Introduction: The relationship between operation volume and quality has been an ongoing debate for years. With an anonymous questionnaire amongst members of the German Society of Visceral Surgery (DGVC) we try to describe the situation in Germany.
Methods: A questionnaire was sent to Surgeons-in-chief with questions about frequency and results of pancreas resections in the year 2006. Operations were divided in following volume categories: 1-5 operations (very low), 6-20 (low), 21-40 (medium) and more than 40 (high). Surgeon volume was defined as the total number of operations of each hospital divided by the number of participating surgeons.
Results: 262 of 520 hospitals returned the questionnaire (50.4 %). 40 hospitals (15.3%) did not perform pancreatic surgery. In the remaining 222 hospitals 5253 pancreatic resections were performed. 20 hospitals were very low volume institutions (9%), 114 low volume (51%), 48 medium (22%) und 40 high volume (18%). Comparing very low/low/medium volume hospitals (LVH) with high volume hospitals (HVH) there are 26.1% with more than 5% mortality vs. 20% in HVH (n.s.). However, above 7.5% mortality there are significant more LVH than HVH (17.7% vs. 5%, p=0,04). Considering surgeons volume, mortality is decreasing with higher volume (3.4 % mortality with < 5 operations per year, 3.2% with 6 - 10, 2.4 % with > 10 operations and 1.5 % with > 20). A single surgeon performs pancreatic surgery in 34.3% of all cases (mean mortality 2.2%). Two surgeons operate in 32.4% of hospitals with mean mortality of 2.8 %. With four or more surgeons mean mortality rises to 4.4%. Overall, mortality in university hospitals, teaching hospitals and others was 2.5%, 2.9% and 2.6%. Summary: Pancreatic surgery is performed in many hospitals outside university institutions in Germany. Mortality in high volume centers is lower compared with all other hospitals. However, 20% of HVH have mortality rates > 5%. Two Thirds of all operations are performed by one or two surgeons with a mean mortality of 2.5 %. Hospitals with four of more surgeons showed a higher mortality (4.4 %).
Conclusion: Hospital volume does not seem to influence mortality in pancreatic surgery in Germany. Quality in pancreatic surgery highly depends on specialized surgeons, whereas increase of mortality in centers with four or more surgeons is an effect of the training process. Based upon this questionnaire of the year 2006 no general conclusion can be drawn. However, the implementation of a national pancreatic registry seems to be a suitable way to further analyze the volume-quality relationship in Germany.