The Risk of AKI Varies by Type of Procedure
After accounting for procedure and comorbidities, the relative risks for developing AKI were determined for the CCS-SP categories (Table 2). Using colorectal resection as the reference category, the aRR for AKI ranged from 0.23 in appendectomy to 1.34 in ileostomy, reflecting a nearly 6-fold difference in the adjusted risk of AKI in the highest risk category compared with the lowest risk category. The strongest predictor of AKI was the severity of preoperative renal dysfunction: compared with those with normal renal function (eGFR >90), the aRRs (95% CI) for those with stage 5 (eGFR <15), stage 4 (eGFR 15–30), stage 3 (eGFR 30–60), and stage 2 (eGFR 60–90) CKD were 7.80 (6.35–9.57), 5.10 (4.56–5.69), 2.54 (2.33–2.78), and 1.24 (1.14–1.35), respectively. There was no significant difference in AKI risk comparing those with missing creatinine data to those with normal renal function (aRR 0.97 [0.80–1.18]), indicating that, on average, those with missing creatinine data likely had normal renal function or only mild renal dysfunction. Indeed, with multiple imputation20 of those with missing values, 37% were imputed to eGFR > 90 and 29% to eGFR 60–90. Multiple imputation analysis and complete case analysis did not result in any clinically meaningful changes in the interpretation of our results (Supplemental Digital Content, http://links.lww.com/AA/A999).
Other strong predictors of AKI were hypertension (aRR 1.50 [1.40–1.61]), BMI (aRR 1.97 [1.80–2.15] in BMI > 39 versus BMI < 29; aRR 1.26 [1.19–1.35] in BMI 29–39 versus BMI < 29), ascites (aRR 1.50 [1.35–1.66]), and preoperative sepsis (aRR 1.52 [1.40–1.64]). Female sex and laparoscopic procedure were associated with a reduced risk of AKI with aRRs of 0.54 (0.51–0.57) and 0.52 (0.47–0.58), respectively. Additional significant predictors of AKI were emergency surgery, functional dependence, ventilator dependence, dyspnea, chronic obstructive pulmonary disease, current smoking, diabetes, congestive heart failure, myocardial infarction, bleeding disorders, hematocrit, chronic steroid use, and cancer (aRR < 1.50; not shown).
The predicted marginal risk for AKI varied tremendously among procedure categories, ranging from 0.1% in appendectomy and 0.3% in gastric bypass to 1.5% in both ileostomy and small bowel resection (Table 2).
The Effect of AKI on 30-Day Mortality
In our cohort, 9988 (2.2%) subjects died within 30 days of their procedure, with a mortality of 1.9% in patients who did not develop AKI compared with 31% in those who developed AKI (Table 3). The lowest mortality was observed in patients undergoing appendectomy and gastric bypass (0.2%) and the highest mortality in patients undergoing exploratory laparotomy (12.4%). In adjusted analysis, there was a 5-fold difference in the risk of mortality between the highest and lowest risk procedures.
After adjusting for procedure, comorbidities, and intraoperative factors, the aRR (95% CI) for 30-day mortality with AKI in the overall sample was 3.51 (3.29–3.74), indicating that AKI was a strong predictor of 30-day mortality in this cohort. Other important predictors of mortality were age (aRR 1.044 [1.042–1.046] for each additional year), history of cancer (aRR 2.30 [2.18–2.42]), functional dependence (aRR 2.18 [2.07–2.30]), preoperative sepsis (aRR 1.86 [1.77–1.97]), ascites (aRR 1.74 [1.63–1.85]), emergency (aRR 1.56 [1.49–1.64]), and perioperative transfusion (aRR 1.53 [1.46–1.61]). In addition, dyspnea, ventilator dependence, chronic obstructive pulmonary disease, current smoking, peripheral vascular disease, stroke, bleeding disorders, chronic steroid use, eGFR, and preoperative hematocrit were also significantly associated with 30-day mortality (aRR < 1.50; not shown). Multiple imputation analysis and complete case analysis did not result in any clinically meaningful changes in the interpretation of our results (not shown).
After stratification based on procedure type, there was significant variation in the aRR for 30-day mortality with AKI, ranging from 1.87 (1.62–2.17) in exploratory laparotomy to 31.6 (17.9–55.9) in gastric bypass (Table 3). When further adjusted for other postoperative complications, the aRR for 30-day mortality in the overall sample was 2.05 (1.90–2.22). In addition, the variation in aRR for 30-day mortality with AKI was reduced, ranging from 1.49 (1.00–2.20) in excision and lysis of peritoneal adhesions to 7.02 (3.25–15.2) in appendectomy.
Based on our multivariable model for 30-day mortality without other postoperative complications, the predicted marginal risk of death without AKI ranged from <0.01% (gastric bypass) to 5.5% (exploratory laparotomy) (Table 4). With AKI, the predicted marginal risk of death ranged from 6.2% (other OR lower GI therapeutic procedures) to 29.5% (exploratory laparotomy). When other postoperative complications were accounted for in the analysis, the predicted marginal risk of death without AKI did not materially change. However, the predicted marginal risk of death with AKI changed in many procedures, such as a reduction from 8.0% to 0.8% in gastric bypass. There was a negligible change in predicted marginal risk in a few procedure categories, such as exploratory laparotomy and other OR lower GI therapeutic procedures.
Using a large high-quality national dataset of surgical outcomes, we assessed the procedure-specific risk of perioperative AKI for patients undergoing intraabdominal general surgery, a group that is at high risk for developing AKI.4 Consistent with prior published results, the overall rate of perioperative AKI in our cohort was 1.1%.3,4 However, this rate varied tremendously among the 15 categories of surgical procedures analyzed, ranging from 0.2% in patients undergoing appendectomy to 3.5% in patients undergoing exploratory laparotomy, with a 6-fold difference in the adjusted risk of AKI between the highest and lowest risk procedures. The predicted marginal risk of AKI varied from 0.1% in appendectomy to 1.5% in ileostomy. Only information known before the surgery was used in this analysis, so this model is appropriate for preoperatively assessing AKI risk.
We also evaluated the effect of AKI on 30-day mortality, and after adjusting for procedure type and other factors, AKI was associated with a 3.5-fold increase in the risk of 30-day mortality. However, there was tremendous variability in the adjusted risk of AKI on 30-day mortality within the procedure categories, with the aRR ranging from 1.80 in exploratory laparotomy to 31.6 in gastric bypass. The predicted marginal risk of 30-day mortality with AKI ranged from 6.2% in other OR lower GI therapeutic procedures to 29.5% in exploratory laparotomy. These estimates of the aRRs and predicted marginal risks likely represent the upper bounds of the effects of AKI on mortality because the estimates were reduced when accounting for other adverse outcomes. Taken together, our results demonstrate that intraabdominal general surgery cannot be treated as a single category, and the specific procedure must be accounted for when evaluating the risk of AKI in noncardiac general surgery procedures.
Although intraperitoneal surgery is identified as being high risk for AKI3,4 and other complications,21 this group consists of a diverse collection of procedures, and considerable variation in the risk of complications should be expected. The highest risk of AKI was observed in small intestinal procedures (ileostomy and small bowel resection), colorectal procedures (colorectal resection), hepatobiliary and pancreatic procedures (other OR GI procedures), splenic procedures, and exploratory laparotomy (Table 2). Exploratory laparotomies are typically emergent cases (44% in our sample) in acute settings with high rates of associated morbidity. A high proportion of small bowel resections (40%) were also emergent, likely reflecting acute disease states such as small bowel obstruction22 or acute mesenteric ischemia.23 In addition, preclinical studies suggest that intestinal ischemia–reperfusion injury leads to multiorgan dysfunction and systemic inflammation24 and that the gut is the “motor” of systemic inflammation.25 Manipulation and resection of the small intestine may result in an inflammatory response that leads to AKI. Patients undergoing colorectal resection (22% of total sample) often undergo mechanical bowel preparation despite its questionable clinical utility,26 and this practice may have adverse physiologic consequences affecting their risk of AKI.27
Procedures with the lowest adjusted risk of AKI include cholecystectomy, gastric bypass, and appendectomy. Despite the high rate of emergency procedures with acute inflammation (appendicitis and cholecystitis) in this category, these procedures are usually performed on younger, healthier patients, and it is not surprising that they are associated with the lowest risk of AKI.
In addition to the specific procedure, the strongest predictor of perioperative AKI in our study was preexisting renal dysfunction. Although we excluded patients with preoperative acute renal failure or dialysis, we retained those who had CKD without acute changes in renal function. As expected, stage 5 (eGFR <15) and stage 4 (eGFR 15–30) CKD were associated with an 8-fold and 5-fold increase in AKI risk, respectively, compared with those with normal renal function (eGFR >90). Stage 3 CKD (eGFR 30–60) was associated with a 2.5-fold in the risk of AKI, and even stage 2 CKD (eGFR 60–90) was associated with a significant increase in the risk of AKI (1.24-fold). It should be noted that 58% of patients in the sample had stage 2 CKD or worse, indicating that the majority of patients presenting for inpatient intraabdominal general surgery had at least 1 significant risk factor for the development of AKI.
The known risk factors for AKI were consistent in their associations with increased AKI across the different procedures in stratified analyses (not shown). Impaired preoperative renal function (eGFR), sepsis, hypertension, and ascites were all associated with an increased risk of AKI in the majority of procedures. However, there were some exceptions. For instance, there was no association between emergency status and AKI in appendectomy (>77% emergent) and gastric bypass (<1% emergent), both procedures with low risks of AKI. These findings underscore the need to evaluate risk factors in the context of the patient population being examined.
AKI appears to have the greatest effect on 30-day mortality in procedures where the rates of AKI and mortality are relatively low, such as gastric bypass and appendectomy. It is clear that the effect of AKI on mortality varies dramatically based on the underlying patient population and characteristics of each procedure, and the reasons for these differential effects will need to be investigated. In addition, AKI may be a spectrum of diseases with different causes and consequences that depend on the clinical context in which it occurs. For instance, AKI associated with gastric bypass (generally younger patients with fewer comorbidities) may have different etiologies and mechanisms of illness than AKI associated with small bowel resection, an older patient population with multiple risk factors and comorbidities.
Initially, AKI was seen simply as a marker of illness and not a cause,28 but mounting epidemiologic evidence suggests that AKI contributes directly to mortality, at least in patients with severe AKI requiring renal replacement therapy.29 Potential mechanisms for AKI-associated morbidity include fluid overload, acid–base disturbances, inflammation and multiorgan dysfunction, as well as inadequate antimicrobial therapy.29 However, the specific mechanisms involved remain unclear, and despite all of the available research, no known therapies reduce the incidence of perioperative AKI.30 AKI may contribute to mortality alone, but more likely, it is part of a cascade of events leading to systemic inflammation and multiorgan dysfunction.31 Indeed, our analysis demonstrates that other postoperative complications, such as sepsis and myocardial infarction, at least partially explain the apparent associations between AKI and mortality.
As with any observational study of large datasets,32 this study is subject to some limitations. The study of AKI has historically been limited by varying definitions of renal failure,33 and ACS-NSQIP only identifies patients with severe changes in creatinine (>2 mg/dL above baseline) or dialysis as having developed AKI. Creatinine changes as small as 0.3 mg/dL may constitute clinically significant AKI,8,16 and the ACS-NSQIP definition of AKI may thus underestimate the overall incidence of AKI in the NSQIP cohort. Although we account for only the most severe cases of AKI, the large sample size and quality of the ACS-NSQIP data provide the opportunity to obtain valuable insights on variations in the risk of AKI among the different categories of intraabdominal surgery. Factors such as intraoperative fluid management34,35 and the use of vasopressors and diuretics3 may affect AKI rates, but the dataset does not report this information.
Categorization of individual procedures into larger groups may lead to biased parameter estimates for the other independent variables.36 However, when we incorporated the individual procedure in our models, the parameter estimates did not change in a meaningful way (Supplemental Digital Content, http://links.lww.com/AA/A999), and our findings were robust to the classification of procedures.c Additionally, modeling the actual procedure instead of the scheduled procedure may lead to misclassification bias.36 Unfortunately, we do not have data on the scheduled procedure, and although we do not expect significant miscategorization of procedures, this is a limitation of our analysis. Despite these limitations, our analysis serves to highlight important aspects of the epidemiology of perioperative AKI.
In conclusion, we have demonstrated that among patients undergoing intraabdominal general surgery procedures, the risk of AKI and 30-day mortality varies considerably depending on the specific procedure. Prior studies identify intraabdominal surgery as a risk factor for adverse outcomes, but it is clear that the risk of adverse outcomes is not uniformly distributed among this group. These results highlight the importance of preoperative risk stratification and identify procedure type as a significant risk factor for AKI and 30-day mortality.
Name: Minjae Kim, MD, MS.
Contribution: This author was involved in study design, conduct of the study, data analysis, and manuscript preparation.
Attestation: Minjae Kim approved the final manuscript and attests to the integrity of the original data and analysis reported in this manuscript. Dr. Kim is the archival author.
Name: Joanne E. Brady, PhD.
Contribution: This author was involved in study design, data analysis, and manuscript preparation.
Attestation: Joanne E. Brady approved the final manuscript and attests to the integrity of the original data and analysis reported in this manuscript.
Name: Guohua Li, MD, DrPH.
Contribution: This author was involved in study design, data analysis, and manuscript preparation.
Attestation: Guohua Li approved the final manuscript and attests to the integrity of the original data and analysis reported in this manuscript.
This manuscript was handled by: Avery Tung, MD.
a The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
b Available at: http://www.hcup-us.ahrq.gov/toolssoftware/ccs_svcsproc/ccssvcproc.jsp. Accessed June 15, 2012.
c We used a random effects (random intercept) logistic regression model, first with the CCS category of procedure as the random effect and then with the individual CPT codes as the random effect. The variances (on the logit scale) were 0.6608 (SE: 0.2527) by CCS code and 0.7158 (SE: 0.0907) by CPT. Absence of significant differences in the variances demonstrates that our findings are robust to the classification of procedures.
1. Kumar AB, Suneja M. Cardiopulmonary bypass-associated acute kidney injury. Anesthesiology. 2011;114:964–70
2. Wald R, Waikar SS, Liangos O, Pereira BJG, Chertow GM, Jaber BL. Acute renal failure after endovascular vs open repair of abdominal aortic aneurysm. J Vasc Surg. 2006;43:460–6
3. Kheterpal S, Tremper KK, Englesbe MJ, O’Reilly M, Shanks AM, Fetterman DM, Rosenberg AL, Swartz RD. Predictors of postoperative acute renal failure after noncardiac surgery in patients with previously normal renal function. Anesthesiology. 2007;107:892–902
4. Kheterpal S, Tremper KK, Heung M, Rosenberg AL, Englesbe M, Shanks AM, Campbell DA Jr.. Development and validation of an acute kidney injury risk index for patients undergoing general surgery: results from a national data set. Anesthesiology. 2009;110:505–15
5. Goldman L, Caldera DL, Nussbaum SR, Southwick FS, Krogstad D, Murray B, Burke DS, O’Malley TA, Goroll AH, Caplan CH, Nolan J, Carabello B, Slater EE. Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med. 1977;297:845–50
6. Rahbari NN, Knebel P, Diener MK, Seidlmayer C, Ridwelski K, Stöltzing H, Seiler CM. Current practice of abdominal wall closure in elective surgery: is there any consensus? BMC Surg. 2009;9:8
7. Borthwick E, Ferguson A. Perioperative acute kidney injury: risk factors, recognition, management, and outcomes. BMJ. 2010;341:c3365
8. Bihorac A, Yavas S, Subbiah S, Hobson CE, Schold JD, Gabrielli A, Layon AJ, Segal MS. Long-term risk of mortality and acute kidney injury during hospitalization after major surgery. Ann Surg. 2009;249:851–8
9. Chertow GM, Burdick E, Honour M, Bonventre JV, Bates DW. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol. 2005;16:3365–70
10. Fink AS, Campbell DA Jr, Mentzer RM Jr, Henderson WG, Daley J, Bannister J, Hur K, Khuri SF. The National Surgical Quality Improvement Program in non-Veterans Administration hospitals: initial demonstration of feasibility. Ann Surg. 2002;236:344–53
11. Hua M, Brady J, Li G. The epidemiology of upper airway injury in patients undergoing major surgical procedures. Anesth Analg. 2012;114:148–51
13. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, Kusek JW, Van Lente FChronic Kidney Disease Epidemiology Collaboration. . Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–54
14. National Kidney Foundation. . K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:S1–266
15. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care. 2004;8:R204–12
16. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11:R31
17. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol. 2005;162:199–200
18. Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–6
19. Raghunathan TE, Lepkowski JM, Van Hoewyk J, Solenberger P. A multivariate technique for multiply imputing missing values using a sequence of regression models. Surv Methodol. 2001;27:85–95
20. van der Heijden GJ, Donders AR, Stijnen T, Moons KG. Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: a clinical example. J Clin Epidemiol. 2006;59:1102–9
21. Lee TH, Marcantonio ER, Mangione CM, Thomas EJ, Polanczyk CA, Cook EF, Sugarbaker DJ, Donaldson MC, Poss R, Ho KK, Ludwig LE, Pedan A, Goldman L. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation. 1999;100:1043–9
22. Oyasiji T, Angelo S, Kyriakides TC, Helton SW. Small bowel obstruction: outcome and cost implications of admitting service. Am Surg. 2010;76:687–91
23. Acosta-Merida MA, Marchena-Gomez J, Hemmersbach-Miller M, Roque-Castellano C, Hernandez-Romero JM. Identification of risk factors for perioperative mortality in acute mesenteric ischemia. World J Surg. 2006;30:1579–85
24. Kim M, Park SW, Kim M, D’Agati VD, Lee HT. Isoflurane post-conditioning protects against intestinal ischemia-reperfusion injury and multiorgan dysfunction via transforming growth factor-β1 generation. Ann Surg. 2012;255:492–503
25. Clark JA, Coopersmith CM. Intestinal crosstalk: a new paradigm for understanding the gut as the “motor” of critical illness. Shock. 2007;28:384–93
26. Wille-Jørgensen P, Guenaga KF, Matos D, Castro AA. Pre-operative mechanical bowel cleansing or not? An updated meta-analysis. Colorectal Dis. 2005;7:304–10
27. Holte K, Nielsen KG, Madsen JL, Kehlet H. Physiologic effects of bowel preparation. Dis Colon Rectum. 2004;47:1397–402
28. Kellum JA, Angus DC. Patients are dying of acute renal failure. Crit Care Med. 2002;30:2156–7
29. Hoste EA, De Corte W. Clinical consequences of acute kidney injury. Contrib Nephrol. 2011;174:56–64
30. Zacharias M, Mugawar M, Herbison GP, Walker RJ, Hovhannisyan K, Sivalingam P, Conlon NP. Interventions for protecting renal function in the perioperative period. Cochrane Database Syst Rev. 2013;9:CD003590
31. Faubel S. Acute kidney injury and multiple organ dysfunction syndrome. Minerva Urol Nefrol. 2009;61:171–88
32. Ward RA, Brier ME. Retrospective analyses of large medical databases: what do they tell us? J Am Soc Nephrol. 1999;10:429–32
33. Josephs SA, Thakar CV. Perioperative risk assessment, prevention, and treatment of acute kidney injury. Int Anesthesiol Clin. 2009;47:89–105
34. Nisanevich V, Felsenstein I, Almogy G, Weissman C, Einav S, Matot I. Effect of intraoperative fluid management on outcome after intraabdominal surgery. Anesthesiology. 2005;103:25–32
35. Kambhampati G, Ross EA, Alsabbagh MM, Asmar A, Pakkivenkata U, Ejaz NI, Arif AA, Ejaz AA. Perioperative fluid balance and acute kidney injury. Clin Exp Nephrol. 2012;16:730–8
36. Dexter F, Dexter EU, Ledolter J. Statistical grand rounds: importance of appropriately modeling procedure and duration in logistic regression studies of perioperative morbidity and mortality. Anesth Analg. 2011;113:1197–201
Supplemental Digital Content
© 2014 International Anesthesia Research Society