Using machine learning and natural language processing, researchers analyzed Twitter posts by academic emergency medicine physicians from 2018 to 2022. During the COVID-19 pandemic, the language and themes of the posts reflected significantly more anxiety, anger, depression, and loneliness than posts these users made before the pandemic. The study suggests that social media could help gauge the well-being of health care workers in real time. An understanding of clinician experiences can inform strategies to support the health care workforce and trainees.