Variabilidad intraobservador e interobservador de las líneas de Gilula en el diagnóstico de las luxaciones carpianas

Abstract

Objective

Alteration of Gilula's lines in the posteroanterior (PA) radiology of the wrist is a sign of perilunate dislocation. Our objective is to evaluate the intraobserver and interobserver variability of the Gilula lines in perilunate dislocations and the agreement between general orthopaedic surgeons, hand surgeons and residents of orthopaedic surgery, who participated in the study.

Materials and methods

Six observers evaluated 30 PA carpal X-rays, including 15 with a diagnosis of perilunate dislocation and 15 without carpal lesions. These observers were the residents of orthopaedic surgery; orthopaedic surgeons not specialised in hand pathology and hand surgeons. All had to classify the X-rays as normal or pathological attending only to Gilula's lines. We calculated the intraobserver and interobserver variability using the Kappa coefficient.

In this study, we have discussed indistinctively of concordance, agreement and variability as synonyms.

Results

Intraobserver concordance was very good (0.867–1.000) in residents and in hand surgeons, and moderate or good (0.553–0.795) in general orthopaedic surgeons.

Interobserver concordance was very good (0.875) among hand surgeons, while it was good among general orthopaedic surgeons (0.679) and residents (0.751).

Interobserver concordance was greater among residents and hand surgeons (Kappa 0.875, very good agreement) than among residents and general orthopaedic surgeons (Kappa 0.702, good agreement).

Conclusion

The observation of Gilula lines is a good diagnostic method in case of suspicion of perilunate dislocation, with good intraobserver and interobserver agreement both in traumatologists in training and those who are already trained, whether or not they are specialists in hand surgery.