Algorithm of weighted mean and standard deviation calculation.
First, different average types are converted to mean according to the following formulas:
Mean=Species value in unified units
(a) for patient number <= 25:
Mean=(RangeMinValue+2* Species value in unified units +RangeMaxValue)/4
If Dispersion type is not range then
Mean= Species value in unified units
(b) for patient number > 25:
Mean= Species value in unified units
Mean=Species value in unified units
Second, different dispersion types are converted to standard deviation (SD) according to the following formulas:
SD=Species dispersion left
SD=Species Dispersion left*Patient number^1/2
SD=(Species Dispersion right-Species Dispersion left)/1.35
If patient number is more than 15 then
SD=(Range Max Value-Range Min Value)/4
SD=(Species Dispersion right-Species Dispersion left)/3.92
SD=Species Dispersion left*Patient number^1/2
SD=Species Dispersion left/1.35
SD=(Species Dispersion right-Species Dispersion left)/3.92
SD=Species dispersion left /2
Third, we calculate weighted mean:
Weighted mean=(Patient number1*Mean1+Patient number2*Mean2+…)/(Patient number1+Patient number2+…)
Fourth, we calculate weighted SD:
Weighted SD=(((Pt number1-1)*SD1^2+(Pt number2-1)*SD2^2+…)-(Pt number1+Pt number2+…)*weighted mean^2+(Pt number1*Mean1+Pt number2*Mean2+…)/((Pt number1-1)+(Pt number2-1)+…)^0.5
Weighted mean and SD are presented in the last row of the results table.