"Symptoms" are self-reported. Symptomatic seroprevalence is objective. What this should tell you is that "symptoms" are a bad endpoint for a randomized-control

"Symptoms" are self-reported. Symptomatic seroprevalence is objective. What this should tell you is that "symptoms" are a bad endpoint for a randomized-controlled trial, not that if the good endpoint doesn't show you what you want, you should rely on the bad one.