Increased Markers of Disease Severity in COVID-19 Patients with Hospital-Acquired vs. Community-Acquired AKI

Background: The etiology of AKI in COVID-19 correlates strongly with age, comorbidities, and laboratory markers of disease severity. Outpatients with COVID-19 have different exposures that may cause AKI than hospitalized patients;thus, the etiology of AKI occurring before hospitalization [community-acquired AKI (CA)] may differ from those with hospital-acquired AKI (HA). Methods: Excluding ESKD and hospital transfers, all COVID-19 PCR-confirmed cases admitted to 4 hospitals from 3/01/20 to 5/31/20 had data collected electronically through 7/31/20 including readmissions. Baseline C-EPI eGFR was determined by chart review for the period of 6 months prior to 5 months post-admission. AKI and recovery from AKI were scored using KDIGO staging. CA was defined as AKI with the highest level of creatinine (Cr) on admission, rising Cr on admission, or RRT started within 48 hours of admission without a subsequent AKI event after recovery. All AKI occurring > 48 hours was considered HA. To test which laboratory values correlated with CA or HA, we used a model adjusted for demographics, BMI, Elixhauser comorbidity index (ECI), and CKD stage. Results: The table shows patients with HA and CA had similar demographics with only the ECI differing significantly. CA had less severe AKI, improved recovery to baseline, and lower mortality than HA. The lower mortality in CA was directly related to the lower stage of AKI. Within a given stage of AKI, mortality was not different between CA and HA. Recovery of renal function was significantly better in CA stage 1 vs. HA (8% vs. 26%, p = 0.001) but was not different for stage 2 or 3. In an adjusted model, higher maximum dimers, ALT, AST, Bili, BNP, lactic acid, CRP, ferritin, LDH, neutrophils, troponin and lower minimum lymphocyte count were significantly associated with HA compared with CA. In contrast, on admission, only higher BNP, higher CRP, lower CPK and higher total CO2 were associated with HA versus CA. Conclusions: Compared to patients with CA, patients with HA had higher stages of AKI that correlated with higher mortality. They also had worsened recovery from stage 1 AKI and increased markers of COVID severity (except for CPK) in-hospital and on admission. We propose that factors other than COVID-19 disease severity led to CA, with volume and rhabdomyolysis as possible contributors.