Observed to expected 30-day mortality as a benchmark for transcatheter aortic valve replacement

The observed-to-expected 30-day mortality ratio (O:E ratio) is a standard metric by which transcatheter aortic valve replacement (TAVR) trials have been evaluated. Early TAVR trials consistently demonstrated O:E ratio less than 0.6 after TAVR when based on the Society for Thoracic Surgery Predicted Risk of Mortality (STS-PROM) for surgical aortic valve replacement. Recent published results from the Transcatheter Valve Therapy (TVT) Registry have demonstrated O:E ratios of 1.0. We evaluated our own O:E ratios for TAVR to investigate this discordance.

Data were collected prospectively for TAVR patients from 2008 through 2015 (N = 546) and were reviewed retrospectively. The observed mortality and STS-PROM were calculated to formulate O:E ratios and were compared over a variety of subgroups.

Overall, the O:E ratio for 30-day mortality was 0.4 and significantly less than 1 (P < .001; 95% confidence interval, 0.25-0.63). The O:E ratio relationship remained less than 0.5 for patients with low (STS-PROM < 4), moderate (STS-PROM = 4-8) and high risk (STS-PROM > 8). The O:E ratio was significantly higher for transapical patients (O:E ratio = 0.8) when compared with transfemoral patients (O:E ratio = 0.2). Lastly, O:E ratios for both commercial (O:E ratio = 0.5) and research (O:E ratio = 0.3) patients were similar (P = .337), and both were significantly less than 1 (P = .007 and P < .001, respectively).

The STS-PROM consistently overestimated 30-day mortality after TAVR. Achieving an O:E ratio less than 0.6 may be a realistic goal for all TAVR programs. While an accurate and specific risk calculator for 30-day mortality after TAVR remains to be established, our data suggest that current TVT results are not acceptable for commercial TAVR and that programs with an O:E ratio greater than 0.6, based on the STS-PROM, should reevaluate internal processes to improve their results.