after passing peer review Dedicated customer support at every stage For pre-submission enquiries, contact research@f1000.com