See referenced original article on pages 752-9 this issue.
The patient behind the code, the care behind the patient†
Article first published online: 6 JAN 2009
Copyright © 2009 American Cancer Society
Volume 115, Issue 4, pages 702–703, 15 February 2009
How to Cite
Vauthey, J.-N. and Zorzi, D. (2009), The patient behind the code, the care behind the patient. Cancer, 115: 702–703. doi: 10.1002/cncr.24095
- Issue published online: 2 FEB 2009
- Article first published online: 6 JAN 2009
- Manuscript Accepted: 1 AUG 2008
- Manuscript Received: 15 JUL 2008
In the current issue of Cancer, Robertson et al1 analyze Medicare data between 2000 and 2004 from patients who had a diagnosis of primary colorectal cancer and subsequently underwent hepatic resection for colorectal liver metastases. Their study is important because it demonstrates that the results of liver resection in these Medicare patients were worse than current results from units that have surgical expertise. Although Robertson et al1 report a 5-year survival rate of 26%, recent single-center and multicenter studies from units with expertise indicate overall survival rates from 51% to 58%.2-4 The authors conclude that the survival rates reported from single-center series may be overly optimistic because of the small number of patients and follow-up limited to chart reviews.
In interpreting the results from this study, it is important to consider the patient-selection criteria. Patients were selected on the basis of International Classification of Diseases, 9th Revision (ICD-9) codes; however, with this selection method, many patient factors remain unknown. For instance, does a patient with ICD-9 code 197 have a first liver metastasis or a recurrent liver metastasis? In a patient with ICD-9 code 50.22 or 50.3, what was the margin status on pathologic examination of the surgical specimen? Did the patient have an R0 or R1 resection? What was the number of blood transfusions? Furthermore, coding problems may hide inappropriate patient selection or the complexity of a hepatic resection. Robertson et al stratified patients by procedure type (partial hepatectomy and hepatic lobectomy) in an attempt to account for the extent of surgery; however, other, unknown features of liver resection also may have a significant impact on outcome variables. Patient selection and chemotherapy factors cannot be accounted for when patients are selected on the basis of ICD-9 codes. It is now well established that current chemotherapy regimens (either combined 5‒fluorouracil, leucovorin, and oxaliplatin or combined leucovorin, 5‒fluorouracil, and irinotecan plus bevacizumab) have resulted in improved overall and progression-free survival for patients with colorectal metastases.5
The Medicare data as presented by Robertson et al do not reveal how many centers were included, how many procedures were performed at each center, or who performed the procedures—factors that have been shown to influence outcomes. Previous studies have demonstrated that high-volume regional providers can deliver complex care with better outcomes than low-volume centers.6-8 One study demonstrated specifically that hepatic resections performed at a high-volume center were associated with improved safety.9 The reasons for improved results at high-volume centers are multifactorial but most likely are related to the expertise of centers with multidisciplinary teams that perform a larger number of procedures. Improved outcomes at high-volume centers also may be related to the experience of the surgeons and anesthesiologists in the use of new techniques and equipment. A recent study indicated that specialized expertise in hepatobiliary surgery also affected perioperative outcome.10
Robertson et al should be congratulated for several strengths and interesting features of their study. First, their study was based on recent data (2000-2004) and therefore is more reflective of current practice than other studies that used older administrative datasets.11 Second, the authors report survival, whereas other studies on this topic have focused only on perioperative outcome.6-9 Third, Robertson et al report on 90-day mortality rather than the 30-day or in-hospital mortality reported in most surgical series. Ninety-day mortality increasingly is recognized as important, because a significant subset of patients (30%-50%) progress from hepatic insufficiency to liver failure after discharge.12, 13 A standardized definition of perioperative mortality after hepatectomy should include all deaths up to 90 days12, 13 to accurately reflect the true risk associated with major liver resection, which certainly is underestimated in contemporary reports of zero or near-zero mortality after hepatectomy.14, 15
Taken together, the authors' data suggest that disease site-oriented programs to improve outcomes nationally are warranted. Delivering a high standard of care with up-to-date expertise remains critical in this rapidly evolving field. We believe the management of colorectal cancer liver metastases requires a multidisciplinary approach involving not only experts in hepatobiliary surgery and medical oncology but an inclusive team with subspecialty expertise in diagnostic imaging,16 pathology,17 and interventional radiology.18
Conflict of Interest Disclosures
Dr. Vauthey has received grant support and honorarium from Sanofi-Aventis and has acted as a member of the Speakers' Bureau for Genentech.