Errors in distemper virus titrations performed in embryonated eggs

Summary

Distemper virus titrations of 2-, 4-, 5-, and 10-fold steps, each using 5, 8, 10, and 15 C.A.M.'s per dilution, showed differences in the standard error of the median effective dose and mean regression coefficients. For reasonable results, a titration using 2-fold dilutions and eight C.A.M.'s per dilution, repeated three times, was selected as a design for laboratory use. Probit analysis of the titration showed a MED of 4.7±0.076 log₁₀ virus dilution. A 95% confidence interval of the median effective dose was 0.30 log₁₀ virus dilution with 81% of the endpoints occurring 0.10 log₁₀ virus dilution on either side of the median effective dose. The regression coefficient was −1.6±0.29 probits per log₁₀ virus dilution when a 1.5 log₁₀ virus dilution range was used, with the median effective dose occurring midway in the virus dilution range. The regression was valid (t=5.6, t_{[5%, 3df]}=3.2) and the data fitted the line (χ²=∼0.0, χ² _{[5%, 3 dff]}=7.8).