High-volume centers for pancreatic cancer surgery: Does the media message matter?

• Published in: Journal of clinical oncology Vol. 31; no. 4_suppl; p. 297
• Main Authors: Chau, Zeling, Smith, Jillian K, West, James K., Zhou, Zheng, McDade, Theodore P., Ng, Sing Chau, Whalen, Giles Francis, Moser, Arthur J., Tseng, Jennifer F.
• Format: Journal Article
• Language: English
• Published: 01.02.2013
• ISSN: 0732-183X; 1527-7755
• DOI: 10.1200/jco.2013.31.4_suppl.297

Abstract only 297 Background: Hospital procedure volume is associated with pancreatic cancer outcomes. However, differences between high-volume hospitals are less clear. Furthermore, consumers are barraged with grading systems for hospitals, whose relationship to volume and real outcomes are poorly delineated. Methods: Massachusetts Division of Health Care Finance and Policy Hospital Inpatient Discharge Database was queried to identify 5 highest-volume hospitals performing resection for pancreatic cancer 2005-2009. Primary outcomes included mortality, complications, median length of stay (LOS) and composite outcomes score (COS) including above. Ranks were determined for each outcome and compared to procedure volume order (highest to lowest), as well as ratings identified from popular media and internet-based hospital ratings including US News and World Reports (USN) Consumer Reports (CR), Healthgrades (HG) and Hospital Compare (HC). An inter-rater reliability analysis was performed with Spearman’s rank correlation coefficients calculated between outcomes and ratings. Results: 5 hospitals performing 82% (n=681) of statewide resections for pancreatic cancer were identified. In-hospital mortality ranged from 0-5.1%, complications 9.0-32.0%, median LOS 7-11 days. Within 5 highest-volume centers, outcomes did not improve in a consistent linear fashion with increased volume; however, volume remained more correlated with outcomes than consumer ratings. Of consumer ratings, HG ratings were most correlated with mortality, but had Spearman correlation coefficient of only 0.32. Complication measures were correlated with HG and USN equally at 0.22. USN rankings were highly correlated with LOS measures at 0.85. When all outcome measures were combined, HG was most correlated at 0.32. CR ratings were inversely correlated with all outcome measures. Conclusions: Between high-volume hospitals, outcome variation exists that does not completely correlate with either absolute volume numbers or by popular media grading systems, which can be discordant. Both consumer-directed ratings and volume data, where available, should be used only in appropriate contexts to direct patient care.