Geoffrey A Anderson1, Jordan Bohnen2, Richard Spence3, Lenka Ilcisin4, Karim Ladha5, David Chang2. 1. Massachusetts General Hospital, GRB 425, 55 Fruit St, Boston, MA, 02114, USA. Geoffrey.Anderson@mail.harvard.edu. 2. Massachusetts General Hospital, GRB 425, 55 Fruit St, Boston, MA, 02114, USA. 3. University of Cape Town, Cape Town, South Africa. 4. Brigham and Women's Hospital, Boston, MA, USA. 5. Toronto General Hospital and University of Toronto, Toronto, ON, Canada.
Abstract
BACKGROUND: The focus of many data collection efforts centers on creation of more granular data. The assumption is that more complex data are better able to predict outcomes. We hypothesized that data are often needlessly complex. We sought to demonstrate this concept by examination of the American Society of Anesthesiologists (ASA) scoring system. METHODS: First, we created every possible consecutive two, three and four category combinations of the current five category ASA score. This resulted in 14 combinations of simplified ASA. We compared the predictive ability of these simplified scores for postoperative outcomes for 2.3 million patients in the NSQIP database. Individual model performance was assessed by comparing receiver operator characteristic (ROC) curves for each model with the standard ASA. RESULTS: Two of our 4-category models and one of our 3-category models had ability to predict all outcomes equivalent to standard ASA. These results held for all outcomes and on all subgroups tested. The performance of the three best performing simplified ASA scores were also equivalent to the standard ASA score in the univariate analysis and when included in a multivariate model. CONCLUSIONS: It is assumed that the most granular data and use of the largest number of variables for risk-adjusted predictions will increase accuracy. This complexity is often at the expense of utility. Using the single best predictor in surgical outcomes research, we have shown this is not the case. In this example, we demonstrate that one can simplify ASA into a 3-category variable without losing any ability to predict outcomes.
BACKGROUND: The focus of many data collection efforts centers on creation of more granular data. The assumption is that more complex data are better able to predict outcomes. We hypothesized that data are often needlessly complex. We sought to demonstrate this concept by examination of the American Society of Anesthesiologists (ASA) scoring system. METHODS: First, we created every possible consecutive two, three and four category combinations of the current five category ASA score. This resulted in 14 combinations of simplified ASA. We compared the predictive ability of these simplified scores for postoperative outcomes for 2.3 million patients in the NSQIP database. Individual model performance was assessed by comparing receiver operator characteristic (ROC) curves for each model with the standard ASA. RESULTS: Two of our 4-category models and one of our 3-category models had ability to predict all outcomes equivalent to standard ASA. These results held for all outcomes and on all subgroups tested. The performance of the three best performing simplified ASA scores were also equivalent to the standard ASA score in the univariate analysis and when included in a multivariate model. CONCLUSIONS: It is assumed that the most granular data and use of the largest number of variables for risk-adjusted predictions will increase accuracy. This complexity is often at the expense of utility. Using the single best predictor in surgical outcomes research, we have shown this is not the case. In this example, we demonstrate that one can simplify ASA into a 3-category variable without losing any ability to predict outcomes.
Authors: Caleb R Schultz; Henri R Ford; Laura D Cassidy; Barbara L Shultz; Christian Blanc; Leslie W King-Schultz; Henry B Perry Journal: J Trauma Date: 2007-11
Authors: Shukri F Khuri; William G Henderson; Jennifer Daley; Olga Jonasson; R Scott Jones; Darrell A Campbell; Aaron S Fink; Robert M Mentzer; Leigh Neumayer; Karl Hammermeister; Cecilia Mosca; Nancy Healey Journal: Ann Surg Date: 2008-08 Impact factor: 12.969
Authors: Daniel B Sonshine; Jesse Shantz; Raphael Kumah-Ametepey; R Richard Coughlin; Richard A Gosselin Journal: World J Surg Date: 2013-07 Impact factor: 3.352
Authors: John G Meara; Andrew J M Leather; Lars Hagander; Blake C Alkire; Nivaldo Alonso; Emmanuel A Ameh; Stephen W Bickler; Lesong Conteh; Anna J Dare; Justine Davies; Eunice Dérivois Mérisier; Shenaaz El-Halabi; Paul E Farmer; Atul Gawande; Rowan Gillies; Sarah L M Greenberg; Caris E Grimes; Russell L Gruen; Edna Adan Ismail; Thaim Buya Kamara; Chris Lavy; Ganbold Lundeg; Nyengo C Mkandawire; Nakul P Raykar; Johanna N Riesel; Edgar Rodas; John Rose; Nobhojit Roy; Mark G Shrime; Richard Sullivan; Stéphane Verguet; David Watters; Thomas G Weiser; Iain H Wilson; Gavin Yamey; Winnie Yip Journal: Lancet Date: 2015-04-26 Impact factor: 79.321