In-hospital mortality rates for liver resection patients may underestimate true, postoperative mortality by 50%, show the findings of a national study.
In addition, the data, from French hospitals, reveal that a large proportion of patients are treated in high-volume centers, indicating disparities in the availability of treatment.
"Our results may be of use in refining the number of specialist hepatobiliary surgeons, rationalizing the use of ancillary resources and reducing patient transportation," say Olivier Farges (Hôpital Beaujon, Paris, France) and colleagues.
Between 2007 and 2010, there were 28,708 elective hepatectomies recorded in the French national database of acute hospital admissions. Of these surgeries, 8529 were classified as major resections, and 22.4% were performed at the same time as another procedure.
The cumulative 90-day mortality was 5.8% for all hepatectomies and 5.0% among patients who had not undergone simultaneous procedures. In comparison, in-hospital mortality was 3.4% and 3.0%, respectively.
The authors found that the average rate of hepatectomies was low, at a median of four per year in each active hospital. However, over 53% of patients were operated on in hospitals performing more than 50 resections per year.
Surgery at a hospital with a volume of at least 11 hepatectomies per year was independently associated with a decreased risk for both in-hospital and 90-day mortality. For example, patients with liver-only resections had a 50% reduced odds of 90-day mortality if their surgery was performed at a center that averaged 11-25 surgeries per year, relative to being treated somewhere performing five or fewer each year.
The authors also found that the more hepatectomies a center performed, the more likely they were to perform major hepatectomies or to operate on primary liver tumors, such that 90% of major hepatectomies were performed in centers with an average of over 11 surgeries per year.
Writing in the Annals of Surgery, Farges and colleagues say their results indicate that patients at lower volume centers may be preferentially given nonsurgical treatment for their tumors, or that they may be denied surgery when their disease is too widespread.
They add that their finding that 30-day mortality massively underestimated true mortality supports a reconsideration of the way mortality is defined in oncology generally.
"Considering the increasingly long survival associated with oncological treatments, it might now make more sense to use 6- or 12-month mortality as a surrogate marker of the risks and benefits of surgery," they conclude.
Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.