Lead/Presenter: Brian Lund, COIN - Iowa City
All Authors: Sandahl T (Iowa City Veterans Affairs Health Care System)
Mosher HJ (Center for Comprehensive Access and Delivery Research and Evaluation (CADRE) at the Iowa City VA HCS)
Jacobson M (Iowa City Veterans Affairs HCS)
Egge J (Iowa City Veterans Affairs HCS)
Lund BC (CADRE)
Objectives:
Acute kidney injury (AKI) occurs in up to 20% of hospitalized patients and is associated with substantial morbidity, mortality, and cost. Recent population-based studies have suggested that AKI rates in hospitalized patients are increasing. AKI may be related to underlying acute illness, or may be a preventable consequence of hospital-based treatments. Efforts to prevent AKI in the hospital are contingent on a better understanding of its epidemiology and course. The objective was to characterize the incidence and course of AKI in the medically-hospitalized veteran population.
Methods:
Veterans with baseline normal renal function admitted for acute medical hospitalization between fiscal years (FY) 2001 and 2016 were identified from national VHA administrative data. Normal renal function was defined as serum creatinine (SCr) 0.6-1.49mg/dl. AKI was defined based on KDIGO guidelines as increase in SCr to >= 1.5 times baseline or >= 0.3mg/dl within 48 hours. Descriptive statistics were used to trend the incidence and characterize the timing and longitudinal course of AKI during hospitalization.
Results:
Of 2,839,316 patients meeting inclusion criteria, 181,000 (6.4%) developed AKI during hospitalization. Incidence declined from 7.5% of admissions in FY2001 to 5.6% in FY2016. Daily incidence during hospitalization was highest within two days of hospitalization (2.2%) and steadily declined to =< 1.0% by Day 10. Peak SCr was reached the same day patients met criteria for AKI in 71.3% of patients, and 82.5% resolved within two days of onset. For patients without resolution prior to discharge 59.9% had follow-up laboratory study with measured SCr within 30 days, at which time 65.4% were resolved.
Implications:
In contrast to trends observed in other health-care systems, the incidence of AKI in veterans admitted with normal SCr to VA hospitals for medical care declined over the past 15 years. The highest rates of AKI occurred on the second and third hospital days, and generally resolved quickly.
Impacts:
AKI represents a potentially preventable complication of hospital care; our findings of decreasing rates of AKI are reassuring. Future analyses of correlates of new-onset AKI early and late during hospitalization may uncover targets for improving care and preventing harm in patients at risk for AKI.