Clinical predictors and outcomes of patients with pericardial effusion in chronic kidney disease.

Abstract

Pericardial effusion is common in hospitalized patients with chronic kidney disease (CKD). We sought to identify predictors and prognostic impact of pericardial effusion in CKD patients. Clinical and biochemical parameters can predict pericardial effusion in CKD patients. In a retrospective nested case-control design, we analyzed hospitalized adult patients with CKD stage 4, 5, or end-stage renal disease diagnosed with pericardial effusion. Controls were same-stage CKD patients without effusion. Among 84 cases and 61 controls, 44% and 34% were on dialysis, respectively. The mean creatinine was higher among cases versus controls (8.4±6.0 vs. 6.0±3.4 mg/dL, P = 0.002). Effusion was moderate to large in 46% of cases. Independent predictors of any pericardial effusion were serum potassium (OR: 1.95 per 1-mEq/L increment, 95% CI: 1.21-3.13, P = 0.006), serum corrected calcium (OR: 1.33 per 1-mg/dL decrement, 95% CI: 1.11-1.67, P = 0.015), and admission heart rate (OR: 1.29 per 10-bpm increment, 95% CI: 1.03-1.62, P = 0.027). Corrected calcium level was an independent predictor of moderate to large pericardial effusion (OR: 1.38 per 1-mg/dL decrement, 95% CI: 1.04-1.82, P = 0.023). Corrected calcium <8.0 mg/dL demonstrated 95% specificity for moderate to large effusion. Patients with effusion had no significant difference in the composite endpoint of mortality or cardiovascular rehospitalization (P = 0.408). In hospitalized CKD patients, hypocalcemia may be useful in identifying those with moderate to large pericardial effusion. In this population, pericardial effusion does not seem to be associated with adverse outcomes.