The epidemiology and outcomes of pericardial effusion in hospitalized children: a national database analysis

The epidemiology and outcomes of pericardial effusion in hospitalized children: a national database analysis

Sasaki J, Sendi P, Hey MT, Evans CJ, Sasaki N, Totapally BR. Journal of Pediatrics 2022;249:29-34. doi: 10.1016/j.jpeds.2022.07.005. Epub 2022 Jul 11. PMID: 35835227


Take home points:

1)  KIDS inpatient database, authors found a prevalence for pericardial effusion of 0.1% in inpatients; pericardial drainage occurred in 12.3% of these patients

2) Top diagnoses in hospitalized patients with pericardial effusion were structural heart disease (40%), post cardiac surgery (28%), rheumatological diagnosis (15%), hematologic malignancy (11%), solid organ tumor (9%) and organ transplant (2.6%).

3) Mortality among patients with pericardial effusion was 6.8%; mortality was higher in younger patients with solid organ tumors and lower in older patients with cardiac or rheumatologic diagnoses.


Commentary from Dr. Thomas Zellers (Dallas, USA), section editor of Congenital Heart Disease Interventions and ACHD Journal Watch:

Summary:  The authors used the Healthcare Cost and Utilization Projects (HCUP) Kids inpatient database from the Agency for Healthcare Research and Quality (AHRQ). This is the largest and most comprehensive database for all payor pediatric inpatients. The authors performed a retrospective analysis of all hospital admissions for children with pericardial effusions using specific ICD-10  codes that would allow for diagnoses, pericardial drainage procedures and mortality. Clinical classification software (CCS) was also used to identify a broad range of etiological factors associated with pericardial effusion. They created 7 etiologic categories which included cardiac surgery, cardiac structural disease, organ transplant, hematologic malignancy, solid organ tumors, rheumatologic diagnoses and OTHER. Infectious etiologies could not be identified separately. The authors further separated the cohorts by age at the time of diagnosis: neonate (< 28 days), infant (> 28 days but < 1 year), preschool (1-5 years), child (6-12 years) and teen (13-20 years).

There were 6.26 million patients evaluated and 6417 (0.1%) were diagnosed as having a pericardial effusion. The prevalence was highest in infants (0.28%) and lowest in neonates (0.04%). The prevalence was twice as high in black children (0.15%) compared to white children (0.08%).

In children with pericardial effusion, the most common diagnoses were (in order) structural heart disease, post op cardiac surgery, rheumatologic diagnosis, hematologic malignancy, solid organ tumor and post organ transplant. Cardiac structural diagnoses were more prevalent in neonates and infants, whereas the other diagnoses were seen more frequently in older age groups.

Pericardial drainage occurred in 12.3% of the patients. Drainage occurred most frequently in teenagers and occurred least frequently in the 6-12 year old age range. Pericardial effusion drainage was required most commonly in post op cardiac surgery patients and those with hematologic malignancy. It was least frequent in patients with rheumatologic diagnoses.

Mortality occurred in 6.8% of children with pericardial effusion. Of those who required pericardial drainage, 10.8% died. Using an adjusted risk for mortality model, the adjusted risk for mortality was lowest in older age groups and highest in neonates. Cardiac structural disease and rheumatologic disease was associated with the lowest adjusted mortality risk; it was highest in patients with solid organ tumor. Other associated risks that increased risk adjusted mortality included acute respiratory failure, acute kidney injury, need for dialysis and need for pericardial drainage.

This paper represents one of the largest studies- using a single database- offering current insights into the etiologies of pericardial effusion in hospitalized patients. It is a retrospective study, however, and the study was not geared to explore the reasons for drainage or the nuances surrounding mortality. The idiopathic etiologies (often viral illness) were not evaluated except as an OTHER category and little information is given about this group of patients. There is also no information as to why a patient had an echocardiogram or a point of care ultrasound to diagnose the pericardial effusion nor any information about the qualifications of the person doing the ultrasound or the severity of the effusion. So, it is possible that the denominator may be over or underestimated.