Complication timing and association with mortality in the American College of Surgeons' National Surgical Quality Improvement Program database

• Published in: The Journal of surgical research 2015, Vol.193 (1), p.77-87
• Main Authors: Wakeam, Elliot, MD, Hyder, Joseph A., MD, PhD, Tsai, Thomas C., MD, MPH, Lipsitz, Stuart R., ScD, Orgill, Dennis P., MD, PhD, Finlayson, Sam R.G., MD, MPH
• Format: Article
• Language: English
• Subjects: Adult; Aged; Cerebrovascular Disorders - mortality; Complications; Databases, Factual - statistics & numerical data; Failure-to-rescue; General Surgery - education; General Surgery - statistics & numerical data; Heart Arrest - mortality; Humans; Inpatients - statistics & numerical data; Middle Aged; Myocardial Infarction - mortality; Pneumonia - mortality; Postoperative Complications - mortality; Proportional Hazards Models; Quality Improvement - statistics & numerical data; Surgery; Surgical education; Surgical quality; Surgical Wound Infection - mortality; Time Factors; United States
• ISSN: 0022-4804; 1095-8673
• DOI: 10.1016/j.jss.2014.08.025

Abstract Background The relationship between timing of postoperative complications on mortality is unknown. We investigated the time-variable mortality risks of common surgical complications. Methods We identified patients undergoing nonemergent, in-patient surgery in the National Surgical Quality Improvement Program (NSQIP) database during 2005–2011 who experienced any of 13 complications within 2 wk of surgery. “Expected timing” was defined as the median postoperative day of occurrence. Hazard ratios (HRs) for complications earlier or later than expected were calculated using Cox proportional hazards, adjusted for age, procedure, American Society of Anesthesiology (ASA), and functional status. A secondary analysis evaluated the effect of preceding complication burden on the relationship between complication timing and mortality. Results Among 77,443 patients experiencing complications, significantly higher mortality was observed with early wound infections (superficial HR 1.30, confidence interval [CI] 1.01–1.70; deep HR 1.52, CI 1.07–2.16; and organ space HR 1.38, CI 1.11–1.70) despite adjustment for patient and operative factors and complication burden. Early cardiac arrest and unplanned intubation were associated with lower mortality, which persisted after adjustment (HR 0.59, CI 0.51–0.68; HR 0.38, CI 0.33–0.43, respectively). By contrast, late occurrence of acute myocardial infarction, pneumonia, and cerebrovascular accident was associated with significantly greater mortality risk (HR 1.41, CI 1.18–1.69; HR 1.37, CI 1.24–1.52; and HR 1.61, CI 1.31–1.98, respectively), but these associations became nonsignificant after adjustment for complication burden. Conclusions Timing of complications plays an important role in mortality. Surgeons and trainees should be aware of these patterns and tailor their clinical care and monitoring practices to account for the implications of complication timing on mortality.