Paul I Sia1,2, David Curragh1,2, Stuart Howell3, Dinesh Selva1,2. 1. South Australian Institute of Ophthalmology, Royal Adelaide Hospital, Adelaide, South Australia, Australia. 2. Department of Ophthalmology and Visual Sciences, University of Adelaide, Adelaide, South Australia, Australia. 3. Data, Design and Statistics Service, Adelaide Health Technology Assessment, School of Public Health, University of Adelaide, Adelaide, South Australia, Australia.
Abstract
IMPORTANCE: This study assesses the interobserver agreement on dacryocystography (DCG) and dacryoscintigraphy (DSG) findings. BACKGROUND: There are no standard grading criteria to guide the interpretation of conventional DCG and DSG findings and therefore there may be a degree of subjectivity. This study evaluates the level of interobserver agreement in the interpretation of DCG and DSG findings. DESIGN: A retrospective observational study at the Royal Adelaide Hospital. PARTICIPANTS: A total of 165 patients who presented with epiphora with 276 DCGs and 290 DSGs performed were included in this study. METHODS: DCG and DSG images were obtained, anonymized, randomized and interpreted by three independent oculoplastic surgeons. Standard grading criteria were set for both DCG and DSG images. Data from all observers were analysed for interobserver agreement using Kappa (κ) statistics, generated using a variation of Cohen's kappa for multiple observers. MAIN OUTCOME MEASURES: Level of interobserver agreement (κ values) in the grading of DCG and DSG findings. RESULTS: There was an overall moderate interobserver agreement for DCG findings (κ = 0.55), with the highest agreement on interpreting canalicular obstruction (κ = 0.80), followed by proximal nasolacrimal duct obstruction (κ = 0.67) and normal patency (κ = 0.63). There was an overall fair interobserver agreement for DSG findings (κ = 0.36), with the best being moderate agreement (κ = 0.42-0.50) for interpreting pre-sac delay and post-sac proximal delay. CONCLUSIONS AND RELEVANCE: DCG offers good reliability in interpreting patent and obstructed systems. On the other hand, DSG has poor agreement and highlights some of the limitations in the ability to guide epiphora management.
IMPORTANCE: This study assesses the interobserver agreement on dacryocystography (DCG) and dacryoscintigraphy (DSG) findings. BACKGROUND: There are no standard grading criteria to guide the interpretation of conventional DCG and DSG findings and therefore there may be a degree of subjectivity. This study evaluates the level of interobserver agreement in the interpretation of DCG and DSG findings. DESIGN: A retrospective observational study at the Royal Adelaide Hospital. PARTICIPANTS: A total of 165 patients who presented with epiphora with 276 DCGs and 290 DSGs performed were included in this study. METHODS:DCG and DSG images were obtained, anonymized, randomized and interpreted by three independent oculoplastic surgeons. Standard grading criteria were set for both DCG and DSG images. Data from all observers were analysed for interobserver agreement using Kappa (κ) statistics, generated using a variation of Cohen's kappa for multiple observers. MAIN OUTCOME MEASURES: Level of interobserver agreement (κ values) in the grading of DCG and DSG findings. RESULTS: There was an overall moderate interobserver agreement for DCG findings (κ = 0.55), with the highest agreement on interpreting canalicular obstruction (κ = 0.80), followed by proximal nasolacrimal duct obstruction (κ = 0.67) and normal patency (κ = 0.63). There was an overall fair interobserver agreement for DSG findings (κ = 0.36), with the best being moderate agreement (κ = 0.42-0.50) for interpreting pre-sac delay and post-sac proximal delay. CONCLUSIONS AND RELEVANCE: DCG offers good reliability in interpreting patent and obstructed systems. On the other hand, DSG has poor agreement and highlights some of the limitations in the ability to guide epiphora management.