Re: Risk-Based Selective Referral for Cancer Surgery
We read with keen interest the article by Bilimoria et al1 published in Annals of Surgery. We are aware of the pioneering work that the authors have undertaken to investigate strategies to reduce postoperative mortality and improve surgical quality.2,3
We are pleased to see that the authors have categorized hospitals as specialized and community hospitals rather than using high volume or low volume as a surrogate marker for quality.
This study, in fact, makes it clear that volume may not have as much impact on improving outcomes as has been postulated and the volume-to-outcome relationship needs to be critically evaluated. It is evident that it is the quality of care at an individual hospital that matters. This is illustrated by the fact that low-volume community hospitals achieve similar postoperative results in low-risk patients. The improved outcomes in specialized centers for high-risk patients are most likely because of better facilities and clinical care pathways available at these institutes. Furthermore, a lack of difference in the perioperative outcomes in high-volume versus low-volume community hospitals highlights the flaw in using volume alone as a proxy for good quality.
Specialized centers invariably have high volumes and better outcomes, but the impact of many other factors is untested. The authors have acknowledged this and have noted that there are unmeasured structural features and clinical care processes that improve the outcomes at the specialized centers. If these are replicated at the community hospitals, equality of results may be achieved. Indeed, some low-volume community hospitals have excellent postoperative results in this study.
It is interesting, though not very surprising, that results for pancreatic and esophageal resections are significantly better at the specialized centers irrespective of the risk stratification. These surgeries are more complex than any of the other procedures, and care in specialized centers is therefore advisable. However, hospital volume alone should not be used to categorize a center as a specialized center, but rather quality-based outcomes of individual units should be utilized.
In Australia, we are faced with a unique situation where many tertiary specialized institutions are low-volume centers for cancer surgeries because of a low- and widely dispersed population. These centers possess the necessary infrastructure and the clinical care processes along with the surgical expertise that can categorize them as specialized centers. We therefore have valuable data on the performance of low-volume specialized centers.
We introduced specialized care processes for pancreatic and periampullary cancer patients at our low-volume tertiary referral center after 2 postoperative mortalities in 1999. The preoperative evaluation since then includes multidisciplinary meetings and decision making in conjunction with oncologists, gastroenterologists and radiologists. Anesthetic and intensive care physician advice is sought when indicated. The surgical procedure is standardized and all surgeries are performed by 2 experienced surgeons working together. Vascular surgeons are readily available, if required. Postoperative care is standardized and expert intensive care input is always utilized. Specialist imaging, interventional radiology and gastroenterology support is readily available at all times. We have a very low threshold for reexploration based on imaging and clinical observations. Using these strategies, we have performed 37 pancreaticoduodenectomies including 11 in patients older than 75 years without a single postoperative mortality in the last decade. Two major leaks have been treated successfully using a combination of interventional radiology, gastroenterology and surgical interventions with specialized intensive care support underpinning the importance of ready availability of such expertise.
We believe that adherence to such protocols along with the availability of necessary infrastructure are key to obtaining better postoperative outcomes.