INTRODUCTION — Severe malaria is defined as presence of Plasmodium falciparum parasitemia and one or more of the manifestations in the table (table 1).
Most cases of severe malaria are attributable to P. falciparum (90 percent), but Plasmodium vivax and Plasmodium knowlesi can also cause severe disease [1-5]. Approximately 445,000 deaths are caused by malaria each year; over 90 percent of the deaths occur among children in sub-Saharan Africa [1].
Management of patients with severe malaria presents a broad array of clinical challenges given the complex pathophysiology of an infection that involves multiple organ systems. These challenges are increased in endemic areas where access to diagnostic and therapeutic tools may be limited.
The approach to treatment of severe malaria will be reviewed here.
Issues related to diagnosis of malaria are discussed further separately. (See "Malaria: Clinical manifestations and diagnosis in nonpregnant adults and children", section on 'Severe malaria' and "Laboratory tools for diagnosis of malaria".)
Issues related to treatment of uncomplicated malaria are discussed separately. (See "Treatment of uncomplicated falciparum malaria in nonpregnant adults and children" and "Malaria in pregnancy: Prevention and treatment".)
DEFINITION — Severe falciparum malaria is defined as one or more of the following in the absence of an identified alternative cause and in the presence of P. falciparum parasitemia [5]:
●Impaired consciousness – Glasgow coma score <11 in adults or Blantyre coma score <3 in children.
●Prostration – Generalized weakness so that a person is unable to sit, stand, or walk without assistance.
●Multiple convulsions – More than two episodes within 24 hours.
●Acidosis – A base deficit of >8 mEq/L, or, if not available, a plasma bicarbonate level of <15 mmol/L, or venous plasma lactate ≥5 mmol/L. Clinical indicators of acidosis include rapid, deep, labored breathing.
●Hypoglycemia – Blood or plasma glucose <40 mg/dL (<2.2 mmol/L).
●Severe malarial anemia – Hemoglobin concentration ≤5 g/dL or hematocrit ≤15 percent in children <12 years of age (<7 g/dL and <20 percent, respectively, in adults) with parasite count >10,000/mcL.
●Renal impairment – Plasma or serum creatinine >3 mg/dL (265 mcmol/L) or blood urea >20 mmol/L.
●Jaundice – Plasma or serum bilirubin >50 mcmol/L (3 mg/dL) with a parasite count >100,000/mcL (approximately 2 percent).
●Pulmonary edema – Radiographically confirmed or oxygen saturation <92 percent on room air with respiratory rate >30/min, often with chest indrawing and crepitation on auscultation.
●Significant bleeding – Including recurrent or prolonged bleeding from the nose, gums, or venipuncture sites, hematemesis, or melena.
●Shock – Compensated shock is defined as capillary refill ≥3 seconds or temperature gradient on leg (mid to proximal limb) but no hypotension. Decompensated shock is defined as systolic blood pressure <70 mmHg in children or <80 mmHg in adults, with evidence of impaired perfusion (cool peripheries or prolonged capillary refill).
●P. falciparum parasitemia >10 percent (>500,000/mcL).
The definition of severe P. vivax malaria is the same as that of severe falciparum malaria except that there are no P. vivax parasite density thresholds [5]. The definition of severe malaria due to P. knowlesi differs from that of severe falciparum malaria; the threshold parasite density is >100,000/mcL (alone) or >20,000/mcL in patients with jaundice [5]. (See "Non-falciparum malaria: Plasmodium knowlesi", section on 'Severe infection'.)
Administration of empiric treatment for malaria is warranted for patients with the above manifestations while clinical evaluation for alternative causes is pursued.
Issues related to clinical manifestations of severe malaria are discussed further separately. (See "Malaria: Clinical manifestations and diagnosis in nonpregnant adults and children", section on 'Severe malaria'.)
INITIAL MANAGEMENT ISSUES
General principles — Death due to severe malaria can occur within hours of presentation, so prompt assessment and initiation of antimalarial therapy is essential. Patients should be evaluated with attention to findings consistent with malaria as well as additional and/or alternative causes of presenting symptoms. A full neurologic assessment should be performed, including assessment of the Blantyre coma score for children (table 2); the Glasgow coma scale is suitable for adults (table 3). Temperature, heart rate and rhythm, respiratory rate and rhythm, blood pressure, oxygen saturation, and weight should be noted, as should capillary refill and degree of pallor. (See "Stupor and coma in adults" and "Evaluation of stupor and coma in children".)
Of primary importance in the treatment of malaria is the provision of prompt, effective antimalarial therapy and concurrent supportive care to manage life-threatening complications of the disease. Supportive measures (eg, oxygen, ventilatory support, cardiac monitoring, and pulse oximetry) should be instituted as needed. During this time, intravenous catheters should be placed, and fingerprick blood samples should be obtained for laboratory tests needed immediately. Point-of-care testing machines can be used for rapid determination of hematocrit (packed cell volume or hemoglobin), glucose, and lactate. Additional tests can be done if/when indicated: electrolytes, blood gases, renal function tests, full blood count, type and cross, blood culture, and clotting studies. Unconscious patients should have a lumbar puncture to rule out concomitant bacterial meningitis in the absence of contraindications (eg, papilledema). These tasks should overlap with institution of antimalarial treatment as well as other ancillary therapies as needed (including anticonvulsants, intravenous glucose and fluids, antipyretics, antibiotics, and blood transfusion).
Repeat clinical assessments should be performed every two to four hours for prompt detection and management of complications (in an intensive care setting, if possible). If the coma score decreases after initiation of treatment, investigations should focus on the possibility of seizures, hypoglycemia, or worsening anemia. Repeat laboratory assessments of parasitemia, hemoglobin/hematocrit, glucose, and lactate should be performed in six-hour intervals. A flowchart summarizing the vital information may be used to guide management decisions (table 4) [6].
Important independent predictors for fatality among African children with severe malaria include acidosis, impaired consciousness (coma and/or convulsions), elevated blood urea nitrogen, and signs of chronic disease (lymphadenopathy, malnutrition, candidiasis, severe visible wasting, and desquamation) [7]. This was illustrated in a large clinical study including 5246 African children with severe malaria. Other factors that were statistically significant predictors of death in univariate analyses but not in the multivariate model included tachypnea, deep breathing, shock, prostration, low pH, hyperparasitemia, severe anemia, and jaundice. Clinical features previously identified as being poor prognostic features that did not correlate with mortality in this study included age, glucose level, axillary temperature, parasite density, and blackwater fever.
Careful observation and thoughtful responses to changes in clinical status are the most important elements in looking after patients with severe malaria. Patients can make remarkable recoveries, and the time and effort to address the components of clinical care described in the following sections can reap tangible rewards in a relatively short period of time.
Antimalarial therapy — The risk of death due to severe malaria is greatest in the first 24 hours after clinical presentation.
Intravenous therapy should be initiated promptly, with close monitoring of parasite density. There are two major classes of drugs available for parenteral treatment of severe malaria: the artemisinin derivatives (artesunate and artemether) and the cinchona alkaloids (quinine and quinidine) [5].
Patients who have received parenteral therapy for at least 24 hours and can tolerate oral medication may transition to an oral regimen for completion of therapy.
Prereferral treatment — In areas where patients with severe malaria cannot begin intravenous therapy immediately and where artemisinin derivatives are readily available, patients should be treated with a prereferral dose of intramuscular or rectal therapy and triaged to an acute care facility.
A single dose of intramuscular artesunate is preferred [5]. Children weighing less than 20 kg should receive 3 mg/kg/dose. For larger children and adults, the dose is 2.4 mg/kg [5].
If intramuscular artesunate is not available [5]:
●For individuals ≥6 years, intramuscular artemether (loading dose 3.2 mg/kg) should be used; if this is not available, intramuscular quinine (loading dose 20 mg/kg, divided into two 10 mg/kg injections, one in each anterior thigh to avoid damaging the sciatic nerve) should be used.
●For individuals <6 years, rectal artesunate (10 mg/kg) should be used. If rectal artesunate is not available, intramuscular artemether should be used (loading dose 3.2 mg/kg); if this is not available, intramuscular quinine should be used (loading dose 20 mg/kg, divided equally between the anterior thighs).
If referral is not possible, intramuscular or rectal treatment should be continued until the patient can tolerate oral medication, at which point the course of therapy may be completed orally. (See 'Completing therapy' below.)
A single artesunate rectal suppository pending transport has been demonstrated to reduce mortality in children <6 years of age. This was illustrated in a randomized trial of over 12,000 patients in rural Bangladesh, Ghana, and Tanzania with suspected severe malaria [8]. Dosing consisted of 100 mg for children 6 months to 6 years of age and 400 mg for patients >6 years. Mortality was significantly lower among those who received prereferral rectal artesunate than among those who received placebo (1.9 versus 3.8 percent, respectively; risk ratio 0.49, 95 percent CI 0.32-0.77). Rectal artesunate is not advised for older children and adults as it has been associated with increased death in these groups (low-quality evidence) [5,9].
Initiating therapy — Parenteral artesunate (intravenous or intramuscular) is preferred for treatment of adults and children with severe malaria (including infants, pregnant women in all trimesters, and lactating women) [5,10-12]. If parenteral artesunate is not available, intramuscular artemether should be administered. If artemether is not available, intravenous quinine/quinidine may be administered.
Dosing for parenteral therapy is summarized in the table (table 5) [5,13]. Parenteral therapy should be administered for at least 24 hours and until oral medication can be tolerated. (See 'Completing therapy' below.)
Artemisinins
General principles — Artemisinin derivatives clear parasitemia faster than quinine and are associated with lower mortality rates in both adults and children [10,11,14]. Artemisinins are active against a broader life cycle range of blood stage parasites than quinine and they are active against gametocytes [15,16]. Artesunate is the most rapid acting of the artemisinin compounds because of its water solubility [5,15].
Artesunate is the preferred therapy for treatment of severe falciparum malaria in adults and children (in areas where intravenous artesunate of reliable quality is readily available) [5,10-12,14,17]. This was illustrated in a meta-analysis including eight randomized trials enrolling more than 1600 adults and 5700 children [11]; treatment with artesunate significantly reduced the risk of death in adults (risk ratio [RR] 0.61, 95% CI 0.50-0.75) and children (RR 0.76, 95% CI 0.65-0.90). In a retrospective study including more than 100 patients with severe malaria treated with artesunate under a United States Centers for Disease Control and Prevention (CDC) investigational protocol, artesunate was a safe and clinically beneficial alternative to quinidine [18]. (See 'Quinine/quinidine' below.)
Artesunate dosing need not be adjusted for hepatic or renal failure, nor for concomitant or previous therapy with other medications (including mefloquine, quinine, or quinidine) [15]. The most common adverse effects include nausea, vomiting, anorexia, and dizziness, although these may be due to malaria rather than drug toxicity. Delayed onset of anemia has been noted; this is discussed further below. Patients treated with intravenous artesunate should be monitored for delayed hemolytic anemia, with repeat hemoglobin testing at 7, 14, and 30 days after treatment [5,19,20].
Potential benefits of the World Health Organization-recommended regimen over a simplified regimen [21] are described in a model which moves beyond parasite clearance time as the key metric of efficacy and considers the sequestered mass of parasites, the lingering effects of ruptured sequestered red cells, and the stage-specific effects of the artemisinins (earlier life cycle stages being less susceptible) [22].
There is no convincing evidence of neurotoxic effects in humans due to oral or intravenous artemisinins, although neurotoxicity has been described in animals and attributed to fat-soluble artemisinins more frequently than to artesunate [23]. Limited data are available on the use of artesunate for severe malaria during pregnancy. (See 'Pregnancy' below.)
Following intravenous artesunate, oral antimalarial treatment may be administered if the patient is able to tolerate oral therapy. Artesunate can also be administered intramuscularly, orally, or via rectal suppository [8]. (See 'Completing therapy' below and 'Prereferral treatment' above.)
Artesunate in the United States — In the United States, intravenous artesunate was approved by the US Food and Drug Administration in May 2020 [24]. As of June 2021, the drug can be obtained through major drug distributors [25]. In hospitals where patients with malaria are encountered, it is advisable for pharmacies to stockpile a few doses to facilitate timely administration.
Indications for intravenous artesunate include:
●Patient has a positive blood smear AND severe malaria as defined by one or more of the following:
•Hyperparasitemia (>5 percent)
•Impaired consciousness
•Seizures
•Circulatory collapse/shock
•Pulmonary edema or acute respiratory distress syndrome
•Acidosis
•Acute kidney injury
•Abnormal bleeding or disseminated intravascular coagulation
•Jaundice (must be accompanied by at least one other sign)
•Severe anemia (hemoglobin < 7 g/dL)
●Patient has a positive blood smear AND is unable to tolerate oral medications (regardless of malaria severity)
●In exceptional cases, intravenous artesunate can be used empirically (eg, in the setting of strong clinical suspicion for severe malaria but a timely, reliable microscopic diagnosis is not available), in consultation with a CDC malaria expert.
While awaiting delivery, preliminary treatment with an oral antimalarial may be reasonable, following administration of an antiemetic if needed (for comatose patients, a nasogastric tube may be used). The following oral antimalarials may be used in this circumstance:
●Artemether/lumefantrine (one tablet = 20 mg artemether and 120 mg lumefantrine)
•Adults: four tablets as one dose, and if still needed repeat eight hours later
•Children (≥5 kg only) dosing based on weight:
-5 to 14 kg: one tablet per dose
-15 to 24 kg: two tablets per dose
-25 to 34 kg: three tables per dose
-≥35 kg: four tablets per dose
●Atovaquone/proguanil: Adult (250 mg atovaquone/100 mg proguanil) and pediatric (62.5 mg atovaquone/25.0 mg proguanil) formulations are available.
•Adults: four adult tablets as one dose
•Children (≥5 kg only) dosing based on weight:
-5 to 8 kg: two pediatric tablets
-9 to 10 kg: three pediatric tablets
-11 to 20 kg: one adult tablet
-21 to 30 kg: two adult tablets
-31 to 40 kg: three adult tablets
->40 kg: four adult tablets
●Quinine: Adults: 650 mg (salt) every eight hours. Children: 10 mg (salt)/kg every eight hours.
●Mefloquine (because of a risk of severe neuropsychiatric adverse events at treatment doses, mefloquine should only be used if other options are not available, and based on health care provider judgement, treatment is needed prior to the arrival of intravenous artesunate): Adults: 750 mg salt, initially, then if needed, 500 mg salt 6 to 12 hours after initial dose. Children: 15 mg salt/kg, initially, then if needed, 10 mg salt/kg 6 to 12 hours after initial dose.
In addition, intravenous tetracyclines such as doxycycline should not be used; these drugs are slow-acting antimalarials that would not take effect until well after 24 hours and are not effective antimalarials for treatment of severe malaria when used alone.
When intravenous artesunate arrives, the oral medication should be discontinued, and the intravenous artesunate can be administered according to the standard guidelines (table 5).
The CDC Malaria Hotline (770-488-7788) is available Monday through Friday, 9 am to 5 pm EST. Outside these hours, providers should call 770-488-7100 and ask to speak with a CDC Malaria Branch clinician.
Adverse effects and resistance — Delayed onset of anemia following treatment of severe malaria with artesunate has been described [26-32]. In one review, anemia occurred 8 to 32 days after completion of artesunate therapy; the mean hemoglobin nadir was 6.2 g/dL, and transfusion was required in some cases [33]. The anemia improved within four to eight weeks following completion of artesunate therapy.
The mechanism of delayed-onset anemia following administration of artesunate is not fully understood but is associated with increased levels of ring erythrocyte surface antigens (measured by flow cytometry) [19] or by plasma P. falciparum histidine-rich protein-2 (PfHRP2) concentrations (measured by dipstick shortly after parasite clearance) [34]. Artesunate may kill malaria parasites without destroying the red blood cell (RBC), which makes the infected RBCs more deformable than other cells, thereby increasing the apparent number of surviving RBCs following parasitemia relative to the number of RBCs in patients treated with other antimalarial agents [19,35,36]. However, the deformable RBCs have a shorter lifespan than nonparasitized RBCs, which may account for delayed-onset anemia following artesunate treatment. Further data are needed to understand if and how artesunate may affect risk for delayed-onset anemia; in general, patients with malaria should have follow-up surveillance hemoglobin concentration evaluated four weeks following completion of treatment for severe malaria. (See "Drug-induced hemolytic anemia", section on 'Other mechanisms'.)
Emergence of artemisinin resistance is an important concern, and combination of artemisinins with other active agents may protect against the development of resistance to individual drugs [37]. Pending further data, use of intravenous artesunate monotherapy remains appropriate for treatment of severe malaria. After the patient is no longer critically ill, oral combination therapy is typically used to complete the course of treatment. (See 'Completing therapy' below.)
Quinine/quinidine — Intravenous quinine remains the treatment of choice for areas where intravenous artesunate of reliable quality is not readily available.
As of April 2019 in the United States, intravenous quinidine will no longer be procured by United States hospitals and supplies for the treatment of severe malaria will dwindle; intravenous quinine was withdrawn by the CDC as an antimalarial treatment in the 1990s [38]. Parenteral quinine should be administered in conjunction with doxycycline, tetracycline, or clindamycin.
Dosing for parenteral therapy is summarized in the table (table 5) [5,13]. Parenteral therapy should be administered for at least 24 hours and until oral medication can be tolerated. For patients with malaria acquired in Southeast Asia, the total duration of quinine/quinidine-based therapy is seven days; for patients with malaria acquired in Africa or South America, the total duration of quinine/quinidine-based therapy is three days. (See 'Completing therapy' below.)
Quinidine can cause QTc prolongation and should be administered with electrocardiographic monitoring [39]. The QT interval should be monitored hourly, and the infusion should be stopped if the corrected QT interval becomes prolonged by more than 50 percent of the baseline value. The infusion can be renewed (without a bolus) once the QTc falls to <25 percent above the original value. Such monitoring is not necessary in the setting of quinine administration in patients without cardiac abnormalities. (See "Major side effects of class I antiarrhythmic drugs".)
Both quinine and quinidine can act as pancreatic secretagogues, leading to hyperinsulinemic hypoglycemia. Other toxic effects include tinnitus, reversible hearing loss, nausea, vomiting, dizziness, and visual disturbances.
Caution should be exercised for patients recently treated with mefloquine because of the potential for cardiotoxic drug-drug interactions.
Monitoring parasite density — During treatment of severe malaria, parasite density should be monitored every 12 hours during the first two to three days to document declining parasite density and confirm adequate response to therapy [40-42]. Thereafter, daily blood smears should be performed until smears are negative or until treatment day 7 (if discharged prior to complete clearance of parasitemia, then negative parasitemia should be documented as an outpatient) [41]. (See "Laboratory tools for diagnosis of malaria", section on 'Quantification of parasitemia'.)
With artemisinin therapy, median parasite clearance times are <72 hours [43,44]; prolonged parasite clearance times have been observed in areas of Southeast Asia. With quinine/quinidine therapy, parasite density may be expected to fall by 90 percent over the first 48 hours [41].
If parasitemia does not clear as expected, confirm that the correct dose is being administered, and consider the possibility of drug resistance or substandard or counterfeit drugs [45]. Administration of an alternative antimalarial for severe disease could be considered.
Completing therapy — For nonimmune patients with parasite density >1 percent (at least four hours after the third dose of parenteral therapy), intravenous therapy should be continued up to 6 more days until parasite density ≤1 percent, after which an oral regimen should be given [41,46]. For patients in endemic areas who have received at least 24 hours of parenteral therapy and are able to tolerate oral medications, treatment may be switched to oral therapy even if parasitemia is above 1 percent [5].
An oral artemisinin combination therapy (ACT) regimen (three-day course) is preferred; if an oral ACT is not available, an alternative regimen may be used. ACT regimens, alternative regimens, and dosing are summarized in the table (table 6).
Patients treated with intravenous artesunate should be monitored for delayed hemolytic anemia, with repeat hemoglobin testing at 7, 14, and 30 days after treatment [5,19,20].
Regimens containing mefloquine should be avoided for completing therapy in patients who presented with altered consciousness, since there is an increased incidence of neuropsychiatric toxic effects associated with mefloquine in the setting of cerebral malaria [5].
Role of postdischarge prophylaxis — Administration of postdischarge malaria prophylaxis may be a useful tool for reducing morbidity and mortality among children <5 years of age hospitalized for severe anemia in areas of Africa where malaria is endemic. However, further study is needed to inform the optimal timing, dose, duration, and potential synergism of prophylaxis with other preventive tools (such as vaccines, bed nets, and iron and/or folate supplementation).
In a randomized trial including 1049 children <5 years of age in Kenya and Uganda hospitalized with severe anemia (hemoglobin <5 g/dL), 85 percent had confirmed malaria infection or were treated for presumed malarial infection [47]. The children received artemether-lumefantrine (three-day course) at discharge and then were randomly assigned to receive placebo or prophylaxis with dihydroartemisinin-piperaquine (administered as three-day courses at 2, 6, and 10 weeks postdischarge). During the intervention period (weeks 3 to 14), the rate of death or readmission was lower among those who received prophylaxis (death rate 2.2 versus 3 percent [hazard ratio 0.74, 95% CI 0.35-1.56]; readmission rate 34 versus 58 percent [hazard ratio 0.63, 95% CI 0.52-0.77]). However, these outcomes were not sustained through 26-week follow up. No serious adverse events were attributed to dihydroartemisinin-piperaquine.
Issues related to malaria chemoprevention in areas of Africa with seasonal transmission are discussed separately. (See "Malaria: Epidemiology, prevention, and control", section on 'Children in areas with seasonal transmission'.)
Reducing transmissibility — In endemic areas with low transmission, a single dose of oral primaquine, targeting intravascular gametocytes, should be administered to reduce transmissibility of treated P. falciparum infection. This precaution is not necessary in areas in which malaria is not endemic, and it is not cost-effective in moderate-to-high transmission settings. (See "Treatment of uncomplicated falciparum malaria in nonpregnant adults and children", section on 'Reducing transmissibility'.)
Pregnancy — Pregnant women are more likely to develop severe P. falciparum malaria than other adults, particularly in the second and third trimesters. Complications such as hypoglycemia and pulmonary edema are more common than in nonpregnant individuals. Maternal mortality can approach 50 percent, and fetal death and premature labor are common.
Prompt antimalarial therapy and supportive care should be administered as outlined in the preceding sections. Parenteral artesunate is the treatment of choice in all trimesters [5]. Among the alternative agents, artemether is preferred over quinine since quinine is associated with recurrent hypoglycemia [5].
Following administration of parenteral therapy (for at least 24 hours and until oral medication can be tolerated), an oral regimen should be administered. Artemisinin combination therapy (ACT) is preferred; if an oral ACT is not available, an alternative regimen may be used. Regimens and dosing are summarized in the table (table 7).
Other issues related to malaria and pregnancy are discussed in detail separately. (See "Malaria in pregnancy: Prevention and treatment".)
OTHER ASPECTS OF MANAGEMENT
Respiratory status — Hypoxemia and rales are not common in the setting of severe malaria; the presence of either should raise suspicion for a concomitant lower respiratory tract infection [48].
Pulmonary edema may develop, particularly in the settings of renal impairment or severe malarial anemia. Acute respiratory distress syndrome can also complicate severe malaria. The approach to ventilatory management ranges from supplemental oxygen to mechanical ventilation with positive end expiratory pressure. (See "Ventilator management strategies for adults with acute respiratory distress syndrome".)
Deep breathing (Kussmaul respirations) is a clinical indicator of metabolic acidosis and is associated with a worse outcome in patients with falciparum malaria [49]. Often, deep breathing is associated with hyperlactatemia, although a study of 3248 Tanzanian children noted that, among the 164 deaths, 45 children with admission blood lactate concentrations >5 mmol/L had no evidence of deep breathing/respiratory compensation [50]. This same study confirmed previous findings of the prognostic significance of hyperlactatemia; that association was even more pronounced in children with severe nonmalarial illness.
Neurologic status — The clinical case definition of cerebral malaria includes the following criteria [5]:
●Blantyre coma score ≤2 (table 2)
●P. falciparum parasitemia (any density)
●No other identifiable cause of coma (eg, hypoglycemia, meningitis, or a postictal state) [5]
The histologic hallmark of cerebral malaria is cerebral sequestration of parasitized erythrocytes. Autopsy-based studies have demonstrated that cerebral malaria may be incorrectly diagnosed (based on the clinical case definition) in about 25 percent of cases [51,52].
The most reliable clinical indicator for cerebral malaria in patients who meet the above clinical case definition is the presence of one or more elements of malaria retinopathy: white-centered hemorrhages (picture 1A), vessel changes (picture 1B), and whitening in areas of the retina (picture 1C) [53-55]. These manifestations are associated with severe malaria caused by P. falciparum but do not appear to be a major component of severe malaria caused by P. knowlesi [56].
Establishing whether retinopathy is present is an important marker for cerebral malaria. In the absence of this finding, alternative causes for coma (such as bacterial infection) should be pursued and treated, even in the presence of established malaria infection. (See 'Bacterial infection' below.)
In children with cerebral malaria, factors highly associated with mortality include increased brain volume (on magnetic resonance imaging) and associated elevated intracranial pressure [57]. The etiology of increased brain volume remains poorly understood. In the absence of ventilatory support, herniation is a common cause of death.
Issues related to cerebral malaria are discussed further separately. (See "Malaria: Clinical manifestations and diagnosis in nonpregnant adults and children", section on 'Cerebral malaria'.)
Clinical evaluation — Clinical evaluation includes full physical exam, a complete neurologic examination, calculation of Blantyre coma score, and funduscopic evaluation. Patients with altered sensorium should undergo lumbar puncture (in the absence of contraindications) to exclude concomitant bacterial meningitis [5].
●Blantyre coma score – The Blantyre coma score is a clinical indicator of severity in children with altered consciousness due to malaria (table 2) [58]. It was developed based on modifications of Glasgow coma score because some key Glasgow indicators are not appropriate for evaluation of altered consciousness in pediatric malaria. Examples include eye opening in response to pain (open-eyed children can be in coma) and verbal response to pain (many children with severe malaria are not yet able to speak) [59].
The Blantyre coma score is outlined in the table (table 2). Fully conscious children score 5; children who do not respond to painful stimuli score 0. Blantyre coma score ≤2 is associated with high risk of mortality [59]. Among 2030 children admitted to a pediatric ward in Malawi with Blantyre coma score ≤2, the mortality in children with a score of 0, 1, or 2 on admission was 34.2, 16.7, and 10.4 percent, respectively. Patients meeting the standard clinical case definition of cerebral malaria had overall mortality of 17.8 percent. (See 'Neurologic status' above.)
The Blantyre coma score should be reassessed at regular intervals following initiation of therapy. A decrease should prompt re-evaluation for seizures (including consideration of unwitnessed or subclinical events), anemia, and hypoglycemia.
●Funduscopic exam – Malarial retinopathy is pathognomonic for cerebral malaria in patients who satisfy the standard clinical case definition [60]. The optic fundi should be evaluated following instillation of mydriatics for pupillary dilatation. Examination should be performed via direct ophthalmoscope (which provides magnification) and indirect ophthalmoscope (which provides three-dimensional perspective and wide field of view). Features of malarial retinopathy include white-centered hemorrhages, vessel changes, and whitened areas of the retina (picture 1A-C) [53].
Measurement of plasma pHRP2 concentration may be a useful marker for the presence of retinopathy in areas with high endemicity [61-63]; use of this tool could reduce the need for funduscopic examination, which can be difficult to perform. In a series of 64 children in Malawi with cerebral malaria (defined by presence of retinopathy or histologic evidence of sequestration at autopsy), the sensitivity and specificity of a plasma pHRP2 level of >1700 ng/mL were 98 and 94 percent, respectively [61]. Thus far, pHRP2 is detectable only via enzyme-linked immunosorbent assay, which may not be practical in endemic areas.
●Lumbar puncture (LP) – Patients with altered sensorium should undergo LP (in the absence of contraindications) to exclude concomitant bacterial meningitis. In one study including 1075 comatose children with suspected cerebral malaria (of whom 80 percent underwent LP), undergoing LP did not influence the mortality risk [64].
If an LP is not performed, presumptive antibiotic therapy for bacterial meningitis should be initiated. (See "Lumbar puncture: Indications, contraindications, technique, and complications in children" and "Lumbar puncture: Technique, indications, contraindications, and complications in adults".)
In patients with cerebral malaria, the mean opening pressure is about 16 cm of cerebrospinal fluid (CSF). Laboratory examination may be normal or may demonstrate slightly elevated total protein level and cell count. In a study comparing CSF findings from 12 children with cerebral malaria and 14 children with presumed viral encephalitis, patients with cerebral malaria had lower white cell count, glucose, and protein levels [65]. Children with malaria had a mean white cell count of 0 cells/microL (range 0 to 4 cells/microL); children with viral encephalitis had a mean white cell count of 4 cells/microL (range 0 to 9 cells/microL). A CSF glucose concentration below 3.4 mmol/L (61 mg/dL) was the best discriminator of cerebral malaria from presumed viral encephalitis.
Seizure management — Seizures occur in up to 70 percent of children with severe malaria; subclinical seizures occur in 15 to 20 percent of cases [60,66]. In an African study comparing 132 pediatric survivors of retinopathy-positive cerebral malaria with age-matched controls, epilepsy, or other neurobehavioral sequelae were observed in nearly one-third of patients [60]. These observations suggest that enhanced seizure control may improve long-term outcomes.
Seizures may be generalized or focal, and the clinical signs may be subtle (nystagmus, irregular respirations, hypoventilation, or a drop in the Blantyre coma score). It is also important to evaluate for causes of seizure besides cerebral malaria (eg, hypoglycemia, fever) and to treat accordingly as outlined in the following sections.
Benzodiazepines are useful first-line agents for seizure treatment. Diazepam (0.4 mg/kg) can be administered intravenously (IV) or per rectum; lorazepam (0.1 mg/kg) can be administered IV or intraosseously. These doses can be repeated once if seizures do not cease within five minutes of the initial dose. Benzodiazepines should not be combined due to risk of respiratory depression. Intramuscular paraldehyde (2 mg/kg) is another option, especially when respiratory depression is a concern [5].
If seizures are not controllable with benzodiazepines or paraldehyde, other options in endemic areas include phenobarbitone (phenobarbital 15 to 20 mg/kg, slow IV push) or phenytoin (18 mg/kg diluted in 100 mL normal saline, infused over 20 minutes). A 20 mg/kg dose of phenobarbital should not be administered without respiratory support [5]. (See "Management of convulsive status epilepticus in children".)
If seizures recur, repeat single doses of benzodiazepine may be administered. Alternatively, in resource-limited settings, maintenance doses of phenobarbital (5 to 15 mg/kg/day, administered orally, via nasogastric tube, or via slow IV push in divided doses every 12 hours) or phenytoin (10 mg/kg/day IV in divided doses every 12 hours) may be initiated. Additional seizure management should consist of antiepileptics as discussed separately. (See "Management of convulsive status epilepticus in children".)
Routine seizure prophylaxis should not be administered in the absence of clinical seizure activity. In one study of 340 children with cerebral malaria randomized to receive phenobarbital (20 mg/kg) or placebo upon admission to hospital, the mortality among those treated with phenobarbital was significantly higher than that of the placebo group (18 versus 8 percent) [67].
Anemia and coagulopathy — Removal of infected and uninfected erythrocytes from the circulation is associated with rapid development of anemia. Patients with severe anemia may present with or without altered consciousness; in addition, severe anemia has been associated with long-term neurocognitive impairment [68]. In endemic areas, hemoglobin concentration may decrease gradually over the course of repeated malaria infections, so patients can be fully alert with hemoglobin concentrations of 2 to 3 g/dL (hematocrit <10 percent). Evaluation for pallor of the conjunctivae, nail beds, and palms can provide a rough estimate of the degree of anemia, since blood vessels in these areas are close to the surface (picture 2). (See "Anemia in malaria".)
Hemoglobin concentration and hematocrit are routinely measured components of complete blood counts, but this may not be available in resource-limited settings or the results may not be available in a timely manner. In such circumstances, the hematocrit can be measured on a fingerprick sample of blood collected into a heparinized capillary tube and centrifuged using a mechanical device (picture 3). Alternatively, the hemoglobin concentration can be determined from fingerprick samples of blood collected into cuvettes. This method is more expensive than "spinning a hematocrit" but can be performed readily near the bedside.
Transfusion — There is a lack of high-quality evidence on which to base transfusion guidelines [69]. In malaria-endemic areas, blood for transfusion is often in short supply and so transfusion should be reserved for patients with poor prognosis (eg, patients with altered consciousness, high-output heart failure, respiratory distress, cool extremities, hyperlactatemia, and/or high-density parasitemia).
In general, for high-transmission settings, transfusion is appropriate for children with hemoglobin <5 g/dL (hematocrit <15 percent); for low-transmission settings, transfusion is appropriate for patients with hemoglobin <7 g/dL (hematocrit <20 percent) [5]. These guidelines are based on small randomized trials [70,71], large observational studies [72-74], and expert opinion.
In general, 10 mL/kg of packed red blood cells or 20 mL/kg of whole blood transfused over two to four hours is appropriate. Blood should be typed and crossmatched prior to infusion. Monitoring of hemoglobin concentration or hematocrit should continue until the parasitemia clears, since repeat transfusion may be required. Blood transfusions are generally well tolerated in the setting of severe malaria, since patients are relatively hypovolemic; diuretics are rarely needed.
In areas where malaria and human immunodeficiency virus (HIV) are common, blood transfusion is associated with important risks. Most blood banks in endemic areas screen for HIV, hepatitis B, and syphilis, although such screening may miss HIV-infected donors who have not yet mounted a detectable serologic response to the infection.
Exchange transfusion — Exchange transfusion has been proposed as a means of removing infected red blood cells from the circulation, thereby lowering the parasite burden and replacing with unparasitized cells. There is no role for routine exchange transfusion as there is no evidence supporting its efficacy as adjunctive therapy in severe malaria, and there is no consensus on the indications, approach, benefits, or risks of this procedure [5,75,76].
The World Health Organization guidelines indicate that it is not possible to make any recommendations regarding the use of exchange transfusion based on the available evidence [5]. The United States Centers for Disease Control and Prevention (CDC) no longer recommends exchange transfusion for treatment of severe malaria [77], based on a review that demonstrated no differences in outcome among patients who underwent exchange transfusion [76]; previously, the CDC recommended exchange transfusion for patients with parasite density >10 percent and end-organ complications [78].
The American Society for Apheresis supports exchange transfusion as an adjunctive therapy for patients with severe malaria [79], although its consideration of adverse events associated with exchange transfusion for malaria is limited.
Coagulopathy — Clinically evident disseminated intravascular coagulation (DIC) in the setting of severe malaria is rare (<5 percent), but profound thrombocytopenia is common, and the microcirculation in many organs is occluded by fibrin thrombi [51]. The approach to DIC is discussed in detail separately. (See "Evaluation and management of disseminated intravascular coagulation (DIC) in adults" and "Disseminated intravascular coagulation in infants and children".)
Acute kidney injury — Renal dysfunction is increasingly recognized as a feature of complicated malaria in adults and children; it is an independent predictor of a poor outcome [7]. Elevated concentrations of extracellular hemoglobin in the setting of intravascular hemolysis may be associated with oxidative renal tubular damage.
Use of adjunctive acetaminophen in adults with moderate or severe malaria may be renoprotective. In a randomized trial including 62 adults in Southeast Asia with moderate or severe malaria treated with acetaminophen (1 g by enterally every 6 hours for 72 hours) or no acetaminophen, use of acetaminophen was associated with reduction in creatinine (median 23 versus 14 percent), particularly in the setting of prominent intravascular hemolysis (median 37 versus 14 percent) [80].
Fluids and nutrition
Hypoglycemia — Hypoglycemia is a common complication of malaria and a marker of severe disease; it should be suspected in any patient who is comatose or who deteriorates suddenly [81,82].
Hypoglycemia is defined as glucose <2.2 mmol/L; the thresholds for intervention are [5]:
●Children <5 years: glucose <3 mmol/L (<54 mg/dL)
●Children ≥5 years and adults: glucose <2.2 mmol/L (<40 mg/dL)
Clinical manifestations of hypoglycemia include seizure and altered consciousness, although these are not reliable clinical indicators, and blood glucose concentration should be assessed as part of routine evaluation. In resource-limited settings, this may be performed via fingerprick samples of whole blood on indicator strips (with or without handheld glucometers).
Hypoglycemic patients should have intravenous access established promptly, followed by administration of initial bolus of dextrose (0.25 g/kg of body weight), which is usually achieved with 2.5 mL/kg of 10 percent dextrose solution, since extravasation of higher concentrations of glucose can cause severe tissue damage. Blood glucose measurement after 15 minutes should be repeated, with administration of repeat boluses until the patient is normoglycemic. If glucose measurement is not possible, comatose patients with parasitemia at the time of initial assessment should receive a bolus of 2.5 mL/kg of 10 percent dextrose solution. (See "Approach to hypoglycemia in infants and children".)
Maintenance intravenous fluids should contain at least 5 percent dextrose; patients with recurrent hypoglycemia should receive 10 percent dextrose (10 percent dextrose can be prepared quickly by withdrawing 100 mL from a 1 liter bag of a 5 percent dextrose solution and replacing it with 100 mL of a 50 percent dextrose solution).
Patients presenting with normoglycemia can develop hypoglycemia during the course of treatment. In addition, those managed promptly for hypoglycemia at presentation can have subsequent recurrent hypoglycemia. Therefore, blood glucose should be monitored closely during the course of illness with prompt management as outlined above.
Low blood glucose is associated with a worse prognosis. In a prospective study of 437 children in Mali with presumed severe malaria (85 percent of whom had microscopic evidence of P. falciparum infection), a significant difference in admission glucose concentration was observed between those who died and those who survived (median 4.6 versus 7.6 mmol/L, p<0.001) [83]. Children with initial blood glucose concentrations <2.2 mmol/L were classified as hypoglycemic, those between 2.2 and 4.4 mmol/L were classified as low glycemia, those between 4.4 and 8.3 mmol/L were categorized as normoglycemia, and those with concentrations >8.3 mmol/L were classified as hyperglycemia. The case-fatality rates for hypoglycemia, low glycemia, normal glycemia, and hyperglycemia were 61, 46, 13, and 7 percent, respectively. The adjusted odds ratio was 0.75 (0.64 to 0.88) for case fatality for each 1 mmol/L increase in admission blood glucose concentration, suggesting that the traditional cutoff of 2.2 mmol/L may be too low for adequate sensitivity in the setting of pediatric cerebral malaria.
The pathogenesis of hypoglycemia in malaria is not fully understood; it may be related to parasite glucose consumption and/or impaired host gluconeogenesis [82]. Malnutrition, adrenal insufficiency, and hyperinsulinemia are not likely causes of hypoglycemia [82]. In addition to primary hypoglycemia, administration of quinine or quinidine (insulin secretagogues) can cause iatrogenic hypoglycemia [84,85]. Hypoglycemia with artesunate therapy is less common than with quinine or quinidine [10].
Volume management — The intravascular volume status in the setting of severe malaria is uncertain; there are data to both support and refute the presence of hypovolemia in the setting of severe malaria infection [86,87]. Adults with malaria appear to be more vulnerable to fluid overload than children; there is a thin line between underhydration (and thus worsening renal impairment) and overhydration (and risking pulmonary and cerebral edema) [88]. Therefore, fluid requirements should be assessed on an individual basis. Reliable markers of intravascular volume depletion in patients with severe malaria include cool peripheries, delayed capillary refill, low venous pressure, and low urine output. Deep breathing (reflecting lactic acidosis) may also be a reasonable indicator of hypovolemia.
Clinical symptoms of hypovolemia frequently resolve with blood transfusion (when warranted). When transfusion is not indicated, data suggest that repeated boluses of normal saline or albumin may be counterproductive [89]. In a large study including more than 3000 African children with severe infection and impaired consciousness and/or increased work of breathing as well as evidence of impaired perfusion (over half of whom had malaria), patients who received saline or albumin bolus had higher mortality at 48 hours than patients who received no bolus (10.6 and 10.5 percent versus 7.3 percent, respectively) [89]. In an observational study among 154 adults with severe malaria, restrictive fluid management (2 to 3 mL/kg per hour of crystalloid for the first 24 hours) did not worsen kidney function, tissue perfusion, or blood pressure; development of pulmonary edema was not related to the volume of fluid administration [90].
Some data suggest that volume resuscitation with crystalloid is favorable over colloid, although it is uncertain whether crystalloid increases risk of exacerbating cerebral edema in the setting of a fragile blood-brain barrier [91,92]. In resource-challenged settings, crystalloids are generally more readily available and are widely used.
In the setting of acute renal failure, institution of renal replacement therapy is appropriate if feasible. Hemofiltration is associated with lower mortality than peritoneal dialysis [93]; there have been no comparative trials of hemofiltration and hemodialysis. (See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose".)
Issues related to fluid management in children in resource-limited settings are discussed further separately. (See "Initial management of shock in children", section on 'Resource-limited settings'.)
Maintenance fluids — Maintenance intravenous fluids should include 5 percent dextrose. The rate of fluid administration should be determined by weight [94]; intravenous fluid support should be continued until oral intake is tolerated. (See "Maintenance intravenous fluid therapy in children".)
●Patients weighing <10 kg should receive fluids at a rate of 4 mL/kg/hour.
●Patients weighing 10 to 20 kg should receive 40 mL per hour for the first 10 kg, plus 2 mL/kg/hour for each kg above 10 kg.
●Patients weighing 20 to 80 kg should receive 60 mL per hour for the first 20 kg, 1 mL/kg/hour for every kg above 20 kg; total maintenance fluids are generally capped at approximately 2.5 liters daily.
●Patients weighing more than 80 kg should receive approximately 2.5 liters daily.
Most hospitals in endemic areas do not have infusion pumps, and intravenous fluids are usually available in one liter bags only. Therefore, fluid delivery can be difficult to monitor and there is real risk of iatrogenic volume overload. To prevent this problem, interposition of a burette (or "drip chamber") between the patient and the bag may be useful for monitoring intravenous fluid administration. A chamber is filled with a known amount of fluid (generally two hours' worth), labeled with tape, and the drip rate is set. In this way, providers can tell at a glance if the infusion is proceeding as scheduled.
Nutrition — Nutritional supplementation should be provided by nasogastric tube for patients with prolonged coma who are unable to eat and drink within 24 to 48 hours. (See "Overview of enteral nutrition in infants and children".)
In most endemic areas, no commercially prepared enteral products are available; substitutes (such as "eggnog" containing eggs, milk, sugar, and oil) or high-calorie drinks may be used. The volumes calculated for intravenous fluids can be administered via nasogastric tube; the intravenous fluid infusion rate should be decreased accordingly. Most patients are able to eat and drink within five to seven days. (See "Oral rehydration therapy".)
Fever — High fever (>38.5ºC) is common in the setting of malaria infection and may reflect the host response to endogenous pyrogens released at the time of schizont rupture [95]. The optimal approach to treatment of fever is uncertain, although use of antipyretics in patients with high fever is appropriate given the association between high fever and convulsions [5]. Aggressive temperature control may help reduce long-term neurologic outcomes in pediatric patients with retinopathy-positive cerebral malaria [60].
Acetaminophen (15 mg/kg every six hours; maximum dose 1000 mg) is a reasonable antipyretic agent; oral therapy can be used for patients able to swallow. Otherwise, suppository formulations are acceptable [5]. If fever persists, ibuprofen (10 mg/kg every six hours; maximum dose 1200 mg per day) can be administered (orally, via nasogastric tube, or IV) alone or on an alternating schedule with paracetamol every three hours.
Concomitant infection
HIV coinfection — Data on the effect of HIV infection on severe malaria infection are conflicting [96,97]. In one study from Thailand including 74 HIV-infected children with severe malaria, HIV infection was associated with higher parasite burden, more complications, and higher fatality rate [97]. In another study from Malawi including 126 HIV-infected children with severe malaria, HIV infection did not affect parasite density or outcome [96].
In general, the approach to treatment of malaria is the same in HIV-infected and HIV-uninfected patients. Potential drug interactions between antimalarials and antiretroviral drugs should be reviewed carefully [5,98,99].
Bacterial infection — Bacterial infection should be suspected in patients with signs or symptoms of sepsis (hypotension, cool extremities, delayed capillary refill, hyperlactatemia). Blood cultures should be obtained and empiric antibiotic therapy with activity against gram-negative bacilli is appropriate for these patients until bacterial infection has been excluded [5,100].
Concomitant bacteremia may contribute to morbidity and mortality in the setting of severe malaria, particularly in endemic areas. Severe anemia has been implicated as a primary risk factor for invasive bacterial infection [101], most commonly with nontyphoidal Salmonella [102-104]. A study of bacteremia in Kenya identified P. falciparum infection as a major risk factor for bacteremia with multiple organisms [104]. When the community parasite prevalence rate was 29 percent, 62 percent of bacteremia cases were attributable to malaria; the prevalence of bacteremia declined in parallel with the prevalence of malaria infection. A large review of cases of imported malaria noted that community-acquired bacterial coinfection occurred in 8 percent of patients; 2 percent had gram-negative bacteremia [105]. (See "Nontyphoidal Salmonella: Gastrointestinal infection and carriage".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Malaria".)
SUMMARY AND RECOMMENDATIONS
●Severe malaria is acute malaria with major signs of organ dysfunction and/or high level of parasitemia. In endemic areas, young children and pregnant women are at high risk for severe malaria. Older children and adults develop partial immunity after repeated infections and therefore are at relatively low risk for severe disease. Travelers to areas where malaria is endemic generally have no previous exposure to malaria parasites and so are at high risk for severe disease. (See 'Definition' above.)
●For patients in rural endemic areas with suspected severe malaria who cannot begin intravenous therapy immediately, we recommend administration of prereferral treatment (Grade 1A). (See 'Prereferral treatment' above.)
●Initial treatment of severe malaria consists of parenteral therapy. We suggest intravenous artesunate (in areas where intravenous artesunate of reliable quality is readily available) rather than intravenous quinine (table 5) (Grade 2A). This approach is warranted for adults and children (including infants, pregnant women in all trimesters, and lactating women). (See 'Initiating therapy' above.)
●Following administration of parenteral therapy (for at least 24 hours, until oral medication can be tolerated, and parasitemia ≤1 percent in nonimmune patients), an oral regimen should be administered (See 'Completing therapy' above.)
•Artemisinin combination therapy (ACT) regimen (three-day course) is preferred; if an oral ACT is not available, an alternative regimen may be used. ACT regimens, alternative regimens, and dosing are summarized in the table (table 6).
•Quinine or quinidine may be used if artemisinins are not available. The total duration of quinine/quinidine therapy is three days (for infection acquired in South America or Africa) to seven days (for infection acquired in Southeast Asia), administered in conjunction with an additional agent (clindamycin, doxycycline, or tetracycline) (table 5 and table 6). (See 'Quinine/quinidine' above.)
●For treatment of severe P. falciparum malaria during pregnancy in any trimester, we suggest treatment with intravenous artesunate (Grade 2B). Among the alternative agents, artemether is preferred over quinine. Following administration of parenteral therapy (for at least 24 hours and until oral medication can be tolerated), an oral regimen should be administered. ACT is preferred; if an oral ACT is not available, an alternative regimen may be used. Regimens and dosing are summarized in the table (table 7). (See 'Pregnancy' above.)
●Supportive care is critical for patients with severe malaria; death can occur within hours of presentation. Prompt assessment and initiation of antimalarial therapy are essential, followed by concurrent supportive care to manage life-threatening complications of the disease.
•Pulmonary complications of severe malaria include pulmonary edema, acute respiratory distress syndrome, and lower respiratory tract infection. Management requirements may range from supplementary oxygen to mechanical ventilation. (See 'Respiratory status' above.)
•Neurologic complications include altered sensorium, seizure, and coma. Clinical evaluation includes full physical examination, calculation of Blantyre coma score (table 2), funduscopic exam, and lumbar puncture. Seizures should be managed as outlined above. (See 'Neurologic status' above.)
•Hematologic complications include severe anemia and coagulopathy. Decisions regarding transfusion should be tailored to individual patient circumstances. (See 'Transfusion' above.)
•Hypoglycemia is a common complication of malaria and a marker of severe disease; it should be suspected in any patient who deteriorates suddenly. The threshold for intervention among children <5 years is <3 mmol/L (<54 mg/dL); the threshold for intervention among children ≥5 years and adults is <2.2 mmol/L (<40 mg/dL). Hypoglycemic patients should have intravenous access established promptly, followed by administration of initial bolus of dextrose (0.25 g/kg of body weight), which is usually achieved with 2.5 mL/kg of 10 percent dextrose solution. Blood glucose measurement after 15 minutes should be repeated, with administration of repeat boluses until the patient is normoglycemic. (See 'Hypoglycemia' above.)
•Hypovolemia should be assessed on an individual basis. Adults with malaria appear to be more vulnerable to fluid overload than children; there is a narrow threshold between underhydration (and risk of renal impairment) and overhydration (and risk of pulmonary and cerebral edema). (See 'Volume management' above.)
