This thesis designs and validates an accessibility-first healthcare web application that enables patients, including those with visual, motor, or cognitive challenges, to read content, navigate interfaces, complete tasks, and receive feedback without workarounds. Built with React, Next.js, and TypeScript, the system integrates text-to-speech, adjustable typography, and high-contrast themes aligned with WCAG 2.2 standards. A persistent accessibility toolbar provides global control for font size, theme, and read-aloud features. Testing involved keyboard-only walkthroughs, screen-reader checks, and contrast verification, showing consistent focus management and improved readability. The results demonstrate that accessibility can be engineered into modern web applications as a default behavior rather than an afterthought.