Lassa Fever in post-Ebola Sierra Leone: Sociodemographics and case fatality rates of in-hospital patients admitted at the Kenema Government Hospital Lassa Fever Ward between 2016-2018

1Department of Nursing, School of Community Health Sciences, Njala University, Sierra Leone 2Kenema Government Hospital Lassa FeverWard, Sierra Leone 3National Public Health Laboratory, Bo Government Hospital 4Department of Environmental Health Sciences, School of Community Health Sciences, Njala University, Sierra Leone 5Mercy Hospital Laboratory, Bo Sierra Leone 6Center for International Health, University of Munich (LMU), Germany RESEARCH ARTICLE


Introduction
Lassa fever (LF) is a zoonotic acute viral illness mainly found in West Africa. The disease was named after a town in Nigeria where the first case was discovered in 1969 which led to two deaths of female missionaries. 1LF is endemic in Sierra Leone, Liberia, Guinea and Nigeria; while other neighboring countries are at high risk of its outbreak since the animal vectors are distributed throughout the region [1]. The first LF case in Mali occurred in 2009 and that for Ghana in 2011 [2,3]. Few isolated LF cases have been reported in Ivory Coast, Central African Republic, Senegal, Congo and Burkina Faso. The animal reservoirs of LF are Mastomys natalensis, the Guinea multimammate rat, Mastomys erythroleucus and the African wood rat Hylomyscus pamfi [4,5,6,7,8]. Mastomys natalensisis the reservoir in Sierra Leone and is commonly found in rural environments where they breed frequently. These LF vectors are widely distributed throughout central, west and east Africa. Their excreta contain the Lassa virus (LASV) and are the source of transmission of LF. Humans are infected by coming in contact with the contaminated excreta of these rats or by eating the rat vectors themselves.
An estimated 100,000 to 300,000 LF cases with approximately 5,000 deaths is reported annually; although these figures are largely under reported [9]. The case fatality rate (CFR) from LF for healthy population with no underlying comorbidity was estimated between 1-2%, 2-5% for in-hospital LF cases, spiking to 20-60% for laboratory confirmed or nosocomial outbreak LF cases [10][11][12][13].
In Sierra Leone, LF which is one of the few diseases for which weekly epidemiological reporting to the health ministry is requested accounts for 10% -16% of hospitalized cases annually [14].
In this retrospective mixed cohort study we analysed the treatment history containing the sociodemographic and clinical characteristics of 52 laboratory-confirmed LF cases that were admitted to the Kenema Government Hospital Lassa Fever Ward (KGHLFW) during 2016 to 2018 (i.e. during the post Ebola outbreak in Sierra Leone). The LF patients whose treatment history we analysed came from either within or outside Kenema district where the KGHLFW is located. These LF patients were diagnosed at the KGHLFW triage after they presented with signs and symptoms of LF.
The main drug of choice for the treatment of LF patients was injection ribavirin through Intravenous route. The medication was administered over a ten day period starting from the date of admission up to the tenth day. The ribavirin was supplemented by other medications including IV fluids, antibiotics (ceftriazone) given at 1g/bd for five days, antimalarial drugs(artemether and lumefantrine),antipyretics(paracetamol/panadol), analgesics(ibuprofen). Additionally these treatments can however be discontinued or continued after ten days depending on whether the patient has improved or not. The show of signs and symptoms after ten days is indicative that the patient will continue to take the medication until signs and symptoms disappear.
We obtained ethical approval for this study from the Njala University Institutional Review Board which waivered the requirement to obtain individual informed consent from the LF patients whose medical records we analyzed since these data were healthcare facility-specific aggregated patient records. R software package version 3.3.1 was used for all descriptive statistical analysis in this study and p-values < 0.05 were considered significant for all two-sided statistical tests [17]. Our descriptive analysis outputs were presented as frequencies, proportions, means and standard deviations (for normally distributed continuous variables); medians and interquartile ranges (for not normally distributed continuous variables).

Methods
For 2016, most of the LF cases were recorded in January (14.3%, n = 3/21) and February (14.3%, n = 3/21); no LF case was recorded for the months of April, May, October, November and December (   There is paucity of research publications on the treatment outcome of LF outbreaks which is assumed to be due to underreportingsince LF presents with non-specific clinical presentations especially during the early phases of its outbreak coupled, its outbreak dynamics, and the attached societal structures that favour its transmission.
With the exception of 2017, January witnessed high LF cases in the affected regions for the period under review. This finding was also corroborated by Akhmetzhanov Andrei R and colleagues who reported that the high-risk for LF transmission to human peaks in January-March. The early months of the year in Sierra Leone is characterized by winter-like climate of low temperature, high humidity and few rainfalls [22]. The average night-time and daytime temperatures in the interior of Sierra Leone in the early months of the year is around 18/20°C and 32/34°C respectively. These climatic conditions in the interior of Sierra Leone around January-March are out of the optimal climatological factors that favor LAF transmission. Fichet-Calvet and colleagues have shown previously that rainfall was an important abiotic factor for LF transmission and that LASV transmission hike greatly in the raining season than dry season [23]. Sierra Leone's raining season start in May and extend till October. Because the highest LF prevalence in our study is out of the raining season period we are assuming that other factors interplay with climate in the transmission of LASV in Sierra Leone [14,19].
Our in-hospital CFRs for the laboratory-confirmed LF cases was similar to previous studies done in Sierra Leone and Nigeria but also higher than those reported for LF outbreaks in other countries [10,11,12,14,19,24,25,26]. CFR can reach 50% among hospitalized LF patients including those with the severe form of the infections. Our high CFR is interesting considering the low number of LF cases which were expected to produce a decrease in the statistical effect size for this study [25,26].
Generally men recorded high CFR throughout the study period compared to women; though they recorded lower prevalence than women. We are also of the opinion that because men most often stay away from the house for a longer period of time compared to women they may be unfamiliar with the signs and symptoms of LF which also put them at elevated risk of dying from the infection because of delays in seeking treatment. Knowledge about a disease is paramount in preventing and controlling its transmission, as well as obtaining successful treatment.
We reported a high LF prevalence among adults than children < 15 years which indicates the significance of age in the general epidemiology of the disease during outbreaks. Our progressive decline in LF cases as we move away from 2016 to 2018 emphasizes the need for maintaining a robust public health surveillance system in those countries that lies along the LF belt. A weak public health system is a recipe for LF outbreak. Our high LF cases observed in January for 2016 and 2018 once more showed the seasonality of LF; which has implication for its control and preventive measures.Our high CFR among men compared to women could be attributed to the fact that because men stay away from the house for a longer period of time may be unfamiliar with the signs and symptoms of LF which makes them to seek treatment late. Knowledge about a disease is paramount in obtaining a successful treatment outcome.
Our study reported an overall high CFR for the entire period under review. We also reported a disproportionately high LF prevalence for female than male -which is consistent with one Nigerian LF study [18]. Isere EE and colleagues reported that majority (52.6%) of confirmed LF cases during the December 2015 to April 2016 outbreak in Nigeria were females. Ilori EA and colleagues however reported lower female LF cases compared to males (37.9% vs. 62.1%) for the January -May 2018 outbreak [18]. This sex-based difference in LF prevalence once more highlights the changing epidemiology of the disease. Our finding also underpins the influence of gender role in the transmission of LF [19]. Women in Sierra Leone spend more time at home and hence have increased contact period with the rodent vectors than their men counterparts. We also reported high LF cases among adults than children < 15 years; which is also consistent with the finding by Ilori EA and colleagues. There was a progressive decline in LF prevalence as we move away from 2016 to 2018 [19]. The high LF prevalence in 2016 can be attributed to the disruption of LF surveillance, preventive and control activities caused by the 2014 -2016 Ebola (EVD) outbreaks in Sierra Leone. The 2014 -2016 EVD outbreaks greatly affected the health sectors in Sierra Leone, Liberia and Guinea which are also foci for LF outbreak. During the 2014 -2016 EVD outbreaks great human resources and logistics were channeled towards bringing the outbreak to an end. As the EVD outbreak waned the health sector in the affected West African countries picked up and became better organised with foreign aid in the area of disease surveillance, prevention of health care associated infection and control -this was reflected in the successive reduction in the prevalence of LF in 2017 and 2018. Sierra Leone became the first Africa country in 2019 to employ the electronic Integrated Disease Surveillance and Response System (eIDSR) for its disease surveillance and monitoring activities; which provides accurate, timely, and comprehensive health surveillance data in a more efficient manner than the traditional paperbased IDSR system [20,21].

Discussion
Conclusion