T. S. Wahl1,2, L. A. Graham2, J. S. Richman1,2, M. S. Morris1,2, R. H. Hollis1,2, K. F. Itani3,4, T. H. Wagner7,8, H. J. Mull3, G. L. Telford5, A. K. Rosen3, L. A. Copeland6, E. A. Burns5, M. T. Hawn7,8 1University Of Alabama Birmingham,Birmingham, AL, USA 2Birmingham VA Medical Center,Birmingham, AL, USA 3VA Boston Healthcare System,Boston, MA, USA 4Harvard Medical School,Boston, MA, USA 5Milwaukee VA Medical Center,Milwaukee, WI, USA 6Central Texas Veterans Healthcare System,Temple, TX, USA 7VA Palo Alto Healthcare System,Palo Alto, CA, USA 8Stanford University,Palo Alto, CA, USA
Introduction: Proteinuria indicates renal dysfunction and is a risk factor for acute kidney injury (AKI) and mortality among medical patients, but less is understood among surgical populations. We hypothesized that pre-operative proteinuria would be associated with post-operative AKI, 30-day unplanned readmission and post-discharge mortality.
Methods: Patients undergoing elective inpatient surgery performed at 118 Veterans Affairs Hospitals from October 2007-September 2014 were examined using Veterans Affairs Surgical Quality Improvement Project (VASQIP) data. The VA Central Data Warehouse domains provided vital sign, laboratory, medication administration, and prior healthcare (emergency or inpatient) utilization data within 6 months. Pre-operative dialysis, septic, cardiac, transplantation, and urologic patients were excluded. Patients having a pre-operative urinalysis with a urine-protein dipstick were examined using closest values within 6-months. Urine-protein dipstick results include negative (0 mg/dL), trace (15-30 mg/dL), 1+ (30-100 mg/dL), 2+ (100-300 mg/dL), or 3+ (>300 mg/dL). Inpatient, pre-discharge AKI was defined as an increase in post-operative serum-creatinine >0.3 mg/dL from the closest pre-operative baseline. Multivariate logistic regression identified predictors of post-operative inpatient AKI, 30-day unplanned readmission, and 30-day post-discharge mortality.
Results: Of 271,149 surgeries, 154,129 met inclusion criteria with the majority orthopedic (37%) followed by general (28.8%), vascular (13.9%), neurosurgery (8.1%), and non-cardiac thoracic (6.6%). 43.8% of the population showed evidence of proteinuria (trace: 20.6%, 1+: 16%, 2+: 5.6%, 3+: 1.6%) with 20.4%, 14.9%, 4.3%, and 0.9%, respectively, having a normal pre-operative estimated glomerular filtration rate (eGFR>60 mL/min/1.73m2). Proteinuria was significantly associated with AKI (negative: 8.6%, trace: 12%, 1+: 14.5%, 2+: 21.2%, 3+: 27.6%, p<0.001), readmission (9.3%, 11.3%, 13.3%, 15.8%, 17.5%, respectively, p<0.001), and post-discharge mortality (0.5%, 0.9%, 1.3%, 1.5%, 1.1%, respectively, p<0.001). After adjustment, increasing proteinuria was associated with post-operative AKI, readmission, and mortality (Table 1).
Conclusion: Proteinuria was associated with post-operative AKI, 30-day unplanned readmission, and 30-day post-discharge mortality independent of pre-operative eGFR. Simple urine assessment for proteinuria may identify patients at higher risk of AKI, readmission, and mortality to guide perioperative management.
