Simple approximation of sample size for precise estimates of SARS-CoV-2 infection from point-seroprevalence studies