Premium
Long‐term reliability and observer comparisons in the radiographic diagnosis of periapical disease
Author(s) -
Molven O.,
Halse A.,
Fristad I.
Publication year - 2002
Publication title -
international endodontic journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.988
H-Index - 119
eISSN - 1365-2591
pISSN - 0143-2885
DOI - 10.1046/j.0143-2885.2001.00525.x
Subject(s) - endodontist , medicine , kappa , radiography , endodontics , reliability (semiconductor) , cohen's kappa , dentistry , orthodontics , radiology , statistics , mathematics , power (physics) , physics , geometry , quantum mechanics
Aim The aim of this study was to evaluate and compare the long‐term diagnostic consistency of two examiners, an endodontist and a radiologist, and to make comparisons with findings recorded by an observer with more recent scientific and clinical experience in endodontics. Methodology Three groups, each consisting of 20 full mouth series of intraoral radiographs, with 79, 93 and 85 endodontically‐treated roots, respectively, were successively evaluated for periapical disease. Evaluations were at first performed separately by the three observers. Disagreement and difficult, borderline cases were subjected to joint evaluation. Intra‐ and interexaminer comparisons were made. For two of the observers the observations were compared with findings recorded several years before for the same cases in the same radiographs. Results The intra‐ and interobserver long‐term reliability of the two original examiners resulted in 83% overall agreement, the kappa values were 0.54, 0.57 and 0.53. Comparisons between all three observers disclosed 82%, 85% and 86% agreement and kappa values 0.55, 0.58 and 0.60. The joint evaluations and decisions did not indicate a dominating influence from any of the observers. Conclusions The long‐term reliability of the two original observers was judged as being satisfactory. All three observers judged the overall disease status of the material in the same way. The joint discussions of selected cases might reduce observer variation to an acceptable level, avoid a number of false recordings and increase the reliability and validity of the findings.