(1326) Impact of Real-Time Clinical Decision Support Tool on Progression of Acute Kidney Injury in Children

Introduction: Acute kidney injury (AKI) affects a significant proportion of hospitalized children and is associated with increased risk of mortality. The 2012 Kidney Disease Improving Global Outcomes (KDIGO) criteria define AKI stage based on degree of rise in serum creatinine and/ or a reduction in urine output, making the diagnosis complicated. Early identification of AKI can lead to earlier supportive therapy, avoidance of nephrotoxic medications, and possibly prevent progression to severe AKI. This study describes the development and implementation of a real-time alert integrated into the electronic health record (EHR) for early identification of AKI in children 1 and 18 years of age admitted to the pediatric intensive care unit (PICU).

Methods: We developed and implemented a clinical decision support tool (CDS) based on KDIGO criteria for AKI utilizing serum creatinine (all patients) and/or urine output criteria (only in patients with indwelling urinary catheter). Patients with advanced chronic kidney disease were excluded. Clinicians were alerted that the patient may have AKI (with specific stage and rationale) and were provided with options to add AKI to the problem list and open a specific order set based on the AKI stage. Clinicians were alerted to potentially nephrotoxic medication the patient was receiving. Chi squared test was used to compare the proportion of patients who developed progression of AKI before and after implementation.

Results: In the 4 months preceding implementation, the alert was triggered 53 times (34 times for stage 1, 13 for stage 2, and 6 for stage 3) in 37 patients (12% of all admissions to the PICU). In the 4 months following implementation, the alert fired 70 times (48 times for stage 1, 14 for stage 2, and 8 for stage 3) in 54 patients (19% of all admissions to the PICU). Twenty seven percent (10/37) patients had progression of AKI before implementation as compared to 15% (8/54) patients after implementation (p = 0.15). The order set was opened 58 times (83 %), and AKI was added to the problem list 15 times (21%).

Conclusions: We demonstrate that a clinical decision support tool may be incorporated into the EHR to facilitate early recognition and management of AKI. Further research is needed to determine if such a tool may help prevent progression to severe AKI.