Abstract: PD67-06
Sources of Funding: Department of Urology and Department of Biostatistics, Winthrop University Hospital

Introduction

Pathologic findings including surgical positive margin(PM) rates and lymph node(LN) yield have been used as metrics of surgical quality in radical cystectomies(RC) dissections. We aim to evaluate these measures in RC reported in the National Cancer Data Base(NCDB) stratified by hospital cystectomy volume and facility type to assess surgical quality for both robotic and open RC.

Methods

The NCDB was queried for primary RC cases (2010-2013). Stratified by institutional volume (1-2.9, 3-4.9, 5-9.9, 10-19.9 and 20+ RC/year), facility type (academic/research(AR),community, and other), type of surgery [open (ORC) and robotic assisted(RARC)] surgical quality was measured using trends of PM rates, LN dissection rates, and number of dissected LN. We performed chi-square tests for association and tests for trend to assess the relationship between facility characteristics and surgical quality measures. Univariate and multivariable logistic regression models for PM and LN dissection rates were examined.

Results

16,923 RC cases were identified. Using the above volume categories, we observed a significant increase in RARC PM rates in centers performing less than 10 RARC per year(12% vs. 9%; p=0.01). In adjusted models, low center volume was associated with a higher PM rate(p=0.02). ORC PM rate in the lowest volume(LV) category was 13.9% vs. 12.0% in the highest volume(HV) category. RARC PM rate in the LV category was 10.4% vs 9.3% in the HV category. Receipt of a LN dissection was significantly higher in HV compared to LV centers for ORC(91 vs. 84%, p=0.03) not for RARC(96.6 vs 92%; p=0.07). In adjusted models, volume association with LN dissection approached but did not reach significance(p=0.06). Median LN yield increased with higher facility volume in both ORC(17 vs. 7; p<0.001) and RARC(21 vs. 10.5; p<0.001). AR facility type was associated with lower PM rates (p=0.05), higher LN dissection rates (p=0.05), and higher median LN dissected (p<0.0001) in ORC. For RARC, AR facility type was only associated with higher median number of LN dissected in RARC (p<0.0001).

Conclusions

Several metrics of oncologic surgical quality were improved at higher cystectomy volume and academic centers for both open and robotic cystectomy. Databases capturing cancer specific survival should be queried to evaluate the downstream oncologic influences of these findings.

Funding

Department of Urology and Department of Biostatistics, Winthrop University Hospital