Sentiment text classification plays a vital role in extracting valuable insights from discharge summaries. These summaries, containing patient information and medical observations, often carry subjective sentiments that can provide valuable feedback on the patient’s experience and overall satisfaction. By examining the subjective sentiment contained within these summaries, valuable insights can be extracted regarding the patient’s experience and overall satisfaction with healthcare services. Analyzing the sentiment expressed in discharge summaries enables healthcare providers to estimate patient satisfaction and identify potential areas of improvement within healthcare services, fostering an environment for enhancing the overall quality of patient care. In this article, sentiment text classification techniques were applied to the discharge summaries utilizing natural language processing tools and libraries such as Pattern and SentiWordNet. Apart, we propose a new method for sentiment analysis in discharge summary classification. Our novel method refers to a new list of positive and negative words from discharge summaries. These tools enable the identification and classification of sentiment into positive and negative categories conveyed in the text. The sentiment analysis process involved various preprocessing steps, including text cleaning, tokenization, and lemmatization to ensure the accuracy and effectiveness of the classification. The discharge summaries from MIMIC-IV were processed, and sentiment labels were assigned to each text based on the identified sentiment categories. The analysis of sentiment will contribute to a deeper understanding of patient sentiments, identify areas for service improvement, and ultimately empower healthcare providers to deliver better healthcare experiences.