ABSTRACT
Background: Acute kidney injury (AKI) is a significant complication of COVID-19 infection. UK NICE guidelines have been developed. Aim: to examine our local patientlevel COVID-19 Hospitalisation in England Surveillance System (CHESS) database to elucidate potential risk factors for AKI vs guidelines. Methods: 564 COVID positive admissions between 7 March-24 May 2020 at University Hospital Southampton were examined using Python (Anacondas distribution) and SPSSTM. AKI was staged by RIFLE and AKIN criteria consistent with NICE guidance. X2, t-test, Mann-Whitney U test and logistic regression were used to analyse the data. Results: AKI was present in 177 patients (31%). At peak, 108 (61%) stage 1;42 (24%) stage 2;27 (15%) stage 3. There were no significant differences in cohorts with respect to white vs non-white ethnicity, gender, obesity or anti-COVID-19 treatment. 44% of patients with AKI died vs 19% in the non-AKI group (p<0.001). AKI was associated with ICU admission (27% vs 10% p<0.001), requirement of non-invasive (13% vs 4%) and invasive ventilation (14% vs 4%) (both p<0.001). Prior diabetes (18% vs 8%), hypertension (47% vs 34%), chronic respiratory and cardiac disease (both 25% vs 15%) were more common in the AKI group (p<0.004). Increased age was associated with AKI (p=0.02) and length of stay (LOS) positively correlated to AKI stage(p<0.001). Peak levels of biomarkers: ferritin, D-dimer, C-reactive protein, high sensitivity troponin-I, neutrophil count and total white cell count, were all significantly raised (p<0.001) in the AKI group, increasing with stage of AKI (p<0.001). However, in multivariable analysis first clinical observations, neutrophil count, haemoglobin, D-Dimer and albumin came out as the most significant predictors of AKI: Specificity 88.7%, Sensitivity 43.6%. Conclusions: AKI is a frequent complication of COVID-19 and we identified similar risk factors to those in the NICE guidelines. In addition, we found hypertension and chronic respiratory disease to increase risk of AKI whilst ethnicity, gender, obesity and COVID-19 treatments did not. Furthermore, AKI was associated with increased mortality, ICU admissions and LOS, concordant with previous studies. This data also points to several biomarkers as possible predictors of AKI development and severity. Further analysis of this data is ongoing.