Artificial Intelligence Chatbot dalam Pelayanan Kesehatan: Dampaknya terhadap Efisiensi Pelayanan dan Pengalaman Pasien

Abstract

AI chatbots are increasingly adopted in healthcare, but evidence of their effects on service efficiency and patient experience remains fragmented. This systematic literature review synthesized 2021–2026 evidence using PRISMA 2020. Searches of PubMed/PMC, DOAJ, and publisher websites identified 14 relevant articles from 166 records. Findings showed efficiency gains in structured tasks, including a 57.3% reduction in history-taking time, while triage accuracy varied (pooled ChatGPT-4 accuracy: 0.86; I² = 93%). Patient experience was generally positive, with 73% satisfaction among perioperative patients, although chatbots were less effective than physicians in reducing anxiety. Key barriers included privacy, empathy, system integration, and limited real-patient data. AI chatbots can improve healthcare services when applied to structured processes with appropriate human oversight.