Malaria incidence (i.e. cases per 1000 population at risk) reduced from 80 in 2000 to 58 in 2015 and 57 in 2019 globally [1, 8]. Despite these gains, the malaria situation in sub-Saharan Africa (SSA) remains volatile [9]. During the COVID-19 pandemic, there was a reversal of earlier reported gains in a reduction in mortality, with an excess of 69,000 deaths reported in the SSA in 2020 [2, 10]. This vulnerability seems to be driven by multiple factors that have shaped the dynamics of the malaria epidemiological quartet of vector, agent, host and environment [11]. The emergence of novel malaria-transmitting mosquito Anopheles stephensi mosquito in SSA [12], has posed a new threat to malaria control measures. In addition, the recent emergence of independent malaria drug resistance in Rwanda [13], Tanzania [14], and Uganda [15] has caused challenges in malaria control using artemisinin-based combination therapy (ACT).
There are also changing characteristics of the host. In recent studies, whereas children remain vulnerable to malaria, the age strata and clinical disease picture have changed [16,17,18]. Furthermore, tropical weather patterns and topography of marshlands accentuated by global warmings seem to improve the breeding of Anopheles mosquito and enhanced efficiency for its disease transmission capabilities in the last decade [19,20,21]. A constellation of these factors seems to have ushered in a new era of malaria in SSA, associated with variations in the clinical spectrum of severe malaria, but this has remained hitherto poorly described.
A total of 300 children, randomly selected among children being recruited into a longitudinal malaria study. This is a much larger sample size compared to recent similar studies for the work we report here [22]. Therefore, these new phenomena represent a more widespread change in malaria presentation during the malaria epidemic in the paediatric population of Eastern Uganda.
In this study, the common clinical features reported were prostration 236/300 (78.7%), jaundice 205/300(68.3%), severe malarial anaemia 158/300(52.7%), black water fever (BWF) 158/300 (52.7%), and multiple convulsions 51/300 (17.0%). Prolonged hospitalization was found in 56/251(22.3%). Participants with acidosis were more likely to have prolonged hospitalization P = 0.041. More than half 154/300 (51%) of the study participants had already been treated with either oral or parental anti-malarial. Our data show that the proportion of patients with each clinical feature has changed with some increasing tremendously compared to similar data reported in the same setting a decade ago [5]. It is plausible that there was an exponential increase in cases of malaria during the malaria epidemic that the Uganda Ministry of Health declared in June 2022. However, this does not necessarily explain the change in the clinical spectrum from the predominant respiratory distress and severe malarial anaemia a decade ago [19] to a higher proportion of children with prostration and BWF in the current data. This demonstrates a shift in age strata towards older children. Elsewhere, malaria presentation in older children is commonly associated with more cerebral and renal involvement compared to frequencies of these features in younger children [23]. Previous data highlighted that BWF is increasing in Eastern Uganda [18], but this time round, this report shows that it has become even more common. In the earlier and current data, there is an increase in the average age of children suffering from severe malaria. Even though BWF seems to be increasing, especially in Eastern Uganda [18], the pathophysiology is typically different and depends on clinical presentation. Our data demonstrate a triad of severe anaemia, jaundice and BWF, which point towards massive haemolysis.
Conversely, BWF associated with cerebral malaria [23, 24] is a pointer towards myolysis, as myoglobin has previously been detected in children with impaired consciousness and BWF [24]. The pathogenesis of cerebral malaria is due to damaged vascular endothelium by parasite sequestration, inflammatory cytokine production and vascular leakage, which result in brain hypoxia, as indicated by increased lactate and alanine concentrations [25]. It is possible the source of myoglobin in the children with cerebral malaria is multifactorial. Firstly, the muscle damage from generalized tonic–clonic seizures. Secondly, vaso-occlusive phenomenon occurs in muscle blood supply following damaged vascular endothelium by parasite sequestration similar to cerebral malaria. In Eastern Uganda, previous data consistently reported a low frequency of impaired consciousness in severe childhood malaria even with changes in age strata [5, 18, 26].
Other clinical observations in our current study may suggest that earlier treatment with antimalarials may be responsible for the variation in the clinical spectrum and outcomes observed. Over one-half of the study participants had taken anti-malarial before admission. In SSA, there is a paucity of data on the role of ACT in the causation or prevention of BWF. Earlier data reported that BWF was associated with quinine, the first-line drug for treating severe malaria two decades ago [26] or where it is still in use [27]. There have been postulations that artemisinin-based combinations fast-acting intra-erythrocyte activity results in death to malaria parasites, causing pitting of the infected red blood cells and resulting in premature apoptosis, autoimmune and spleen-driven lysis, culminating in massive haemolysis [28]. In some settings, case studies have indicated that ACT could be associated with BWF [29], but further research is needed in these setting. The presence of a rising proportion of BWF could as well be explained by presence of red cell abnormalities probably undiagnosed in this community. Olupot-Olupot et al. [18] had earlier reported prevalences among his patients for HbAS 4.6%, HbSS 4.1%, alpla-thalassaemia homozygous in 3.7% and heterozygous 38.0% and G6PD deficiency in 15.6%.
In this study, the median age was 4.6 years and a proportion of children below 5 years of 164/300 (54.7%). This demonstrates a shift towards older children and a lower proportion of children under the age of 5 years compared to a similar study in the same settings by Olupot-Olupot et al. [5] that reported an average age of 1.5 years for data collected a decade ago. Elsewhere, Kalinga et al. [30], assessed clinical manifestations and outcomes of severe malaria in children admitted to district hospital in Rungwe and Kyela in south-western Tanzania [30]. In their study, 1371 children were selected as screening group of which 409 (29.8%) tested positive for malaria. The mean age of the children was 2.7 (95%CI = 2.5, 2.8) years, and the majority (86%) were under 5 years of age [30]. These results may propose that the current epidemic of malaria-affected older children with waned immunity compared to earlier studies in which younger nonimmune children were affected. It is plausible that these age differences are also responsible for the differences in the clinical spectra of the disease between earlier and current malaria series.
In most malaria series, low rates of clinical jaundice have been reported. In Eastern Uganda, for instance, in 2012 Olupot-Olupot and colleagues [5] reported that jaundice accounted for 26.7% of children with severe malaria and now this report demonstrates a higher proportion 205/300 (68.3%) in the same settings. Therefore, this suggests that massive haemolysis is the underlying cause given the unique triad of severe anaemia, jaundice and BWF reported in the earlier studies [18] and the current data.
In Eastern Uganda, cerebral malaria has consistently remained low [18]. In the current data, cerebral malaria was in 10/137(6.0%) despite the older age of children with severe malaria in this study population. On the other features, the current data reports a reversal in the rates for respiratory distress. For instance, respiratory distress of 6.7% was much lower than in earlier series in the same settings [5, 18, 26]. The main difference is age, and it is well documented that respiratory distress is common in severe malaria in younger children [5]. The transition from quinine to artesunate could have influenced the change in the spectrum especially since artesunate has a rapid parasite clearance and the fact that more than half of the participants had pre-referral artemisinins. This could have reduced many incidences of cerebral malaria especially in the older children. There is also a possibility of increased haemolysis of the pitted red blood cells, which could be contributing to the rising jaundice and BWF with associated severe anaemia in these children. However, there is a paucity of data supporting this as earlier mentioned and more research is warranted especially in the most affected communities in Eastern Uganda.
On the outcomes, 52/251 (22.3%) study participants had prolonged hospitalization, similar to other series previously reported, participants with acidosis (P = 0.041), were more likely to have prolonged hospitalization. The overall in hospital mortality was 4.6%, similar to the 4.6% reported in Tanzania [30] but lower than earlier reports by Olupot-Olupot et al. in the same settings [5, 18, 26].
A retrospective study on adherence to malaria case management at this study site conducted between June 2021 and March 2022 indicated poor adherence to malaria treatment resulted in higher morality at 40/147 (27%) [22]. The difference with the current data could be attributed to several factors and the review period corresponded to the tail end of the COVID-19 lockdown. During this period, it was observed that few, but mainly very critically sick children accessed referral services at this facility, few health workers were working, and drug supplies were erratic as an effect of COVID-19. This effect of COVID-19 on malaria outcomes has been reported previously [2]. Compared to pre-COVID-19 of 63/662 (9.5%) for a similar study done May, 5 2011 until April,30 2012 in Eastern Uganda [5], the mortality in hospital mortality was much lower. This mortality in the first post-COVID-19 lockdown year is an indication of improvement attributable to recovery in the case management process. In addition, the older average age of the study participants in the current data and high rates (> 50%) of pre-admission anti-malarial treatment may have contributed to the observed outcomes.
Overall, the factors that were independently associated with mortality in this study were acidosis (P = 0.016) and impaired consciousness (P = 0.016). These same factors have been reportedly associated with poor outcomes in African children with severe malaria [31,32,33]. In addition, there were many runaways (self-discharges) whose outcomes could not be ascertained in their community. Against this background, the data on outcomes is incomplete. However, this has demonstrated that in the post-COVID-19 malaria epidemic, the clinical spectrum of disease is unique and mortality outcomes remain poor.
The limitations were; short duration, small sample size, lumbar puncture/CSF studies and blood cultures were not done. Adherence to anti-malarial therapy could not be closely tracked and neurological sequelae of severe malaria were not looked for and documented.