<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;">Boa tarde, com esta rotina, <br><br><span style="font-family: courier,monaco,monospace,sans-serif;">require(DiagnosisMed)</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">ele<-matrix(c(55,7,49,84),nrow=2,ncol=2,byrow=T,</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">dimnames=list(Eletro=c("Presente","Ausente"),</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Etsenose=c("Presente","Ausente")))</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">ele</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span
 style="font-family: courier,monaco,monospace,sans-serif;">diagnosis(ele,plot=F)</span><br style="font-family: courier,monaco,monospace,sans-serif;"><br style="font-family: courier,monaco,monospace,sans-serif;">Tenho a saída<br><br><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">---------------------------------------------------------------</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">          Etsenose</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Eletro     Presente Ausente Sum</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;"> 
 Presente       55       7  62</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">  Ausente        49      84 133</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">  Sum           104      91 195</span><br style="font-family: courier,monaco,monospace,sans-serif;"><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">The test has the following parameters [95% confidence interval]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family:
 courier,monaco,monospace,sans-serif;">---------------------------------------------------------------</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Sample size:                   195 </span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Prevalence considered(%):      46.67 </span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Sensitivity(%):                92.31  [ 84.96  -  96.22 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family:
 courier,monaco,monospace,sans-serif;">Specificity(%):                52.88  [ 43.36  -  62.20 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Positive predictive value(%):  63.16  [ 54.70  -  70.88 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Negative predictive value:(%): 88.71  [ 78.48  -  94.42 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Positive likelihood ratio:     1.96   [ 1.58  -  2.44 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family:
 courier,monaco,monospace,sans-serif;">Negative likelihood ratio:     0.15   [ 0.07  -  0.30 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Diagnostic odds ratio:         13.28  [ 5.47  -  37.30 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Error trade off (FN : FP)      0.14 : 1 </span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Error rate(%):                 28.72  [ 22.83  -  35.43 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family:
 courier,monaco,monospace,sans-serif;">Accuracy(%):                   71.28  [ 64.57  -  77.17 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Youden index:                  0.4519  [ 0.4555  -  0.4484 ]</span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">Area under ROC curve:          0.726 </span><br style="font-family: courier,monaco,monospace,sans-serif;"><span style="font-family: courier,monaco,monospace,sans-serif;">-----------------------------------------------------------</span><br><br><br>Quando  chequei os resultados, vi que
 sensibilidade e especificidade estao invertidos. <br><br>Vejam: Sensibiidade = a/(a+c)=55/104=0,5288. Na sáida está como especificidade, ou seja, invertido. O que estou fazendo errado.<br><br>Valores corretos: sensib=0,5288 especif=0,9231<br><br><br>Obrigado a todos.<br><br>Edson Lira<br>
Estatístico<br>
Manaus-Amazonas</td></tr></table>