Survival analysis of heart failure patients: A case study

Abstract

This study was focused on survival analysis of heart failure patients who were admitted to Institute of Cardiology and Allied hospital Faisalabad-Pakistan during April-December (2015). All the patients were aged 40 years or above, having left ventricular systolic dysfunction, belonging to NYHA class III and IV. Cox regression was used to model mortality considering age, ejection fraction, serum creatinine, serum sodium, anemia, platelets, creatinine phosphokinase, blood pressure, gender, diabetes and smoking status as potentially contributing for mortality. Kaplan Meier plot was used to study the general pattern of survival which showed high intensity of mortality in the initial days and then a gradual increase up to the end of study. Martingale residuals were used to assess functional form of variables. Results were validated computing calibration slope and discrimination ability of model via bootstrapping. For graphical prediction of survival probability, a nomogram was constructed. Age, renal dysfunction, blood pressure, ejection fraction and anemia were found as significant risk factors for mortality among heart failure patients.

Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone

“…We analyzed a dataset containing the medical records of 299 heart failure patients collected at the Faisalabad Institute of Cardiology and at the Allied Hospital in Faisalabad (Punjab, Pakistan), during April-December 2015 [52,66]. The patients consisted of 105 women and 194 men, and their ages range between 40 and 95 years old ( Table 1).…”

“…Some features are binary: anaemia, high blood pressure, diabetes, sex, and smoking ( Table 1). The hospital physician considered a patient having anaemia if haematocrit levels were lower than 36% [52]. Unfortunately, the original dataset manuscript provides no definition of high blood pressure [52].…”

“…An abnormally low level of sodium in the blood might be caused by heart failure [70]. The death event feature, that we use as the target in our binary classification study, states if the patient died or survived before the As done by the original data curators [52], we represented this dataset as a table having 299 rows (patients) and 13 columns (features). For clarification purposes, we slightly changed the names of some features of the original dataset (Additional file 1).…”

“…In this paper, we analyze a dataset of medical records of patients having heart failure released by Ahmad and colleagues [52] in July 2017. Ahmad and colleagues [52] employed traditional biostatistics time-dependent models (such as Cox regression [53] and Kaplan-Meier survival plots [54]) to predict mortality and identify the key features of 299 Pakistan patients having heart failure, from their medical records. Together with their analysis description and results, Ahmad and coworkers made their dataset publicly available online ("Dataset" section), making it freely accessible to the scientific community [55].…”

“…Afterwards, Zahid and colleagues [56] analyzed the same dataset to elaborate two different sex-based mortality prediction models: one for men and one for women. Although the two aforementioned studies [52,56] presented interesting results, they tackled the problem by standard biostatistics methods, leaving room for machine learning approaches. We aim here to fill this gap by using several data mining techniques first to predict survival of the patients, and then to rank the most important features included in the medical records.…”