Launching today
DxCoach
Write your clinical reasoning, get attending-level feedback
4 followers
Write your clinical reasoning, get attending-level feedback
4 followers
DxCoach is a free tool for medical students to practice diagnostic reasoning. Work through realistic patient cases, write out your full clinical thinking, and get structured feedback that grades your reasoning, not just your final answer. Feedback is grounded in research on how diagnostic expertise actually develops: building illness scripts and deliberate reflection, not memorizing bias checklists. Inline annotations mark what's strong and where your reasoning broke down.
I built DxCoach because most tools for medical students test recall like multiple choice, flashcards, question banks. Almost nothing evaluates the actual reasoning process: the thinking that connects a patient's presentation to a diagnosis. That's the skill that separates a good clinician from someone who memorized the right answers, and it's the hardest thing to practice on your own.
DxCoach works differently. You read a realistic patient case, write out your full diagnostic reasoning in your own words, and get structured feedback that grades how you reasoned, not just whether your final answer was right. The feedback marks what was strong, where your logic broke down, and what you missed, with inline annotations directly on your writing. A student can name the correct diagnosis through weak reasoning and still get told why the reasoning was thin, or reach a wrong answer through sound thinking and be credited for the process.
The design is grounded in research on how diagnostic expertise actually develops, Norman et al. (2024) and the work on deliberate reflection. The core finding is that diagnostic errors come mostly from underdeveloped illness scripts and knowledge gaps, not from failing to memorize cognitive bias checklists. So the feedback is built to develop reasoning and clinical knowledge, not to make you recite biases. That distinction shapes everything about how the tool responds.
This is an early version, and all cases are fictional and educational. It is a tool for practicing reasoning, not a source of medical advice, and it should never inform real patient care. I'm actively working on clinical review of the case library and would genuinely welcome feedback from medical students, educators, and clinicians, especially critical feedback on the cases and the reasoning the tool rewards. That's exactly what I need to make it better.
It's free. I'd love for you to try a case and tell me where it falls short.