<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>