Evaluation of the Repeatability and Reproducibility of the Results of Lymph Node Cytology for Leishmania Amastigotes in Dogs Without Leishmaniosis and in Dogs with Leishmaniosis Before, During, and After Treatment

Lymph node (LN) cytology is sensitive for the detection of parasites in dogs with leishmaniosis (CanL) and is used for the semiquantitative measurement of parasitic burden. The aim of this study was to investigate the repeatability and reproducibility of the qualitative (detection of <i>Leishmania</i> amastigotes) and semiquantitative (measurement of parasitic burden using a logarithmic scale) results of LN cytology. Three experienced examiners evaluated, blindly and independently of each other, two LN smears from each of the 53 dogs, including dogs without CanL (<i>n</i> = 40 smears), dogs with CanL (<i>n</i> = 32 smears), and dogs during or after treatment of CanL (<i>n</i> = 40 smears). Cohen's kappa coefficient (κ) indicated substantial-to-good agreement of the qualitative results when the same examiner evaluated the same smear twice (intraobserver repeatability; κ = 0.75-0.964), when the same examiner evaluated the two smears from the same dog (inter-sample reproducibility; κ = 0.626-0.964), and when different examiners evaluated the same smears (interobserver reproducibility; κ = 0.626-0.893). For the semiquantitative results, allowing a one-point difference in each pair of measurements, agreement was also substantial to good with κ values of 0.952-1 (intraobserver repeatability), 0.732-0.935 (inter-sample reproducibility), and 0.804-0.941 (interobserver reproducibility). High repeatability and reproducibility, in addition to high sensitivity, indicate that LN cytology is useful for diagnosis and treatment monitoring of CanL.