P0619POINT OF CARE CREATININE FOR EARLY DIAGNOSIS OF COMMUNITY ACQUIRED ACUTE KIDNEY INJURY IN NIGERIA. TECHNOLOGY EVALUATION AND DESIGN OF QUALITY IMPROVEMENT PROJECT

Abstract

Abstract Background and Aims Acute Kidney Injury (AKI) in low- and middle-income countries is mostly a community-acquired potentially reversible syndrome and has high morbidity and mortality. Due to limited laboratory infrastructure diagnosis of AKI is often delayed until life threatening complications have developed and dialysis treatment is largely unavailable. Decisions for hospital referral from primary health care centers and triage decisions for hospital admission are not based on laboratory results in Port Harcourt Nigeria. To address the need for early diagnosis and treatment of AKI we established a collaboration between the Renal Unit of the University of Port Harcourt Teaching Hospital, Primary Health Care Board Rivers State and the Renal Department of Salford Royal NHS Foundation Trust, aiming at the evaluation of the use of point of care (POC) Creatinine (Cr) for early detection and management of community acquired AKI. Method The first stage of the project evaluated the accuracy of POC Cr technology. Following informed consent patients underwent concurrent measurement of Cr using the central laboratory (Lab) assay (Jaffe) from a venous sample and a point of care Cr measurement using a capillary sample (fingerstick) with the NOVA Stasensor Xpress Cr analyser. Pearson Correlation and Bland-Altman plots were used to assess correlation and agreement between the two methods. During the second stage, the results were discussed at a focused AKI workshop and guidance for the use of POC Cr was developed. Results During the first phase 96 concurrent POC Cr capillary and venous Lab Cr samples were analysed. Mean age was 49±14 years and 66 subjects were females. POC Cr values were 127±122 umol/l and Lab Cr values were 100 ±85 umol/L, mean positive bias of 27.2±47.94 umol/L. Overall, correlation between POC Cr and Lab Cr was very good, with Pearson correlation r=0.956) Figure 1A. All 4 out of 96 values that were outside the limits of agreement (set at mean ±2 standard deviations) were for Lab Cr values >200 umol/L. A Bland-Altman Plot is presented for paired samples with Lab Cr values <200 umol/L (Figure 1B).