In this study we assess the mortality risk after 6 months, 1 year, and 2 years of follow-up in a sample of frail patients diagnosed with delirium who had been readmitted to hospital before 30 days after discharge (premature readmission). We compute Kaplan Meier estimates of survival probability functions at the selected censoring times for the premature readmission cohort versus the non-premature readmission cohort. Addtionally, we compute Independent Samples T-Test, Logistic Regression, Random Forest and Boosting Classification to indetify significant risk factors for premature hospital readmission. We find that premature hospital readmission is associated with increased mortality at 6 months, 1 year, and 2 years. The study identifies as high risk factor for premature hospital readmission the following ones: suffering congestive heart failure (CHF), diabetes, the use of anti-arrhythmic and antidepressant drugs as well as variables that are indicators of frailty such as older age, low weight, male sex, polypharmacy, lack of balance, weight loss measured through calf circumference, and high dependency to perform activities of daily living are indicative of higher risk of premature hospital readmission.