Influence of Tet Reliability on the creening Performance of Frequency-Doubling Perimetry

Purpoe To invetigate the influence of tet reliability on the creening performance of frequency-doubling perimetry (FDT). Deign Cro-ectional tudy. Method FDT enitivity and pecificity were calculated three time uing three different trategie for handling unreliable (that i, >0 reliability indice outide normal limit) and unfeaible (that i, could not be completed) tet in 452 glaucoma patient from our outpatient department and 237 healthy ubject recruited outide the hopital. Reult Unreliable (P = .01) and unfeaible (P < .001) tet occurred more frequently in glaucoma patient when compared with healthy ubject. Bet creening performance wa obtained by interpreting unreliable tet a if they were reliable and conidering unfeaible tet a poitive. With thi trategy, enitivity wa 0.90 (95% confidence interval 0.87–0.93) and pecificity 0.81 (95% confidence interval 0.76–0.86). Concluion Unreliable and unfeaible FDT tet appear to contain ueful information