Low-volume centers using ECMO have poorest survival rates

Key clinical point: Low-volume lung transplantation centers in the United States typically have the poorest survival rates compared to those with higher volumes when using ECMO.

Major finding: Of 85 LTx subjects bridged via ECMO, 20 (23.5%) of these were bridged in low, 30 (35.3%) in medium, and 35 (41.2%) in high-volume centers; in the ECMO cohort, the lowest 5-year survival rate (13.61%) was observed at low-volume centers.

Data source: Retrospective analysis of 16,603 adult LTx recipients in the International Registry for Heart and Lung Transplantation during 2005-2010.

Disclosures: Dr. Hayanga reported no financial conflicts of interest.




SAN DIEGO – Lung transplantation centers that are considered low volume tend to have lower rates of survival than do those of their medium- and high-volume counterparts when patients are bridged via extracorporeal membrane oxygenation (ECMO), according to researchers.

Even so, there is a point at which survival outcomes begin to improve for low-volume centers, they added.

“Increasingly, [ECMO] is used as a bridge to lung transplantation; indeed, the use of ECMO has tripled over the past 15 years and survival has increased by the same magnitude,” Dr. Jeremiah A. Hayanga said at the annual meeting of the Society of Thoracic Surgeons.

Dr. Jeremiah A. Hayanga
Dr. Jeremiah A. Hayanga

“An entire body of literature has linked high-volume [centers] to improved outcomes in the context of complex surgical procedures. Lung transplantation [LTx] falls within the same domain, and has been considered subject to the same inverse volume-outcome paradigm,” said Dr. Hayanga of Michigan State University, Grand Rapids.

He and his coinvestigators conducted a retrospective analysis of 16,603 LTx recipients in the International Registry for Heart and Lung Transplantation (ISHLT) who underwent ECMO as their bridging strategy between 2005 and 2010. Centers were stratified into categories of low, medium, and high based on the volume of LTx procedures performed over the study interval: Low was defined as fewer than 25, medium as 25-50, and more than 50 as high volume.

Overall, 85 of the 16,603 transplant recipients in the study population were bridged via ECMO: 20 (23.5%) of them in low-volume centers, 30 (35.3%) in medium-volume centers, and 35 (41.2%) in high-volume centers. The researchers used Cox proportional hazard modeling to identify predictors of both 1- and 5-year survival rates, which were found to be significantly lower in low-volume centers – 13.61% at 5 years post LTx.

Looking at just the high-volume and low-volume centers, the researchers noted “significant differences” in both 1-year and 5-year survival rates when ECMO was used for bridging. One-year survival probability was roughly 40% in low-volume centers and roughly 70% in high-volume centers, while 5-year survival probability was well under 25% for recipients from low-volume centers and around 50% for those from high-volume centers (P = .0006). No significant differences existed for non-ECMO patients, regardless of center volume.

“No differences existed in survival in medium- and high-volume centers,” said Dr. Hayanga. “Transplanting without ECMO as a bridge showed fewer survival differences for both 1-year and 5-year survival. However, when ECMO was used as a bridge, the low-volume center [survival rates] were dramatically lower at both 1 year and 5 years.”

When Dr. Hayanga and his colleagues examined procedural volume as a continuous variable, however, a single inflection point was determined as the point at which survival outcomes steadily improve – 19 procedures. Centers that performed at least 19 LTx procedures between 2005 and 2010 experienced an uptick in survival rates, even though centers that saw 19-25 procedures were still considered low volume, the researchers noted.

“The corresponding c-statistic, however, is just under 60%,” cautioned Dr. Hayanga. “The C-statistic is a measure of the explanatory power of a variable – in this case, [center] volume – in accounting for the variability in outcome, or survival in this case. To put that number into context, a C-statistic of 50% means ‘no explanatory power’ whatsoever.”

Dr. Hayanga explained that he and his coauthors compared transplant recipient and donor characteristics using analysis of variance (ANOVA) and chi-square tests to compare variables, cumulative survival using Kaplan-Meier curves, and significance using log-rank tests.

Dr. Hayanga reported no financial conflicts of interest.

Next Article:

   Comments ()