Survival differences following lung transplantation among US transplant centers.

CONTEXT: Although case loads vary substantially among US lung transplant centers, the impact of center effects on patient outcomes following lung transplantation is unknown.

OBJECTIVE

To assess variability in long-term survival following lung transplantation among US lung transplant centers.

DESIGN, SETTING, AND PATIENTS: Analysis of data from the United Network for Organ Sharing registry for 15,642 adult patients undergoing lung transplantation between 1987 and 2009 in 61 US transplantation centers still active in 2008.

MAIN OUTCOME MEASURES: Mixed-effect Cox models were fitted to assess survival following lung transplantation at individual centers.

RESULTS

In 2008, 19 centers (31.1%) performed between 1 and 10 lung transplantations; 18 centers (29.5%), from 11 to 25 transplantations; 20 centers (32.8%), from 26 to 50 transplantations; and 4 centers (6.6%), more than 50 transplantations. One-month, 1-year, 3-year, and 5-year survival rates among all 61 centers were 93.4% (95% confidence interval [CI], 93.0% to 93.8%), 79.7% (95% CI, 79.1% to 80.4%), 63.0% (95% CI, 62.2% to 63.8%), and 49.5% (95% CI, 48.6% to 50.5%), respectively. Characteristics of donors, recipients, and surgical techniques varied substantially among centers. After adjustment for these factors, marked variability remained among centers, with hazard ratios for death ranging from 0.70 (95% CI, 0.59 to 0.82) to 1.71 (95% CI, 1.36 to 2.14) for low- vs high-risk centers, for 5-year survival rates of 30.0% to 61.1%. Higher lung transplantation volumes were associated with improved long-term survival and accounted for 15% of among-center variability; however, variability in center performance remained significant after controlling for procedural volume (P < .001).

CONCLUSIONS

Center-specific variation in survival following lung transplantation was only partly associated with procedural volume. However, other statistically significant sources of variability remain to be identified.