J. Li1, E. Chang1, N. C. Wang1, D. Cron1, P. Zhang1, S. C. Wang1 1University Of Michigan,Ann Arbor, MI, USA
Introduction: Acute kidney injury (AKI) occurs in approximately 20% of hospitalized adults in the US and is associated with increased morbidity, mortality, and costs for patients of general surgery procedures. Numerous studies have suggested the effectiveness of using patient CT imaging data to predict risk of AKI. Analytic Morphomics, a quantitative approach to processing cross-sectional imaging data, has shown utility in evaluating perioperative risk and postoperative outcome. Thus, we hypothesized that individualized morphomic factors would be significantly associated with AKI after major general surgery procedures.
Methods: 326 adult patients undergoing major, elective, intra-abdominal general surgery operations from January 2008 to September 2011 were selected. Patients undergoing isolated appendectomy or cholecystectomy procedures were excluded, as were patients with a preoperative creatinine value of 1.6 or greater, a history of ESRD, or a prior nephrectomy. Preoperative CT scans within 365 days prior to the operation were analyzed for kidney, vertebral, and fascia-associated parameters such as length, area, volume, and radiodensity. Demographic factors included patient age, gender, and BMI. Univariate analysis was utilized to determine the strength of the association between patient morphomic and demographic factors and development of postoperative AKI as defined by the KDIGO criteria within 30 days of the surgery.
Results: Of the 238 eligible patients, 31 (13%) developed AKI. Average pixel radiodensity of the total kidney (p<0.001) and kidney parenchyma (p<0.001) were found to be significantly lower for the group that developed postoperative AKI than for those that did not. No significant differences between AKI and non-AKI patients were found for total kidney volume (p=0.19), parenchymal volume (p=0.16), anterior body depth (p=0.32), and patient demographic factors (p>0.15). Table 1 contains the results of the univariate analysis including the means between the AKI and non-AKI group for several factors.
Conclusion: For patients undergoing noncardiac, nonvascular surgical procedures, our analysis showed that kidney pixel radiodensity factors are the most significant in predicting postoperative AKI risk.
