This is a multicentric retrospective cohort study from 3 tertiary centers (King Abdulaziz University Hospital, King Khalid University Hospital, and East Jeddah General Hospital) in the Kingdom of Saudi Arabia (KSA).
Inclusion criteria
All children who were positive for COVID-19, determined by polymerase chain reaction (PCR) of nasal swab samples, requiring hospital admission to one of the participating centers between March 1st and mid-July 2020 were included. Children were defined as ≤14 years old. All included patients were considered ill because of the required hospital admission. All participating centers used the same visual triage checklist for acute respiratory infection in suspected cases to determine who needed screened using a nasopharyngeal swab.
All patients labeled as positive for COVID-19 had their nasopharyngeal and\or endotracheal swab samples screened by PCR. Patients were considered infected if the initial swab result was positive, or if it was initially negative with a positive repeat test. Only the first admission was considered for patients who had multiple hospital admissions.
Exclusion Criteria
We excluded positive COVID-19 patients with insufficient data, those who were seen in the emergency department with positive swabs but did not require admission, or any patient with a previous admission for COVID-19. We also excluded neonates (defined as those aged <28 days at presentation) and patients with evidence of preexisting chronic kidney disease stages III–V.
Data Collection
Clinical and laboratory data were collected and merged from electronic health records of all participating centers. Data included patient demographic information (age, gender, and nationality), anthropometric measures and centile, presenting symptoms and signs, associated comorbidities, and need for pediatric intensive care unit (PICU). For clinical information, we examined the severity of respiratory distress, blood pressure interpretation, oxygen saturation, need for oxygen therapy, and mechanical ventilation.
Renal function assessment was performed using the following data: baseline creatinine, baseline estimated glomerular filtration rate (eGFR), highest creatinine, and lowest eGFR recorded during the initial 7 days of admission. We calculated urine output (UOP) in children with available UOP data. We identified patients with AKI using the Kidney Disease Improving Global Outcomes (KDIGO) consensus definition [14]. KDIGO defines AKI as an increase in serum creatinine by ≥0.3 mg/dL (≥26.5 µmol/L) within 48 hours; OR an increase in serum creatinine ≥1.5 times the baseline level within the prior 7 days; OR a urine volume ≤0.5 mL/kg/hour for the first 48 hours of admission (for patients who had UOP information). We categorized AKI patients into 3 main stages. Stage I consisted of patients with an increase in serum creatinine by ≥0.3 mg/dL (≥26.5 µmol/L), or an increase in serum creatinine to ≥1.5 times the baseline level within the prior 7 days, and/or a UOP of ≤0.5 mL/kg/hour over 6-12 hours. Stage II patients had an increase in serum creatinine 2 to 2.9 times from the baseline level and/or a decrease in UOP of ≤0.5 mL/kg/hour for 12 to 24 hours. Stage III patients had an increase in serum creatinine by ≥353.6 mmol\L or greater than 3 times the baseline level, a decrease in GFR to <35 ml\min\1.73 m2, and/or a decrease in UOP of ≤0.5 mL/kg/hour for more than 24 hours.
We also calculated the renal angina index (RAI=1-40) for all participants using the multiplication of both risk level and injury level. A score ≥8 is considered a positive RAI [15]. Estimated creatinine clearance was calculated using the modified Schwartz formula [16]. Baseline creatinine was defined as the last creatinine within the previous 6 months prior to the PICU. For those patients admitted for the first time, we used an average GFR according to age, sex, and height of the child [17].
For any patient that required ICU admission, we calculated the Pediatric Risk of Mortality (PRISM version IV) during the first 24 h of ICU admission [18].
All laboratory workups were recorded and followed for identification of significantly associated hematological and\or laboratory abnormalities, such as significant anemia, leucopenia or leukocytosis, neutropenia or neutrophilia, thrombocytopenia, and disturbed coagulation profiles. Other associated laboratory abnormalities, such as hypernatremia, hyperkalemia, and high troponin levels were reported. Evidence of any associated infection (blood and urine culture) and the high acute phase reactants (C-reactive protein and ferritin) were recorded and highlighted.
Clinical outcome indicators included mortality rate, abnormally high creatinine, high blood pressure, or proteinuria.
Definitions
Significant exposure to nephrotoxic medications (nephrotoxic injury negated by just-in-time action = NINJA score): exposure to 3 or more common nephrotoxic medications for 1 day or to aminoglycosides for 3 or more days [19].
Hypertension: 2 or more systolic and diastolic blood pressure readings at 2 different time points >95th percentile for gender, age, and height.
Hypotension: 2 or more systolic and diastolic blood pressure readings at 2 different time points <50th percentile for gender, age, and height.
Hyperkalemia: serum potassium level >5.5 mmol\L
Hypernatremia: serum sodium level >146 mmol\L.
Anemia: a reduction of the hemoglobin concentration to 2 standard deviations below the mean, based on age-specific normal levels.
Leukocytosis and leucopenia: total leucocytic count higher than the mean normal value based on age.
Neutropenia: a neutrophilic count < 500/mmз.
Oliguria: a UOP <0.5 ml\kg\hr.
We used the World Health Organization definition for MIS-C and adolescents aged 0-19 years. Their criteria were fever for 3 days or more and two of the following: skin rash or signs of mucocutaneous inflammation hypotension or shock, features of myocardial dysfunction, evidence of coagulopathy, acute gastrointestinal problems, and elevated markers of inflammation, such as erythrocyte sedimentation rate, C-reactive protein, or procalcitonin, with evidence of COVID-19 and no other obvious microbial cause of inflammation [20].
The study was approved by the Biomedical Ethics Research Committees of the three recruiting centers. Consent from participants was not required as this was a retrospective study using data collected for routine clinical practice.
Statistical analysis
All analyses were performed using STATA (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX). The proportion and mean for dichotomous and continuous variables, respectively, were measured to describe a patient’s characteristics. The association between AKI development and different outcomes (prolonged hospitalization, ICU admission, and mortality) and predictors of residual renal impairment at discharge were measured using the chi-square test. Statistical significance was determined using a 95% confidence interval and a p-value of 0.05.