John Nicholas Melvan1, Onkar Khullar2, Sreekanth Vemulapalli3, Andrzej S Kosinski4, Allan Pickens2, Seth D Force2, Shuaiqi Zhang4, Manu S Sancheti2. 1. Division of Cardiothoracic Surgery, Holy Cross Hospital, Fort Lauderdale, Florida; Division of Cardiothoracic Surgery, Department of Surgery, Emory University School of Medicine, Atlanta, Georgia. Electronic address: johnnicholas.melvan@holy-cross.com. 2. Division of Cardiothoracic Surgery, Department of Surgery, Emory University School of Medicine, Atlanta, Georgia. 3. Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Division of Cardiology, Duke University Medical Center, Durham, North Carolina. 4. Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina.
Abstract
BACKGROUND: Socioeconomic factors play key roles in surgical outcomes. Socioeconomic data within The Society of Thoracic Surgeons (STS) General Thoracic Surgery Database (GTSD) are limited. Therefore, we utilized community size as a surrogate to understand socioeconomic differences in lung cancer resection outcomes. METHODS: We retrospectively reviewed all lung cancer resections from January 2012 to January 2017 in the STS GTSD. This captured 68,722 patients from 286 centers nationwide. We then linked patient zip codes with 2013 Rural-Urban Continuum Codes to understand the association between community size and postoperative outcomes. Demographic and clinical variables were evaluated for relationships with 30-day mortality, major morbidity, and readmission. RESULTS: Zip codes were included in 47.2% of patients. Zip-coded patients were older, were more comorbid, had less advanced disease, and were more commonly treated with minimally invasive approaches than were those without zip code classification. For geocoded patients, multivariable analyses demonstrated that sex, insurance payor, and hospital region were associated with all 3 major endpoints. Community size, based on Rural-Urban Continuum Codes coding, was not associated with any primary endpoint. Invasive mediastinal staging was related to morbidity, greater pathological stage predicted mortality, and worsened clinical stage was associated with readmission. More invasive surgery and greater extent of lung resection were associated with all primary endpoints. CONCLUSIONS: Incomplete data capture can promote selection bias within the STS GTSD and skew outcomes reporting. Moreover, community size is an insufficient surrogate, compared with sex, insurance payor, hospital region, for understanding socioeconomic differences in lung cancer resection outcomes.
BACKGROUND: Socioeconomic factors play key roles in surgical outcomes. Socioeconomic data within The Society of Thoracic Surgeons (STS) General Thoracic Surgery Database (GTSD) are limited. Therefore, we utilized community size as a surrogate to understand socioeconomic differences in lung cancer resection outcomes. METHODS: We retrospectively reviewed all lung cancer resections from January 2012 to January 2017 in the STS GTSD. This captured 68,722 patients from 286 centers nationwide. We then linked patient zip codes with 2013 Rural-Urban Continuum Codes to understand the association between community size and postoperative outcomes. Demographic and clinical variables were evaluated for relationships with 30-day mortality, major morbidity, and readmission. RESULTS: Zip codes were included in 47.2% of patients. Zip-coded patients were older, were more comorbid, had less advanced disease, and were more commonly treated with minimally invasive approaches than were those without zip code classification. For geocoded patients, multivariable analyses demonstrated that sex, insurance payor, and hospital region were associated with all 3 major endpoints. Community size, based on Rural-Urban Continuum Codes coding, was not associated with any primary endpoint. Invasive mediastinal staging was related to morbidity, greater pathological stage predicted mortality, and worsened clinical stage was associated with readmission. More invasive surgery and greater extent of lung resection were associated with all primary endpoints. CONCLUSIONS: Incomplete data capture can promote selection bias within the STS GTSD and skew outcomes reporting. Moreover, community size is an insufficient surrogate, compared with sex, insurance payor, hospital region, for understanding socioeconomic differences in lung cancer resection outcomes.
Authors: Kimberly J Song; Rowena Yip; Michael Chung; Qiang Cai; Yeqing Zhu; Ayushi Singh; Erik E Lewis; David Yankelevitz; Emanuela Taioli; Claudia Henschke; Raja Flores Journal: JTCVS Open Date: 2022-02-23