RESULTS
695 patients have been hospitalized in infectious clinics with Covid-19 diagnosis in study time. Of these 695 patients, 244 patients were included in the analysis. Flow chart of study is shown in Figure 1.
Mean age was 59.6 ± 13.7 and 65.2% of patients were male. The median time from diagnosis to hospital admission was 5.5 (IQR, 1-8) days. 130 patients (53.3%) had hypertension and 91 patients (37.3%) had diabetes mellitus. Thirty-seven (15.7%) patients admitted with fever and 164 (67.2%) of patients admitted with oxygen saturation (SaO2)≤ 93% in a resting state.
On admission of the 244 patients, median serum urea and SCr were 34 (IQR, 26.8-48) and 0.86 (0.72-1.05) mg/dL, respectively. eGFR<60 mL/min per 1.73 m2 was reported in 44 (18%) of patients. During hospitalization, median peak SCr was 0.95 (IQR, 0.80-1.25) mg/dL. Mean HbA1c of diabetic patients was 7.91 ± 1.99%. Most of the patients (98.8%) admitted with elevated CRP values. IL-6 was available in 96 patients with a median value of 28.3 (IQR, 7.03-113.6) pg/mL.
Thirty-three patients (13.5%) admitted with AKI. AKI was detected in 63 patients (25.8%) on any time of hospitalization including stage 1 in 41/63 (65.1%) patients, stage 2 in 7/63 (11.1%) patients, and stage 3 in 15/63 (23.8%) patients. Table 1 shows the baseline demographic and clinical characteristics and laboratory values of all patients and comparison of the patients grouped according to the presence of AKI. Compared with patients without AKI, patients who developed AKI were significantly older, had more comorbidities; hypertension and diabetes mellitus and admitted to emergency in a shorter time after Covid-19 diagnosis. Moreover patients with AKI had higher leukocytes, CRP and D-dimer values than patients without AKI. Median value of serum urea, SCr, eGFR and percentage of decreased eGFR on admission were significantly higher in patients who developed AKI than patients who did not.
Urinalysis data of study patients are shown in Table 2. The median pH value was 6 (IQR, 5.5-6) and the median urine-specific gravity was 1019.5 (IQR, 1012-1028). After excluding glycosuric patients, the median urine-specific gravity was 1017 (IQR, 1010-1023) in 156 patients. Glycosuria was found in 88 (36.1%) patients and median blood glucose level at the time of urinalysis was 268 (IQR, 210.5-321.5) (data regarding blood glucose at the time of urinalysis in glucosuric patients were available in 48 patients). Only six patients of glucosuric patients(6/48, 12.5%) had a bloodglucose value under renal threshold defined.By urine dipstick, 189 patients (77.5%) had no heme and 188 patients (77.1%) had no proteinuria. The percentage of patients with proteinuria, hematuria and pyuria were significantly higher in patients with AKI. In contrast, urine pH was significantly lower in patients with AKI than patients without AKI.
Most patients received antiviral therapy (favipiravir, 93.4%; remdesivir, 5.3%), low-molecular-weight heparin (LMWH) (93.4%) and corticosteroid therapy (dexamethasone, 82.4%, pulse methylprednisolone, 33.6%). Patients with AKI received hydroxychloroquine treatment less frequently than those without AKI, however patients with AKI received more antibacterial therapies than patients without AKI. The treatments of the study patients; all patients and patients grouped according to the presence of AKI are shown in Table 3.
According to multivariate logistic regression analyses of risk factors on admission associated with the development of AKI in patients with Covid-19 are shown in Table 4. AKI development was associated with higher WBC and decreased eGFR as well as with proteinuria on admission.
During median 8 (IQR, 5-12) days of follow-up, thirty-five patients (14.3%) were admitted to the ICU and 33 patients (13.5%) died. Patients with AKI had significantly higher ICU admission and in-hospital mortality rates than patients without AKI (39.5% vs. 5.5%,P =0.000; 38.1% vs. 5%, P =0.000). Five patients (7.9% of patients with AKI) required continuous renal replacement therapy (CRRT).
Comparison of the demographic, clinical, and laboratory characteristics on admission between patients who survived and who died are shown in Table 5. Compared to the patients who survived, deceased patients were older and they had significantly higher SCr, CRP, ferritin, D-dimer and lower SaO2, lymphocyte and urine pH levels. Moreover, patients who died had significantly higher percentages of AKI, decreased eGFR, proteinuria andhematuriathan patients who survived. According to the multivariate logistic regression analysis; older age, higher CRP levels and proteinuria were independent predictors of in-hospital mortaliy (Table 6). Kaplan-Meier analysis revealed a significantly higher AKI and in-hospital mortality rate for patients with proteinuria (P =0.013) (Figure 2).