Can available mathematical models predict serum digoxin levels in Thai patients?

• Main Authors: J. Jiratham-Opas, R. Kanjanavanit, W. Wongcharoen, B. Punyawudho, P. Arunmanakul, A. Amaritakomol, P. Topaiboon, S. Gunaparn, A. Phrommintikul
• Format: Journal
• Published: 2018
• Subjects: Medicine; Pharmacology, Toxicology and Pharmaceutics
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85041023177&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/58899
• Institution: Chiang Mai University

© 2018 John Wiley & Sons Ltd What is known and objective: Digoxin is commonly prescribed for heart failure patients with reduced ejection fraction (HFrEF) and patients with atrial fibrillation (AF). Due to digoxin's narrow therapeutic range, monitoring the serum digoxin concentration (SDC) is important. However, the SDC measurement is not widely available. Equations using clinical parameters can be employed to estimate the SDC but have never been studied in the Thai population. Therefore, we conducted this study to evaluate the correlation between the measured SDC and predicted digoxin level using 2 commonly used equations: the Konishi equation and the Koup and Jusko equation. Methods: This report describes prospective, cross-sectional study conducted at Chiang Mai University. One hundred and fourteen patients were recruited in the study. All of the patients were diagnosed as having HFrEF, AF or both and had been receiving digoxin for at least 4 weeks. The SDC of each patient was measured at steady state and assigned to one of 3 groups according to the classifications of the Digitalis Investigation Group (DIG) trial: in the therapeutic range, over the therapeutic range and in the suboptimal range. Results and discussion: There were significant correlations between the measured and predicted SDCs using both the Konishi equation and the Koup and Jusko equation, which had correlation coefficients (R) of 0.69 and 0.31 (P <.05 for both), respectively. The percentages of patients with measured SDCs in the therapeutic range, over the therapeutic range and in the suboptimal range were 27.2%, 9.6% and 63.2%, respectively. The sensitivity and specificity of the Konishi equation in predicting SDCs in the over the therapeutic range were 72.73% (95% Confidence interval (CI): 39.03%-93.98%) and 80.58% (95% CI: 71.62%-87.72%), respectively. Of the 5 patients (4.4%) who were rehospitalized, 2 patients (0.01%) were readmitted due to acute decompensated heart failure (ADHF). One of the patients had an SDC that was over the therapeutic range. None of the readmitted patients had ventricular arrhythmia. What is new and conclusions: The Konishi equation yielded better predictions of the SDC, especially in the subgroup of HFrEF patients. Furthermore, the prediction of SDCs in the over the therapeutic range using this equation was superior to that of the Koup and Jusko equation. With further validation in a larger population, this equation should facilitate the detection of patients who are over the therapeutic range in clinical practice.