Studies have proposed that malaria may lead to electrocardiographic (ECG) changes and pericardial inflammation. We aimed to investigate the frequency of ECG alterations, determined by ECG and Holter monitoring, and pericardial effusion in patients with malaria infection. We performed a prospective observational study of adult patients with uncomplicated malaria in Amazonas, Brazil. Peripheral blood smears, ECG, and bedside echocardiography were conducted before antimalarial treatment and repeated at follow-up after completed treatment. We evaluated the diagnostic value of PR-segment depression, PR-segment elevation, and Spodick’s sign for detecting pericardial effusion. A subset of patients underwent Holter monitoring at baseline. Among 98 cases of uncomplicated malaria (55% men; mean age 40 years; median parasite density 1,774/µl), 75 had Plasmodium vivax , 22 Plasmodium falciparum , and 1 had mixed infection. At baseline, 17% (n = 17) had PR-segment depression, 12% (n = 12) PR-segment elevation, 3% (n = 2) Spodick’s sign, and the prevalence of pericardial effusion was 9% (n = 9). ECG alterations had sensitivities of 22% to 89% and specificities of 88% to 100% for detecting pericardial effusion at baseline. PR-segment depression had the best accuracy (sensitivity 89%, specificity 90%). Of the 25 patients, 4 patients who did not have pericardial effusion, displayed nonsustained ventricular tachycardia, determined by Holter monitoring (median duration 43 hours). Follow-up examination data were obtained for 71 patients (median 31 days), for whom PR-segment depression, elevation, and pericardial effusion had reduced significantly (p <0.05). In conclusion, our findings suggest that ECG alterations may be useful to detect pericardial effusion in malaria and that these findings decrease after completed antimalarial treatment.
Malaria affects more than 220 million people worldwide each year, and common complications involve anemia, cerebral malaria, hypoglycemia, and acute kidney injury. Studies have also reported that malaria may be able to affect the heart. One study found electrocardiographic (ECG) alterations in up to 23% of hospitalized patients infected with Plasmodium falciparum , and 2 studies reported pericardial effusion in malaria. , Pericardial effusion often appears owing to inflammation associated with pericarditis, and timely identification is important for treatment because of risk of progression to cardiac tamponade. Owing to sparse availability of ultrasound imaging in malaria endemic areas, pericardial effusion and pericarditis may be underdiagnosed, and ECG is typically the most common diagnostic tool. Consequently, we aimed to determine the frequency of ECG alterations and pericardial effusion in patients with uncomplicated malaria in the Brazilian Amazon Basin. Our hypothesis was that ECG alterations yield diagnostic value for detecting pericardial effusion in malaria.
Methods
The study was conducted in the municipality of Cruzeiro do Sul, Acre, located in the Western part of the Brazilian Amazon rain forest. The municipality has an area of 8,783 km 2 , and a population of approximately 80,000 people. The climate is tropical, with an average annual temperature of 25°C. The municipality has one of the highest malaria transmission burdens in the Americas, with an annual parasite index >10 and nearly 15,000 cases reported in 2015. Transmission occurs throughout the year. Three-fourths of cases are of Plasmodium vivax , and one-fourth of cases are of Plasmodium falciparum .
This was a prospective, observational study of 102 adults with uncomplicated malaria included from June to December 2020 ( ClinicalTrials.gov : NCT04445103). Patients were examined in community health care clinics and at a free-standing emergency care clinic in Cruzeiro do Sul, Acre, Brazil (Unidade Pronto Atendimento). Inclusion criteria were clinical symptoms and a positive peripheral blood smear of any Plasmodium species, regardless of parasite density. Exclusion criteria were pregnancy, <18 years old, unable to complete the examination, suspected concomitant infection based on examination by a physician, intake of antimalarial medication for the current infection, or known chronic pericardial effusion. After exclusion (n = 4), a total of 98 patients were included ( Figure 1 ). A convenience sample of 25 patients with malaria underwent Holter monitoring (Cortrium, C3+). A single follow-up examination was scheduled for patients with malaria approximately 30 days after inclusion and involved repetition of all data collection procedures. If patients could not participate on the follow-up date, this was rescheduled to be as close as possible to the original date. By follow-up, information was obtained on relapse of malaria since the inclusion date and whether symptoms of malaria had improved.
Age- and gender-matched controls without known heart disease (n = 98; ratio 1:1) were included from the same health care clinics. Controls were eligible to participate if they did not have malaria, verified by negative peripheral blood smears. The same exclusion criteria as mentioned previously was applied. For comparison of Holter findings, a convenience sample of 18 unmatched controls with clinical indication for Holter monitoring were included. Indications included daily palpitations at rest, syncope ≥2 times within 6 months, and a history of stroke.
Participation in the study involved answering a structured questionnaire administered by an interviewer in Portuguese, measurements of height and weight, blood samples, peripheral blood smear, ECG, and echocardiography. Details on blood sampling, laboratory procedures, and microscopy are displayed in Supplementary Methods. In addition, all patients underwent a full physical examination by a physician (PB) and were assessed for symptoms of severe malaria. Hypertension was defined as a history of hypertension and/or intake of antihypertensive medication. We collected information on QT-prolonging medication and cardiotoxic drugs ( Supplementary Table 1 ).
A 12-lead ECG (Cardioline Instruments, TouchECG+) was recorded and analyzed offline by 2 trained investigators, blinded to clinical data. Time intervals were manually counted and, based on studies assessing ECG and pericardial effusion, the following alterations were assessed: (1) PR-segment depression >0.5 mm (leads: I, II, aVL, aVF, and V 2 to V 6 ), (2) PR-segment elevation >0.5 mm (lead: aVR), (3) Spodick’s sign (all leads), and (4) electrical alternans >5 mm. Each of these criteria were considered positive if present in ≥2 leads. According to contemporary guidelines, we assessed for ST-segment elevation and depression >1 mm. QTc was assessed by the Bazett and Fridericia formulas. Holter data were analyzed according to contemporary guidelines , using Cubeholter (Cardioline Instruments, Trento, Italy) by a single investigator (PB), blinded to malaria status and clinical data. Nonsustained ventricular tachycardia was defined as ≥3 consecutive ventricular beats at heart rate >110 beats per minute for <30 seconds. Additional information on classification of Holter events are listed in Supplementary Table 2 . All examinations were included in the analyses regardless of monitoring period.
Echocardiographic examinations were performed at bedside (VIVID IQ, GE Healthcare, Horten, Norway) by a single investigator (PB) to ensure high-quality echocardiograms. Examinations were analyzed offline (EchoPac203, GE Vingmed) by a single investigator (AW), blinded to malaria status and clinical data. Left ventricular ejection fraction (LVEF) was assessed by Simpson’s biplane method in the apical 2- and 4-chamber views. Pericardial effusion was examined and quantified in the subcostal view and considered present when ≥0.5 cm in width. Patients with pericardial effusion were evaluated by a physician for signs of cardiac tamponade.
First-line standard malaria treatment was administered on the day of diagnosis and inclusion into the study after baseline data collection. Patients with malaria caused by Plasmodium vivax were treated with chloroquine (3 days) and primaquine (7 days), and those with malaria caused by Plasmodium falciparum were treated with artemether and lumefantrine (3 days) and a single dose of primaquine. All dosages were individually administered based on weight. One patient with mixed infection ( Plasmodium vivax and Plasmodium falciparum ) was treated with artemether and lumefantrine (3 days) and primaquine (14 days).
The study was approved by the institutional review committees at Federal University of Acre and University of São Paulo (CAAE: 26552619.6.0000.510 and 32947520.4.0000.5467), local health care authorities, and leaders of health care clinics. The study complies with the second Declaration of Helsinki, and all patients provided written informed consent in Portuguese after receiving written and oral information about the study. Illiterate participants provided fingerprints on consent forms, which were signed by 2 independent witnesses.
Two-sided p values <0.05 were considered statistically significant. Histograms and Q-Q plots were used to assess the distribution of continuous variables. Changes in variables between baseline and follow-up ( Table 1 ) were assessed using Student’s paired t test, Wilcoxon signed-rank test, or McNemar test as appropriate. Variables in cases and controls ( Supplementary Tables 3 to 5 ) were compared using analysis of variance, Wilcoxon rank-sum, and chi-square test as appropriate. The relation between pericardial effusion (categorical variable) and LVEF was examined by linear regression models, and the relation with parasite density by negative binomial regression models. We followed this approach, as parasite density represents highly dispersed count data. Both models were unadjusted. Sensitivity, specificity, and positive and negative predictive values for electrocardiography to detect pericardial effusion were calculated using the diagt command in STATA, in which the appropriate prevalence level was chosen. Statistical analyses were performed in STATA SE version 14.2 (StataCorp, College Station, Texas).
All patients (n = 98) | Completed follow-up (n = 71) | |||
---|---|---|---|---|
Variable | Baseline | Baseline | Follow-up | p * |
Demographics | ||||
Age (years) | 40 ± 15 | 40 ± 15 | ||
Men | 54 (55%) | 38 (54%) | ||
Body mass index (kg/m 2 ) | 26 ± 5 | 26 ± 5 | ||
Present smoker | 40 (41%) | 27 (38%) | ||
Hypertension | 20 (20%) | 13 (18%) | ||
Family income (Brazilian reals) | 1,580 (1,000 to 2,500) | 1,540 (1,000 to 2,000) | ||
Education | ||||
No formal education | 12 (12%) | 8 (11%) | ||
Primary school | 24 (25%) | 21 (30%) | ||
Secondary school | 42 (43%) | 27 (38%) | ||
Higher academic | 20 (20%) | 15 (21%) | ||
Clinical | ||||
Chest pain | 13 (15%) | 10 (14%) | 0 | <0.001 |
Dyspnea | 3 (3%) | 3 (4%) | 1 (1%) | 0.001 |
Systolic blood pressure (mm Hg) | 123 ± 17 | 123 ± 17 | 127 ± 16 | 0.04 |
Diastolic blood pressure (mm Hg) | 76 ± 11 | 76 ± 10 | 79 ± 11 | 0.01 |
Heart rate (bpm) | 89 ± 18 | 86 ± 18 | 75 ± 12 | <0.001 |
Axillary temperature (°C) | 36.7 ± 0.9 | 36.6 ± 0.8 | 36.4 ± 0.4 | 0.05 |
Microscopy | ||||
Plasmodium species vivax positive falciparum positive vivax + falciparum positive | 75 (77%) 22 (22%) 1 (1%) | 52 (73%) 19 (26%) 1 (1%) | 2 (3%) | |
Parasite density (µl −1 ) | 1,774 (527 to 6,591) | 2,096 (526 to 6,585) | 128 (80 to 176) | |
Electrocardiogram | ||||
PR (ms) | 144 ± 17 | 144 ± 18 | 146 ± 21 | 0.17 |
QRS (ms) | 88 ± 10 | 88 ± 8 | 88 ± 9 | 0.11 |
QT (ms) | 369 ± 32 | 373 ± 32 | 392 ± 29 | <0.001 |
QTc Bazett (ms) | 443 ± 30 | 443 ± 31 | 435 ± 29 | 0.08 |
QTc Fridericia (ms) | 416 ± 24 | 418 ± 25 | 420 ± 26 | 0.81 |
PR-segment depression | 17 (17%) | 15 (21%) | 2 (3%) | <0.001 |
PR-segment elevation | 12 (12%) | 10 (14%) | 3 (4%) | 0.008 |
Spodick’s sign | 2 (2%) | 2 (3%) | 0% | 0.16 |
Paraclinical | ||||
C-reactive protein (mg/L) | 34 ± 26 | 33 ± 28 | 2 ± 5 | <0.001 |
Hemoglobin (g/dL) | 13.9 ± 1.3 | 13.9 ± 1.5 | 13.9 ± 1.4 | 0.51 |
Leukocytes (mm 3 ) | 5,526 ± 2,158 | 5,492 ± 2,223 | 6,296 ± 1,766 | 0.008 |
Reticulocytes | 0.8 ± 0.2 | 0.8 ± 0.2 | 0.8 ± 0.2 | 0.76 |
Platelets (× 10 3 /mm 3 ) | 132 ± 72 | 129 ± 68 | 207 ± 61 | <0.001 |
Creatinine (mg/dL) | 0.97 ± 0.25 | 0.94 ± 0.20 | 0.90 ± 0.21 | 0.09 |
Bilirubin total (mg/dL) | 0.83 ± 0.52 | 0.92 ± 0.52 | 0.42 ± 0.18 | <0.001 |
Sodium (mmol/L) | 139.4 ± 3 | 139.6 ± 3 | 141.2 ± 3 | 0.003 |
Potassium (mmol/L) | 4.5 ± 0.7 | 4.5 ± 0.7 | 4.7 ± 0.4 | 0.05 |
Blood glucose (mg/dL) | 109 ± 34 | 107 ± 26 | 104 ± 17 | 0.44 |
Prothrombin time (s) | 13.1 ± 1.3 | 13.1 ± 1.3 | 12.9 ± 1.4 | 0.49 |
International normalized ratio | 1.03 ± 0.1 | 1.02 ± 0.1 | 1.01 ± 0.11 | 0.50 |
Lactate dehydrogenase (U/L) | 313 ± 103 | 319 ± 106 | 235 ± 75 | <0.001 |
Echocardiography | ||||
Pericardial effusion | 9 (9%) | 9 (13%) | 2 (3%) | 0.008 |
Left ventricular ejection fraction (%) | 54 ± 6 | 53 ± 6 | 54 ± 5 | 0.3 |