Can RENEX, a decision support system for Tc-99m MAG3 diuretic renography, teach residents better diagnostic skills?

540 Objectives RENEX is a decision support system that interprets Tc-99m MAG3 diuretic studies based on patient demographics and data generated from the renogram; in addition, RENEX provides the rationale for its interpretation and could serve as an educational tool. The objective was to determine if training of nuclear medicine residents by RENEX would lead to improved diagnostic performance. Methods Two sets of 50 adult patients with Tc-99m MAG3 diuretic studies were randomly selected from an archived database and independently interpreted by 3 expert readers, RENEX, and 5 nuclear medicine residents with at least one year of experience. The mean ages of the patients in the 1st and 2nd groups were 57.2 ± 17.7 and 58.7 ± 15.8 years, respectively; all studies contained a 24 min baseline and a 20 min post-furosemide acquisition. Each kidney was scored from +1.0 to -1.0; scores 0.20 indicated obstruction with higher scores indicating greater confidence in the diagnosis of obstruction. Scores from +0.19 to -0.19 were equivocal and scores -0.20 indicated non-obstruction with lower scores indicating greater confidence in the diagnosis of non-obstruction. Residents interpreted the 1st set of 50 studies unaided and were subsequently trained by reinterpreting the 1st set of 50 studies with access to their original scores and to RENEX. To evaluate a training effect, residents then interpreted a 2nd set of 50 studies without access to RENEX. The CCC (concordance correlation coefficient) was used to compare the scores of experts and residents. The CCC was also calculated after replacing one of the expert readers with a resident; if resident interpretations were equivalent to expert readers, the CCC of the “one resident/two experts” combination would be the same as that of the 3 experts. Results There was no significant difference in CCC of the 3 experts between the 1st and 2nd sets of patients, 0.90 (CI: 0.86, 0.93) vs 0.89 (CI: 0.84, 0.92), p = 0.62. The CCC of the 5 residents for the 2nd 50 patients after training by RENEX was significantly improved compared to the CCC obtained for the 1st 50 patients, 0.71 (CI: 0.63, 0.77) for the 2nd set vs 0.49 (CI: 0.40, 0.53) for the 1st set, p Conclusions Resident agreement with expert interpretations significantly improved after training with RENEX, p < 0.01. Interobserver agreement of diuretic renography interpretations by nuclear medicine residents with at least one year of training also improved after training with RENEX, p < 0.01. RENEX has the potential to be a useful educational and diagnostic aid for Tc-99m MAG3 diuretic renography. Research Support: NIDDK, R01 EB008838