SO042A COLLABORATIVE QUALITY IMPROVEMENT PROJECT TO IMPROVE MANAGEMENT OF COMMUNITY ACQUIRED PRIMARY CARE AKI (CAPAKI) USING E ALERTS AND AN EDUCATIONAL OUTREACH PROGRAMME

Abstract

INTRODUCTION AND AIMS: Acute Kidney Injury (AKI) is associated with high financial cost, mortality and length of stay. 60% of all patients have established AKI when admitted to hospital. Primary care AKI (CAP-AKI) is acquired and diagnosed in the community. CAP-AKI is poorly recognised and impacts on long term patient mortality.Aim 1) Reduce the response time for severe community acquired acute kidney injury (stages 2 and 3) detected in primary care (CAP-AKI). METHODS: Baseline incidence of CAP-AKI was collected between January 2015 and April 2016. Salford GP practices were assigned into 4 groups using a restricted randomisation technique. A 2x2 factorial design was constructed using 2 interventions; an educational AKI outreach programme and an AKI e-alert to accompany blood tests taken in the community. All patients in Salford CCG were included. Dialysis patients and patients <18 years old were excluded. Observation period for the effect of interventions was April to August 2016. Response time was defined as number of hours between AKI alert and repeat blood test or admission to hospital. Comparison between pre and during intervention response time was performed with Mann-Whitney U test. The relative effects of the interventions are calculated and visualised based on the Yates algorithm and control charts. RESULTS: 0.8% of all 182,142 primary care creatinine blood tests taken between January 2015 and October 2016 demonstrated CAP-AKI, AKI1(0.6%), AKI2 (0.1%) and AKI3 (0.06%). There were 47 AKI 2 and 3's during the intervention period. Response time during this period decreased compared to the response time of 47 consecutive alerts prior to the intervention phase (median 16 hours, IQR 107 from median 49 IQR 123 respectively, p=0.031). Relative effects of the interventions to the mean response time calculated by the Yates algorithm are shown in Figure 1 (left) and depicted graphically in the response plot in Figure 1 (right). The e-alert reduced the mean response time to repeat blood test or hospital admission by -29.3hours. There was more leverage change when education was combined with the alert (-8.4 hours).The control chart (figure 2) demonstrates a significant shift in mean response time to AKI. This shift was unique to the e-alert and educational outreach group. CONCLUSIONS: This study is the first to demonstrate that the introduction of an ealert reduces the response time to AKI 2 and 3. The outreach programme was welcomed and provided additional benefit when combined with e-alerts.