Inter-observer agreement in the reporting of Tc-99(m)-DMSA renal studies

The early identification of renal cortical scarring secondary to urinary reflux and/or renal infection is important in the management of patients with recurrent disease. Scintigraphic imaging of the renal cortex using Tc-99(m)-dimercaptosuccinic acid (DMSA) is often considered the standard method for the diagnosis of renal scars in both adults and children. Consistent reporting of Tc-99(m)-DMSA is nevertheless essential in ensuring that the clinician can act on the data reported. In this study, seven experienced observers were asked to report, independently, Tc-99(m)-DMSA data sets from 32 patients. The observers were asked to note the presence of a space occupying lesion or of a renal cortical scar(s) and to conclude if the kidney was normal or abnormal. There was marked variation in the number of renal cortical scars reported, the total number of scars seen and whether or not the kidney was normal or abnormal. After peer review of all data, consensus and referenced criteria were arrived at for a reanalysis of data. All scans were reread by the observers. There was some improvement in the concordant reporting of kidneys with renal scars (51 vs 61%) and whether or not the kidney was normal or abnormal (53 vs 63%). However, this was not statistically significant (at a level of P < 0.05). This study demonstrated significant inter-observer variation regarding the reporting of Tc-99(m)-DMSA studies. A consensus with agreed guidelines for data interpretation did not significantly affect this apparent lack of consistency of reporting.