Original Article
Is an activity volume threshold really realistic for lung cancer resection?
Abstract
Background: We analyzed volume as a continuous variable to estimate threshold, which is a methodology rarely seen in the literature. The objective of this work was to assess hospital volume for lung cancer (LC) surgery and to establish the associated threshold for acceptable in-hospital mortality (IHM). Data was obtained from the French national medico-administrative database.
Methods: From January 2005 to December 2016, data from 108,571 patients operated for LC in France were collected from the national administrative database. To estimate the volume threshold, hierarchical logistic regression models were developed.
Results: The crude IHM rate was 5.2% in low volume centers and 3.5% in high volume centers (P<0.0001). Centers performing more than 70 LC surgeries per year reduced the risk of postoperative death by 35% [adjusted odds ratio (OR): 0.65; 95% confidence interval (CI): 0.5–0.84]. Among the 4 models, the use of fractional polynomial of the volume had the lowest Akaike’s information criterion (AIC) index. The threshold volume was reached once a hospital’s annual volume reached 70 patients (95% CI, 40–85). In our analyses, the proportion of patients who were admitted in hospitals with an annual volume that was less than identified threshold were 34% of patients operated for LC. A hospital with an annual volume of 10 patients for lung resection, increasing the annual volume by 60 procedures would be associated with a 31% reduction in the odds of death within 30 days.
Conclusions: From the medico-administrative database, we have been able to estimate a minimum volume threshold that may be useful to help regionalize thoracic surgery centers.
Methods: From January 2005 to December 2016, data from 108,571 patients operated for LC in France were collected from the national administrative database. To estimate the volume threshold, hierarchical logistic regression models were developed.
Results: The crude IHM rate was 5.2% in low volume centers and 3.5% in high volume centers (P<0.0001). Centers performing more than 70 LC surgeries per year reduced the risk of postoperative death by 35% [adjusted odds ratio (OR): 0.65; 95% confidence interval (CI): 0.5–0.84]. Among the 4 models, the use of fractional polynomial of the volume had the lowest Akaike’s information criterion (AIC) index. The threshold volume was reached once a hospital’s annual volume reached 70 patients (95% CI, 40–85). In our analyses, the proportion of patients who were admitted in hospitals with an annual volume that was less than identified threshold were 34% of patients operated for LC. A hospital with an annual volume of 10 patients for lung resection, increasing the annual volume by 60 procedures would be associated with a 31% reduction in the odds of death within 30 days.
Conclusions: From the medico-administrative database, we have been able to estimate a minimum volume threshold that may be useful to help regionalize thoracic surgery centers.
