Rakhi Maiwall1, Suman Kumar2, Manoj Kumar Sharma1, Zeeshan Wani1,3, Mulu Ozukum1, Shiv Kumar Sarin1. 1. Departments of Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India. 2. Department of Hematology, Command Hospital (Eastern Command), Kolkata, India. 3. Department of Gastroenterology, Jawaharlal Nehru Hospital, Srinagar, Kashmir, India.
Abstract
BACKGROUND: The prevalence and clinical significance of hyponatremia in cirrhotics have been well studied; however, there are limited data on hyperkalemia in cirrhotics. AIM: We evaluated the prevalence and prognostic significance of hyperkalemia in hospitalized patients with cirrhosis and developed a prognostic model incorporating potassium for prediction of liver-related death in these patients. METHODS: The training derivative cohort of patients was used for development of prognostic scores (Group A, n = 1160), which were validated in a large prospective cohort of cirrhotic patients. (Group B, n = 2681) of cirrhosis. RESULTS: Hyperkalemia was seen in 189 (14.1%) and 336 (12%) in Group A and Group B, respectively. Potassium showed a significant association that was direct with creatinine (P < 0.001) and urea (P < 0.001) and inverse with sodium (P < 0.001). Mortality was also significantly higher in patients with hyperkalemia (P = 0.0015, Hazard Ratio (HR) 1.3, 95% confidence interval 1.11-1.57). Combination of all these parameters into a single value predictor, that is, renal dysfunction index predicted mortality better than the individual components. Combining renal dysfunction index with other known prognostic markers (i.e. serum bilirubin, INR, albumin, hepatic encephalopathy, and ascites) in the "K" model predicted both short-term and long-term mortality with an excellent accuracy (Concordance-index 0.78 and 0.80 in training and validation cohorts, respectively). This was also superior to Model for End-stage Liver Disease, Model for End-stage liver disease sodium (MELDNa), and Child-Turcott-Pugh scores. CONCLUSIONS: Cirrhotics frequently have impaired potassium homeostasis, which has a prognostic significance. Serum potassium correlates directly with serum creatinine and urea and inversely with serum sodium. The model incorporating serum potassium developed from this study ("K"model) can predict death in advanced cirrhotics with an excellent accuracy.
BACKGROUND: The prevalence and clinical significance of hyponatremia in cirrhotics have been well studied; however, there are limited data on hyperkalemia in cirrhotics. AIM: We evaluated the prevalence and prognostic significance of hyperkalemia in hospitalized patients with cirrhosis and developed a prognostic model incorporating potassium for prediction of liver-related death in these patients. METHODS: The training derivative cohort of patients was used for development of prognostic scores (Group A, n = 1160), which were validated in a large prospective cohort of cirrhotic patients. (Group B, n = 2681) of cirrhosis. RESULTS:Hyperkalemia was seen in 189 (14.1%) and 336 (12%) in Group A and Group B, respectively. Potassium showed a significant association that was direct with creatinine (P < 0.001) and urea (P < 0.001) and inverse with sodium (P < 0.001). Mortality was also significantly higher in patients with hyperkalemia (P = 0.0015, Hazard Ratio (HR) 1.3, 95% confidence interval 1.11-1.57). Combination of all these parameters into a single value predictor, that is, renal dysfunction index predicted mortality better than the individual components. Combining renal dysfunction index with other known prognostic markers (i.e. serum bilirubin, INR, albumin, hepatic encephalopathy, and ascites) in the "K" model predicted both short-term and long-term mortality with an excellent accuracy (Concordance-index 0.78 and 0.80 in training and validation cohorts, respectively). This was also superior to Model for End-stage Liver Disease, Model for End-stage liver diseasesodium (MELDNa), and Child-Turcott-Pugh scores. CONCLUSIONS: Cirrhotics frequently have impaired potassium homeostasis, which has a prognostic significance. Serum potassium correlates directly with serum creatinine and urea and inversely with serum sodium. The model incorporating serum potassium developed from this study ("K"model) can predict death in advanced cirrhotics with an excellent accuracy.