Can the Brucellacapt (R) test be substituted for the Coombs test in the diagnosis of human brucellosis?

BACKGROUND. Throughout this work we have studied the capacity of Brucellaca pt((R)) test to replace Coombs test in the serological diagnosis of human b rucellosis. METHODS. A total of 66 initial sera from patients with diagnostic of brucel losis were studied. The patients were divided in two groups: 42 patients sh owing a primo-infection (group1), and 24 patients with a previous case of b rucellosis (group 2). As a controls, for the group 1 we have used 100 sera from healthy donors, and for group 2, 28 sera from people that have had cli nical brucellosis but actually are in good health. All serum samples were t ested in either Coombs and Brucellacapt((R)) tests. The diagnostic yield wa s calculated using ROC (receiver-operating characteristic) plots. Moreover, the results obtained in Coombs and Brucellacapt((R)) tests with 397 serum samples from 66 patients with brucellosis were compare with a nonparametric method. RESULTS. The sensibility and specificity for group1 were respectively 1 and 0.98 for Coombs and, 1 and 0.95 for Brucellacapt((R)) tests. For group 2, the results in Coombs test were 1 and 0.80, and in Brucellacapt((R)) test 0 .95 and 0.74. In this second group, the area under the ROC plot was 0.950 f or Coombs and 0.904 for Brucellacapt((R)) tests. Non statistical difference s were observed comparing both serological tests using the Wilcoxon method (Z= 0.213; p=0.8). CONCLUSIONS. Brucellacapt((R)) and Coombs tests yield similar diagnostic re sults in the follow-up of serological samples from patients with brucellosi s, and its should consider as interchangeables.