|Year : 2013 | Volume
| Issue : 2 | Page : 51-56
The clinical, radiographic, and electrocardiographic correlates of childhood pericardial effusion diagnosed with echocardiography
Wilson E Sadoh1, Stanley U Okugbo2, Isoken A Isah1
1 Department of Child Health, Division of Cardiothoracic Surgery, University of Benin/University of Benin Teaching Hospital, Benin City, Edo State, Nigeria
2 Department of Surgery, Division of Cardiothoracic Surgery, University of Benin/University of Benin Teaching Hospital, Benin City, Edo State, Nigeria
|Date of Web Publication||13-Feb-2014|
Wilson E Sadoh
Department of Child Health, University of Benin Teaching Hospital, PMB-1111, Benin City, Edo State
Source of Support: None, Conflict of Interest: None
Background: Pericardial effusion (PE) is best detected with echocardiography. In settings where echocardiography is unavailable, other methods are relied on to diagnose PE.
Objective: To evaluate the clinical, radiologic, and electrocardiographic (ECG) correlates of PE in children with echocardiographically detected PE.
Methods: Children referred for echocardiography over a 4 year period at the University of Benin Teaching Hospital, were recruited if they also had PE. The clinical features, chest radiographic, and ECG findings were correlated with echocardiographically determined PE size using Spearman's correlation test. Statistical Package for Social Sciences (SPSS) 16 was used in the analysis of data.
Results: Of the 1,023 echocardiograms, 54 (5.2%) had PE. The mean age was 66.4 ± 53.4 months with 27 (50.0%) being males. PE was incidentally detected echocardiographically in 48 (88.89%) children. PE was mild, moderate, and severe in 12 (22.2%), 25 (46.3%), and 17 (31.5%), respectively. Of the 40 (74.1%) children managed in our center, 32 (80.0%) had heart failure. Four (10.0%) and six (15.0%) children, respectively had clinical tamponade and distant heart sound. Of the 15 cases with globular heart silhouette on chest radiograph, 12 (80%) had severe PE. Low voltage on ECG was seen in six (15.0%) children, who all had severe PE. Cases with distant heart sound, cardiomegaly on chest radiograph, and low voltages on ECG were significantly positively correlated with more severe PE, ρ, and (P-values), respectively were 0.49 (<0.0001), 0.54 (<0.0001), and 0.40 (0.005). There were five deaths of which two had shock from tamponade.
Conclusion: Childhood PE is commonly caused by diseases causing heart failure. Clinical evaluation, ECG, and radiographic findings are not sensitive in identifying mild and moderate PE, while mortality is high in cases of tamponade.
Keywords: Echocardiography, heart failure, pericardial effusion, tamponade
|How to cite this article:|
Sadoh WE, Okugbo SU, Isah IA. The clinical, radiographic, and electrocardiographic correlates of childhood pericardial effusion diagnosed with echocardiography. Nig J Cardiol 2013;10:51-6
|How to cite this URL:|
Sadoh WE, Okugbo SU, Isah IA. The clinical, radiographic, and electrocardiographic correlates of childhood pericardial effusion diagnosed with echocardiography. Nig J Cardiol [serial online] 2013 [cited 2020 Sep 23];10:51-6. Available from: http://www.nigjcardiol.org/text.asp?2013/10/2/51/127000
| Introduction|| |
Pericardial effusion (PE) is the presence of fluid within the pericardial sac. It occurs from a variety of diseases which could be infective or noninfective. The noninfective causes could be traumatic or nontraumatic.  In infective pericarditis, the commonly isolated organisms are , Mycobacteria tuberculosis and Staphylococcus aureus especially in resource limited settings.  In a study of infective pericarditis in Ibadan, Nigeria,  53 cases were recorded over a 10 year period. The predominant organisms isolated were M. tuberculosis and S. aureus in 11 cases. Viruses are important nonbacterial causes of pericarditis.  In noninfective cases, transudative PE has been reported to occur in heart failure, rheumatic fever, chronic renal failure, human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS), and malignancy. ,,
The clinical manifestations of PE depend on the amount and rapidity of fluid accumulation.  The spectrum of clinical presentation ranges from being asymptomatic in mild to moderate effusion to severe cardiovascular compromise as seen in cardiac tamponade due to severe effusion.  The clinical features include; fever, tachypnea, tachycardia, pulsus paradoxus, pericardial rub, distant heart sounds, and shock. Mild to moderate PE could escape clinical detection as they may be asymptomatic. Suspicion of PE on clinical ground may be hampered in cases of noninfective PE as they may not present with fever.  PE in patients without symptoms could build up and subsequently present with signs of cardiac tamponade. Cardiac tamponade occurs when the pericardial fluid is under pressure, thus compromising ventricular filling and its functions.  Cardiac tamponade is the most severe presentation of PE and its presentation is described as a continuum with mild to severe types and acute, subacute, and chronic forms of tamponade.  It may also be associated with a high mortality. In Ilesha, Nigeria Okeniyi  recorded a mortality of 62.5% among 15 children with cardiac tamponade from various causes.
Although PE may be detected by other modalities such as chest radiograph and ECG, echocardiography is the best method of detecting the presence of PE.  Echocardiography has enhanced the evaluation of severity and hemodynamic aberrations associated with PE including tamponade.  It also affords a better delineation of the underlying disease associated with PE. In resource limited settings where echocardiography may not be available; the clinical, ECG, and radiologic correlates of PE diagnosed with echocardiography becomes useful in identifying PE early. Besides, there is paucity of studies on childhood PE in Nigeria.
In this study; the clinical, chest radiographic, and ECG findings in children with PE are correlated with echocardiographic findings. The spectrum of underlying diseases complicated by PE was also documented.
| Materials and Methods|| |
Children referred for echocardiography between February 2009 and January 2013 at the Pediatric Cardiology Unit of the University of Benin Teaching Hospital, Benin City, Nigeria in whom PE was confirmed echocardiographically, were recruited for the study. The following data were retrospectively collected; age, gender, and indication for the echocardiography for all the children. The source of referral of the children were noted and categorized into those from other facilities and those from our facility.
The children from our facility were further evaluated for the presence of heart failure and clinical cardiac tamponade. Heart failure was determined using the clinical diagnostic criteria outlined below. 
- Significant tachycardia for age (>160 beats/min in infancy, >140 beats/min at 2 years, >120 beats/min at 4 years, and >100 beats/min above 6 years). Where fever was present, a 10 beats/min for every 1°C rise in temperature was allowed for.
- Significant tachypnea for age (> 60 cycles/min in the newborn, >40 cycles/min in <24 months, 30 cycles/min in 2-5 years, >28 cycles/min in 5-10 years, and >25 cycles/min in >10 years).
- Cardiomegaly (displaced apex beat with a central trachea or cardiothoracic ratio (CTR) >60% in <5 years and >50% in >5 years).
- Tender hepatomegaly of at least 3 cm size below the right costal margin.
The fulfillment of at least three of the four criteria above was diagnostic of heart failure.
Signs of suspected clinical cardiac tamponade were worsening or presence of new signs of tachycardia, tachypnea, and respiratory distress. The presence of distended neck veins, hypotension, pulsus paradoxus, and shock were also noted as signs of tamponade. Radiographic evaluation included measurement of the CTR on the chest radiograph. Cardiomegaly was defined as CTR > 60% in children younger than 5 years; while in children > 5 years, CTR > 50% was the criterion for cardiomegaly. Cardiac silhouette suggestive of enlargement of all the heart chambers, that is, globular or water-bottle silhouette was noted. The children with suspected cardiac disorder also had a 12 lead electrocardiogram using a portable Schiller AT-1, Switzerland. The ECGs were read according to the standard guidelines.  The ECG features of pericarditis/PE such as low voltages defined as QRS amplitude <5 mm in all the limb leads and electrical alternans were identified. ST segment elevation in limb and precordial leads which suggested pericarditis were similarly noted.
Echocardiogram was performed using an Aloka Prosound SSD-4000SV (Aloka, Meerbach, Germany). A trans thoracic echocardiography was done utilizing 2-dimensional (2D), M mode, and Doppler interrogation from various views. PE was diagnosed when there was an echo free space between the pericardium and cardiac wall in both systole and diastole on M mode. The largest size of the measured echo free space was taken as the dimension of the PE. The severity of the PE was graded using the dimension of the PE as follows; mild measured <5 mm, moderate was between 5 and 10 mm, while severe effusion was >10 mm and above, as was described in a previous report.  Echo free space diameter of >5 mm was considered pathologic.  The effects of the PE on the heart such as collapse of the atrial or ventricular walls were documented as well.
Congenital heart diseases were diagnosed echocardiographically in standard fashion.  Rheumatic heart disease (RHD) was diagnosed using the World Health Organization (WHO) 2006 echocardiographic criteria.  for case definition of RHD. Dilated cardiomyopathy (DCM)/myocarditis were diagnosed when there were dilated heart chambers with poor fractional shortening (FS) and ejection fraction on 2D echocardiography.  Myocarditis was further suspected when there was tachycardia unaccounted for by fever. Because of the echocardiographic similarity of DCM and myocarditis, and the absence of histological confirmation of myocarditis, DCM and myocarditis were grouped together. Hypertrophic cardiomyopathy was diagnosed when there was asymmetrical septal hypertrophy, septal dysorganization, and systolic anterior motion of the anterior mitral valve leaflet on M-mode. When the ratio between the septal and ventricular wall thickness in systole was >1.3 cm,  it was also suggestive. Cardiac rhabdomyoma was suspected on finding of a brightly echogenic mass within the left ventricular wall on echocardiography. It was confirmed to be rhabdomyoma on histology at autopsy.
The children with mild to moderate effusion without atrial or ventricular collapse were followed-up weekly until the effusion resolved spontaneously. Those with moderate to severe effusion with effects on the heart or those with clinical tamponade were drained. Children with massive PE were drained surgically, while the others had pericardiocentensis. Samples from the drained effusion were sent for microbiological analysis. Pericardial tissues were taken and also sent for histology in those that had surgical drainage. Patients with heart failure and those with persistent PE were treated with diuretics. Severe cases of heart failure also had digoxin. The outcome of the patients managed in our facility was noted. It was documented if they were discharged, discharged against medical advice or died.
Data were coded, entered into, and analyzed with Statistical Package for Social Sciences (SPSS) version 16.0. Chicago, IL. Simple proportions were presented in percentages, while continuous variables were presented as means ± standard deviation (SD). Differences in means were compared with Student's t-test or one-way analysis of variance (ANOVA). The cases with distant heart sounds, cardiomegaly on chest radiograph, and reduced voltages on ECG were correlated with echocardiographically determined PE size, using Spearman correlation test. Statistically significant P value was taken as ≤0.05.
| Results|| |
Characteristics of the study population
Of the 1,039 echocardiograms performed over the study period, 54 (5.2%) cases had PE. They consisted of 27 (50.0%) males and 27 (50.0%) females. The children were aged 1 week to 17 years with a mean age of 66.4 ± 53.4 months. Majority of the children 35 (64.9%) were aged <5 years, while the group aged between 5 years and <10 years and those ≥10 years were eight (14.8%) and 11 (20.4%), respectively. Forty (74.1%) of the studied children were managed in our facility, while 14 (25.1%) were referred for echocardiography from other centers.
Suspected PE was the reason for echocardiography in only six children, thus PE was an incidental finding in 48 (88.89%). The most common indications for echocardiography were suspected congenital heart disease in 28 (51.9%), half of whom (14, 25.9%) also had features of Down's syndrome, and cardiomegaly noticed on chest radiograph in nine (16.7%). The other indications for echocardiography are shown in [Table 1]. Of the 40 children managed in our facility, 32 (80.0%) either had or were on treatment for heart failure at the time of the echocardiogram
Clinical, radiographic, and electrocardiographic features of the study population
Of the 40 children managed in our facility, four (10.0%) only, had pericardial rub and presented with clinical tamponade, while two (50.0%) of the cases with cardiac tamponade presented in shock. Six (15.0%) cases had distant heart sound. The cases with distant heart sound were significantly correlated with severe PE, ρ = 0.49, P ≤ 0.0001. Chest radiographs were available for the 36 of the 40 children managed in our facility, of which 31 (86.1%) had cardiomegaly. Cardiomegaly was significantly correlated with severe PE; ρ = 0.54, P ≤ 0.0001. In almost half, 15 (41.7) of those with cardiomegaly, the cardiac silhouette looked globular. The remaining five (13.9%) children's chest radiograph had normal CTR.
There were six cases including the four with clinical tamponade that had low voltages on ECG. The presence of low voltage was also significantly correlated with severe PE; ρ = 0.40, P = 0.005. One of the six cases with low voltages also had electrical alternans. Two of the six cases had ST segment elevation.
[Table 2] shows the distribution of the underlying echocardiographically diagnosed diseases in the 54 cases. There were 11 (20.4%) cases of PE without an underlying condition on echocardiography. Of these 11 cases, three were pericarditis of uremic origin. Other types of acquired heart diseases such as RHD, DCM/myocarditis, and HCM constituted 23 (42.6%) of the series. The other diagnoses are shown in [Table 2]. Truncus arteriosus was the commonest cyanotic congenital heart disease in three (5.6%) cases. Of the 15 cases with globular cardiac silhouette on chest radiograph; 12 (80.0%) had severe PE, while the other three had DCM in two (13.3%) and RHD in one (6.7%).
|Table 2: Distribution of echocardiographically diagnosed cardiac conditions in children with pericardial effusion|
Click here to view
Of the 54 children, majority (25, 46.3%) had moderate PE, followed by those with severe PE (17, 31.5%), and mild PE (12, 22.2%). There were 20/54 (37.0%) cases with PE size between 1 and 5 mm whose PE could have been physiologic. With respect to severity of PE, the children managed in our facility were nine, 16, and 15, respectively with mild, moderate, and severe PE. The echocardiogram of one of the children with severe effusion showed strands within the pericardial space, which were confirmed to be fibrinous materials on pericardiostomy. Two others had visible particles within the pericardial space. The mean age of the children with mild, moderate, and severe PE were 39.1 ± 64.6, 27.4 ± 45.6, and 99.4 ± 67.9 months, respectively; the difference between the mean age was statistically significant, P = 0.007. Post hoc test, showed that the mean age of children with mild PE is significantly lower than those with severe PE, P = 0.05. The mean age of children with moderate PE was significantly lower than those with severe PE, P = 0.001. There is no gender difference in distribution of the severity of the PE, P = 0.77. [Table 3] shows the distribution of distant heart sound, low voltages, and cardiomegaly on chest radiograph with respect to the severity of PE.
|Table 3: The distribution of clinical, radiologic, and electrocardiographic features of 40 children by severity of pericardial effusion|
Click here to view
The FS of the children with heart failure 31.8 ± 7.0 was significantly lower than in those without heart failure 38.7 ± 5.3, P = 0.013 (CI = -12.3 to -1.52). Of the 32 cases with heart failure, acquired heart diseases; DCM/myocarditis, RHD, and HCM were the commonest cause of heart failure in 16 (50.0%) of cases. [Table 4] shows the distribution of heart failure by the underlying heart diseases. On evaluation of the effect on the heart by the PE, compression of the right atrium alone was seen in six (12.2%) cases. Compression of both right atrium and ventricle were demonstrated in nine (18.4%) cases. Most of the children with cardiac wall collapse were seen in the 17 cases of severe PE.
|Table 4: The distribution of underlying cardiac conditions associated with heart failure in 32 patients with pericardial effusion|
Click here to view
Outcome measures are presented for the 40 cases managed in our facility. Of 10 cases with severe effusion, nine had drainage. One child with severe PE that was planned to have pericardiocentesis was discharged against medical advice. Of the nine who had drainage of their PE; four (44.4%) cases had pericardiocentensis, while five (55.6%) required surgical drainage. There were five (12.5%) deaths, of which two were related to the PE. The two deaths were cases of chronic renal failure complicated by massive PE and presented with cardiac tamponade and shock. The PE in 14 (35.0%) cases resolved spontaneously over 2-10 weeks period. Of the 14 cases that resolved spontaneously, five had shown an initial increase in fluid accumulation before spontaneous resolution. The remaining seven cases were lost to follow-up. Of the nine PE cases that were drained, one (11.1%) case was confirmed Mycobacterium tuberculosis, two (22.2%) had Staphylococcus aureus, while no organism was isolated in the remaining six (66.7%).
| Discussion|| |
The wide spectrum of underlying cardiac diseases associated with PE in this study, is similar to that found in the study conducted by Kessler et al.,  in 1980. Unlike the Kessler study, HIV associated PE was seen in the present study. This has however been similarly described in more recent reports. , RHD continues to be a cause of morbidity amongst children in resource limited settings as documented in this study. We note a high number of cases with suspected Down syndrome. Down syndrome has been reported to be complicated by PE in those with severe hypothyroidism.  We could not however determine thyroid status in this study because of cost. The resolution of the PE in the children with Down syndrome without the administration of L-thyroxine may suggest that the Down syndrome cases we managed may not have had severe hypothyroidism. We however acknowledge the inability to determine the thyroid function tests as a limitation in this study.
Heart failure is a common cause of PE in this study as has similarly been reported by previous authors. , The heart failure could be a manifestation of the underlying disease such as in DCM, RHD ventricular septal defect, and myocarditis. It may also be a manifestation of the severe PE, especially those with cardiac tamponade.
The detection rate of PE by clinical, radiographic, and electrocardiographic assessment prior to echocardiographic evaluation in this study was low. There were only four cases with pericardial rub detected clinically. This low value is consistent with the findings from the Ibadan  study on infectious pericarditis in which only one case had pericardial rub. Shock seen in cardiac tamponade and heart failure are important pointers to the presence of PE in this study. Cardiac tamponade presenting in shock, is severe and is often associated with mortality as seen in this study. The two mortalities related to PE in this study presented with clinical tamponade and shock. A high index of suspicion is required in cases without clinical tamponade.
The chest radiographic findings which should complement the physical examination findings may only show a globular cardiac silhouette in cases of massive PE. From our study, PE is a common cause of globular cardiac silhouette, being responsible for a globular heart in 12/15 (80.0%) cases. Although PE is suspected in a chest radiograph showing globular heart with normal pulmonary vascular markings, other possibilities should be considered, such as DCM in younger children and RHD in older children as was found in this study. Low voltage on ECG caused by the pericardial fluid was seen in 40% of severe cases of PE. Chest radiograph and ECG which are relatively available investigative stools in resource limited setting, have a low detection rate of nonsevere PE as shown in this study.
The ability of echocardiography in detecting clinically undiagnosed PE is demonstrated in the incidental detection of PE in almost 90% of cases in this study. This is different from an earlier study where 70 of 171 (40.9%) cases were incidentally detected by echocardiography. The higher proportion of moderate and severe PE in the earlier study may have enabled more cases to be detected clinically as they would be more likely to be symptomatic in that study compared to the present study. The importance of non-detection of PE especially in settings without echocardiography, is that the mild to moderate cases may be at the earlier stages of accumulation of pericardial fluid and if left unattended, may increase in size and present in cardiac tamponade which could be fatal. Mortality from cardiac tamponade in this study was high which is similar to the report of Okeniyi  in Ilesa. Since cardiac tamponade is a continuum, echocardiographic finding of compression of the right atrium which is an early finding allows the clinician to be proactive in draining the effusion before more severe forms of tamponade ensues.
Resolution of mild to moderate PE in children has similarly been reported in a previous study.  In this study, almost half of the children had resolution of their PE, drawing attention to the fact that not all PE needs to be drained. The resolution of the PE may have been due to the diuretic the patients on heart failure were on. There were only three of the nine cases who had drainage in this study where organisms were isolated. Thus, infective pericarditis was recorded in a third of the cases investigated. The absence of bacteria in the other six cases could be due to possible viral infection or a transudate from cardiac failure which was common in this study. Viral pericarditis is a common cause of PE from previous reports. , We note the inability to perform viral studies as a limitation in this study. The isolation of M. tuberculosis and S. aureus in this study is similar to a previous report on PE in Nigeria. 
| Conclusion|| |
Heart failure was a common cause of PE in this study. Clinical, chest radiographic, and ECG evaluations were poor in detecting most cases of mild to moderate PE. Mortality was commonly seen in cases of cardiac tamponade. A high index of suspicion for PE is needed in the children with suspected cardiac diseases complicated by heart failure in settings where echocardiography is not available.
| References|| |
|1.||Rheuban KS. Pericardial disease. In: Allen HD, Gutgesell HP, Clark EB, Driscoll DJ, editors. Moss and Adams Heart Disease in Infants, Children and Adolescents. Including Fetal and Young Adult. Philadelphia: Lippincott Williams and Wilkins; 2001. p. 1287-96. |
|2.||Jaiyesimi F, Abioye AA, Antia AU. Infective pericarditis in Nigerian children. Arch Dis Child 1979;54:384-90. |
|3.||Zreik H, Li J, Garson Jr TA. Etiology and danger of pericardial effusion in infants and children. Cardiol Young 1996;6:162-5. |
|4.||Chen Y, Brennessel D, Walters J, Johnson M, Rosner F, Raza M. Human immunodeficiency virus-associated pericardial effusion: Report of 40 cases and review of literature. Am Heart J 1999;137:516-21. |
|5.||Spodick DH. Acute cardiac tamponade. N Engl J Med 2003;349:684-90. |
|6.||Khandaker MH, Espinosa RE, Nishimure RA, Sinak LJ, Hayes SN, Melduni RM, et al. Pericardial disease: Diagnosis and management. Mayo Clin Proc 2010;85:572-93. |
|7.||Okeniyi JA. Cardiac tamponade in Ilesa, Nigeria. S A J Child Health 2008;2:162. |
|8.||Pepi M, Muratori M. Echocardiography in the diagnosis and management of pericardial disease. J Cardiovasc Med (Hagerstown) 2006;7:533-44. |
|9.||Omokhodion SI. Childhood heart failure. In: Omokhodion SI, Osinusi K, editors. Pediatric Cardiology and Respiratology. WACP update series (West African College of Physicians, Lagos, Nigeria); 1996. p. 72-82. |
|10.||Park MK, Guntheroth WG. Basic measurements. In: Park MK, Guntheroth WG, editors. How to Read Pediatric ECGs. St Louis: Mosby Year Book; 1992. p. 10-32. |
|11.||Ahmad S, Ellis J, Nesbitt A, Molyneux E. Pericardial effusion in children with severe protein energy malnutrition resolve with therapeutic feeding: A prospective cohort study. Arch Dis Child 2008;93:1033-6. |
|12.||Snider AR, Serwer GA, Ritter SB. Echocardiography on pediatric heart disease. 2 nd ed. St. Louis: Mosby Publishing; 1997. |
|13.||Carapetis JR, Paar J, Cherian T. Standardization of epidemiologic protocols for surveillance of post-streptococcal sequala: Acute rheumatic fever, rheumatic heart disease and acute post-streptococcal glomerulonephritis. WHO Technical Report 2006. |
|14.||Kessler KM, Rodriguez D, Rahim A, Dheen M, Samet P. Echocardiographic observations regarding pericardial effusions associated with cardiac disease. Chest 1980;78:736-40. |
|15.||Okoromah CA, Ojo OO, Ogunkunle OO. Cardiovascular dysfunction in HIV-infected children in a sub-Saharan African country: Comparative cross-sectional observational study. J Trop Pediatr 2011;8:3-11. |
|16.||Anah MU, Ansa VO, Etiuma AU, Udoh EE, Ineji EO, Asindi AA. Recurrent pericardial effusion associated with hypothyroidism in Down syndrome: A case report. West Afr J Med 2011;30:210-3. |
|17.||Berger M, Bobak L, Jelvel M, Goldberg E. Pericardial effusion diagnosed by echocardiography. Clinical and elctrocardiographic findings in 171 patients. Chest 1978;74:174-9. |
|18.||Sagrista-Sauleda J, Angel J, Permanyer-Miralda G, Soler-Soler J. Long term follow-up of idiopathic chronic pericardial effusion. N Engl J Med 1999;341:2054-9. |
[Table 1], [Table 2], [Table 3], [Table 4]