Diuretic Response Prediction With MELD Score in Heart Failure

BACKGROUND: Heart failure (HF) is still an important disease with high mortality rates worldwide. HF treatment is also difficult due to different phenotypes. Diuretic response (DR) is one of the main differences across these subgroups. Novel urinary parameters are used for DR prediction. We sought to determine whether the MELD score could be used as an additional parameter for predicting the DR. METHODS: Eighty-one consecutive patients diagnosed with decompensated HF between June and October 2020 were included. The second hour urine sodium (UNa) level after the first intravenous diuretic administration and serum parameters were recorded. All patients underwent a comprehensive echocardiographic examination. MELD score derivatives were tested to assess the DR. RESULTS: A total of 81 patients (mean age: 66.4 ± 13.5; mean ejection fraction: 29.6 ± 12.7%) were divided into two groups according to UNa. 26 (32%) patients had poor DR. MELD Na score was independently associated with DR (OR = 0.88 -0.21 to (-0.03); p = 0.008). Furthermore, MELD Na score was correlated with urinary sodium (r = -0.354; p = 0.004). Daily furosemide dose was higher (237.9 ± 204.7 vs. 129.3 ± 83.5 mg; p = 0.001) and length of hospital stay was longer (15.6 ± 10.8 vs. 8.5 ± 6.1 days; p < 0.01) in the low UNa group. CONCLUSION: The MELD score derivative was associated with DR according to urinary sodium and may be used as an additional parameter to predict the DR.