Francois Cauchy1, Olivier Farges, Eric Vibert, Emmanuel Boleslawski, Francois René Pruvot, Jean Marc Regimbeau, Jean Yves Mabrut, Olivier Scatton, Mustapha Adham, Christophe Laurent, Emilie Grégoire, Jean Robert Delpero, Philippe Bachellier, Olivier Soubrane. 1. *Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Beaujon, Assistance Publique Hôpitaux de Paris, Université Paris 7, Clichy, France †Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Paul Brousse, Villejuif, France ‡Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Claude Huriez, Lille, France §Department of Hepatobiliopancreatic Surgery, Hôpital Nord, Amiens, France ¶Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Croix Rousse, Lyon, France ||Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Saint Antoine, Paris, France **Department of Hepatobiliopancreatic Surgery, Hôpital Edouard Herriot, Lyon, France ††Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital Saint André, Bordeaux, France ‡‡Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital de la Conception, Marseille, France §§Department of Surgical Oncology, Institut Paoli-Calmettes, Marseille, France ¶¶Department of Hepatobiliopancreatic Surgery and Liver Transplantation, Hôpital de Hautepierre, Hôpitaux Universitaires de Strasbourg, Université de Strasbourg, Strasbourg, France.
Abstract
OBJECTIVE: Investigate if involving surgeons in outcome prediction-research and having them use a dedicated Electronic-Health-Record that provides feedback, improves patients' outcome. BACKGROUND: Improvement of clinical outcome mainly relies on the declaration of adverse events, identification of their predictors, self-assessment, and feedback. METHODS: Thirteen French Hepato-Pancreato-Biliary-centers made commitment to include all patients undergoing elective hepatectomies in an observational study. Each center was given access to a dedicated website, where perioperative data were prospectively collected. The website provided real-time individual and comparative feedback of outcome and was also intended to perform prognostication studies. The hypothesis was that by using this strategy, the length-of-stay would be reduced by 10%. Power-calculation implied the inclusion of 1720 patients. Secondary endpoints were 90-day mortality, severe morbidity, and the comprehensive-complication index. RESULTS: Only 5 of the 13 participating centers were fully compliant in enrolling their patients and the inclusion period was extended by 1-year (October 2012-October 2015) to meet the objective. During this period, the collaborative group published 9 studies based on the study data (median impact factor = 8.327) that identified quantitative clinical variables, qualitative clinical variables, and nonclinical variables influencing outcome. For patients enrolled by the 5 active centers (n = 1752), there was no improvement in length of stay (13.3 vs 12.4 days, P = 0.287), severe complications (23.6 vs 20.5%, P = 0.134), the complication comprehensive index (24.0 vs 24.9, P = 0.448), mortality (4.1 vs 3.9%, P = 0.903), or unplanned readmissions (7.2 vs 8.4%, P = 0.665), even after adjusting for confounders. CONCLUSION: Simply sensitizing surgeons to their outcome has no measurable short-term clinical benefit.
OBJECTIVE: Investigate if involving surgeons in outcome prediction-research and having them use a dedicated Electronic-Health-Record that provides feedback, improves patients' outcome. BACKGROUND: Improvement of clinical outcome mainly relies on the declaration of adverse events, identification of their predictors, self-assessment, and feedback. METHODS: Thirteen French Hepato-Pancreato-Biliary-centers made commitment to include all patients undergoing elective hepatectomies in an observational study. Each center was given access to a dedicated website, where perioperative data were prospectively collected. The website provided real-time individual and comparative feedback of outcome and was also intended to perform prognostication studies. The hypothesis was that by using this strategy, the length-of-stay would be reduced by 10%. Power-calculation implied the inclusion of 1720 patients. Secondary endpoints were 90-day mortality, severe morbidity, and the comprehensive-complication index. RESULTS: Only 5 of the 13 participating centers were fully compliant in enrolling their patients and the inclusion period was extended by 1-year (October 2012-October 2015) to meet the objective. During this period, the collaborative group published 9 studies based on the study data (median impact factor = 8.327) that identified quantitative clinical variables, qualitative clinical variables, and nonclinical variables influencing outcome. For patients enrolled by the 5 active centers (n = 1752), there was no improvement in length of stay (13.3 vs 12.4 days, P = 0.287), severe complications (23.6 vs 20.5%, P = 0.134), the complication comprehensive index (24.0 vs 24.9, P = 0.448), mortality (4.1 vs 3.9%, P = 0.903), or unplanned readmissions (7.2 vs 8.4%, P = 0.665), even after adjusting for confounders. CONCLUSION: Simply sensitizing surgeons to their outcome has no measurable short-term clinical benefit.