Enhancing Patient Safety in Prehospital Environment: Analyzing Patient Perspectives on Non-Transport Decisions With...

Area of Science:

• Prehospital Care
• Computational Health Informatics
• Patient Experience Research

Background:

• Understanding patient decisions to refuse hospital transportation after prehospital emergency care is crucial for service improvement.
• Advanced computational techniques offer novel ways to analyze patient feedback and behavior in emergency medical services.

Purpose of the Study:

• To explore patient experiences and perspectives on declining hospital transport post-prehospital care.
• To utilize advanced computational methods, including Natural Language Processing and machine learning, to analyze patient sentiments and predict post-refusal actions.

Main Methods:

• Interviews with 210 patients in Qatar who refused transportation after treatment by Hamad Medical Corporation Ambulance Service (HMCAS).
• Sentiment analysis and topic modeling (Latent Dirichlet Allocation) on patient responses regarding reasons for refusal and service satisfaction.
• Machine learning models (Naïve Bayes, KNN, Random Forest, SVM) were employed to predict post-refusal actions.

Main Results:

• Average participant age was 38.61 years; common complaints included chest and abdominal pain.
• Sentiment analysis indicated a generally favorable perception of HMCAS services.
• Latent Dirichlet Allocation identified key themes in refusal reasons and satisfaction.
• Naïve Bayes and SVM models achieved 81.58% accuracy in predicting post-refusal actions.

Conclusions:

• Natural Language Processing and machine learning effectively enhance understanding of patient behavior and sentiment in prehospital settings.
• Computational methodologies provide nuanced insights into patient demographics and sentiments, supporting Quality Improvement initiatives.
• Continuous integration of automated feedback systems is recommended to improve patient-centered prehospital care.