JMIR Publications

ABSTRACT

I developed long COVID after a SARS-CoV-2 infection in April 2023 and spent nearly 3 years largely housebound, and periodically bedbound, with post-exertional malaise, fatigue, cognitive impairment, muscle twitching, and chest pain. Over that period, I consulted 2 primary care physicians, 7 specialists, and 2 long COVID clinics, and enrolled in 1 clinical trial. Every clinician I met was willing to help, and none had treatments to offer, because no established protocol for treating the underlying condition existed. I eventually concluded that if my long COVID was going to be investigated, I would have to be the prime investigator. This article describes what that looked like from the inside. How I built a personal differential diagnosis out of home-designed tests, selectively ordered laboratory work, and used my own consumer genetic data compared against two unaffected relatives. I describe how I used artificial intelligence tools as teaching aids and built software to automate literature surveillance while cognitively impaired. I further describe how I decided what level of evidence was good enough to act on - when waiting for certainty carried its own risk. And finally, I describe how I tried to keep my self-directed protocols safe.