Faiz Gani1, Joseph K Canner1, Timothy M Pawlik2. 1. Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, MD; Johns Hopkins Surgery Center for Outcomes Research, Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, MD. 2. Department of Surgery, Wexner Medical Center at the Ohio State University, Columbus, OH. Electronic address: tim.pawlik@osumc.edu.
Abstract
BACKGROUND: Variations in hospital billing practices may reflect differences in patient risk or may represent the "upcoding" of patients in response to payer incentives/policies. The current study sought to assess whether coding practices for gastrointestinal surgery have changed over time and to evaluate the association between upcoding and in-hospital costs. METHODS: A total of 1,344,152 patients aged >18 years undergoing a gastrointestinal operation between 2001 and 2011 were identified using the National Inpatient Sample. Coding practices were compared by hospital and patient characteristics. Multivariable analysis was performed to evaluate the association between coding practices and in-hospital costs. RESULTS: The mean and median number of codes per admission were 8.8 (standard deviation = 4.58) and 9 (interquartile range: 5-11), respectively. Over time, the proportion of admissions being upcoded (>9 codes/admission) increased from 14.1% to 32.9% (∆ = +133.3%, P < .001). This trend was observed for each gastrointestinal operation and was greatest for hepatectomy (∆ = +73.3%). Although admissions that were upcoded were more likely to be for patients with greater comorbidity and Medicare enrollees, an increase in the proportion of patients upcoded was also observed regardless of the primary payer, among patients presenting without comorbidity, and among patients undergoing an elective operation (all P < .001). On adjusted analysis, admissions that were upcoded were independently associated with a $13,754 (95% confidence interval: $13,638-$13,870) greater in-hospital cost. CONCLUSION: The number of "upcoded" patients was observed to increase with time. Greater education, regulation, and scrutiny are required of coding practices.
BACKGROUND: Variations in hospital billing practices may reflect differences in patient risk or may represent the "upcoding" of patients in response to payer incentives/policies. The current study sought to assess whether coding practices for gastrointestinal surgery have changed over time and to evaluate the association between upcoding and in-hospital costs. METHODS: A total of 1,344,152 patients aged >18 years undergoing a gastrointestinal operation between 2001 and 2011 were identified using the National Inpatient Sample. Coding practices were compared by hospital and patient characteristics. Multivariable analysis was performed to evaluate the association between coding practices and in-hospital costs. RESULTS: The mean and median number of codes per admission were 8.8 (standard deviation = 4.58) and 9 (interquartile range: 5-11), respectively. Over time, the proportion of admissions being upcoded (>9 codes/admission) increased from 14.1% to 32.9% (∆ = +133.3%, P < .001). This trend was observed for each gastrointestinal operation and was greatest for hepatectomy (∆ = +73.3%). Although admissions that were upcoded were more likely to be for patients with greater comorbidity and Medicare enrollees, an increase in the proportion of patients upcoded was also observed regardless of the primary payer, among patients presenting without comorbidity, and among patients undergoing an elective operation (all P < .001). On adjusted analysis, admissions that were upcoded were independently associated with a $13,754 (95% confidence interval: $13,638-$13,870) greater in-hospital cost. CONCLUSION: The number of "upcoded" patients was observed to increase with time. Greater education, regulation, and scrutiny are required of coding practices.