The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Abstract & Commentary
Synopsis: Patients who undergo resection for lung cancer at high-volume hospitals are likely to have fewer complications and survive longer than those who have surgery at low-volume hospitals.
Source: Bach P, et al. The influence of hospital volume on survival after resection for lung cancer. N Engl J Med. 2001;345:181-188.
The study included patients who met the following requirements: 1) 65 years of age or older; 2) insured by Medicare; 3) diagnosis of primary cancer of the lung; 4) non-small-cell histology; 5) Stage I, II, or IIIA disease; 6) lung resection < 4 months after diagnosis; and 7) resident of an area covered by the Surveillance, Epidemiology, and End Results Program (SEER) and the Nationwide Inpatient Sample (NIS) databases. The NIS is a stratified random sample of 1012 hospitals in 22 states. Five of these states overlap with the SEER database, which includes 7 metropolitan areas (San Francisco, Oakland, San Jose, Detroit, Atlanta, Seattle, and Los Angeles County) and 5 states (Connecticut, Utah, New Mexico, Iowa, and Hawaii). Data were analyzed for 1 year (1997). The final sample included 2118 patients and 76 hospitals.
There was considerable variation in the volume of lung cancer resections performed at the 76 institutions. Thirty-four (45%) of the hospitals performed fewer than 9 procedures per institution. In contrast, 16 (21%) performed 20-66 procedures and 2 (3%) performed 67-100 procedures per institution. The volume of procedures performed at the hospital was positively associated with the survival of patients (P < 0.0001). Five years after surgery, 44% of patients who underwent operations at either of the 2 hospitals with the highest volume of procedures were alive, compared with 33% of those who underwent operation at the hospitals with fewer than 9 procedures performed. At the highest volume hospitals, patients also had a lower rate of postoperative complications (20% vs 44%) and lower 30-day mortality (3% vs 6%). In addition, the rate of 5-year survival among patients who underwent surgery at a teaching hospital was higher (42%) than for a nonteaching hospital (34%) (P < 0.001). Survival was improved at teaching vs. nonteaching hospitals regardless of volume.
Comment by Leslie A. Hoffman, RN, PhD
The major findings of this study were that: 1) rate of survival at 5 years was higher by 11 percentage points (44% vs 33%) among patients who underwent resections for lung cancer at hospitals with the highest volumes of such procedures than among those at the hospitals with the lowest volumes; 2) serious postoperative complications occurred at hospitals with the lowest volume twice as often as at those with the highest volume (44% vs 20%); and 3) survival was significantly better among patients who underwent surgery at teaching vs. nonteaching institutions, regardless of the number of procedures performed.
The association between volume (number of patients with a particular disease or procedure managed in a given hospital) and morbidity and mortality is not new. Prior studies have shown similar findings for patients diagnosed with other types of cancer (eg, breast, colon, and prostate cancer). In addition, a prior study has shown an association between increased volume and improved outcomes for elderly patients diagnosed with acute myocardial infarction.1 Thus, the volume of patients managed appears to have a major influence on morbidity and mortality, regardless of whether the underlying condition is medical or surgical. This finding is not surprising, given the learning curve associated with gaining the highest level of expertise. In essence, "practice makes perfect."
One response to these findings would be to recommend that patients be channeled to centers of excellence. This change would likely be impossible to implement and could have untoward consequences (eg, increasing surgical volume beyond capacity). A more productive approach might be to refer patients from very low-volume centers (eg, < 9 procedures per year), to a regional higher volume center. If that were accomplished, findings of this study suggest that the consequence could be a decrease in morbidity, mortality, and health care costs.
There are a number of limitations to this study. Only data included in the SEER, NIS, and Medicare databases could be used for analysis and the analysis was limited to 1 year. Hospitals were coded as either a teaching or a nonteaching institution. Given the complex organizational frameworks that exist today, this distinction was likely imperfect. Finally, the database did not provide complete information regarding adjuvant treatment, a factor that may have influenced survival.
1. N Engl J Med. 1999;340:1640-1648.