Fig. 5

Clinical evaluation of our newly established diagnostic test algorithm for HTLV-1 infection. The suitability of the established HTLV-1 diagnostic test algorithm was evaluated clinically. a Eight hundred and seventy-three samples determined reactive by a primary test were in parallel confirmed for positivity by WB, LIA, and PCR at each domestic clinical laboratory. The rates of each judgment were calculated and are indicated as percentages in the pie graphs. b To determine the best test algorithm, 285 WB-indeterminate samples, 60 LIA-indeterminate samples, 432 PCR-positive samples, and 494 PCR-negative samples, which were all primary test-reactive, were additionally (secondly) confirmed by PCR or LIA on the former two and the latter two samples, respectively, at each domestic clinical laboratory. Herein the rates of each re-judgment by PCR or LIA were calculated, as indicated by percentages (upper and lower pie graphs, respectively). P: positive, N: negative, and I: indeterminate