Preliminary Assessment of Acute Kidney Injury in Critically Ill Children Associated with SARS-CoV-2 Infection

Abstract

The virus, severe acute respiratory syndrome coronavirus 2, that causes coronavirus disease 2019 (COVID-19) infected >12 million people around the world as of July 2020 (Johns Hopkins Coronavirus Tracker). Knowledge on pathophysiology, management, multiorgan effects, and postinfection complications is evolving. Several studies revealed variable AKI prevalence among hospitalized adults with COVID-19, as high as 76% in critically ill adults with higher risk of mortality (1). To our knowledge, no specific epidemiologic data on AKI in critically ill children with COVID-19 have been published, leading to gaps in our ability to plan for its implications in this ongoing pandemic. We are conducting a multicenter, point-prevalence AKI study among critically ill children with COVID-19. We provide an epidemiologic snapshot of AKI among this cohort. This cross-sectional analysis captured weekly data points. This interim analysis includes children (1 month to 18 years) from 41 centers (six countries; 32 in the United States) with confirmed severe acute respiratory syndrome coronavirus 2–positive diagnosis (on the basis of center testing protocols) admitted to intensive care units (ICUs) between April 15 and May 20, 2020. The primary outcome was AKI defined by Kidney Disease Improving Global Outcomes creatinine criteria. Admission and peak creatinine on data capture days (or previous day if unavailable) were used. Baseline creatinine was the lowest within 3 months before admission or, if unknown, estimated using previously validated methods (2). Hospital outcomes censored at 14 days were (1) deceased, (2) hospitalized in ICU ≥14 days, or (3) presumed ICU hospitalization <14 days (when a patient is not included in two weekly date capture time points or not included in three date capture points and length of stay dates are <14 days). Final 28-day hospital mortality was determined. Characteristics of participants with AKI versus participants without AKI were compared using distribution-appropriate univariable analyses. For characteristics with P value ≤0.10, univariable relative risks (RRs; 95% confidence intervals [95% CIs]) of AKI were calculated. All centers’ ethical review boards approved the study and waived consent requirement. Deidentified data were electronically entered into REDCap, and analysis was conducted in SAS, version 9.4. Only 26 of 41 (63%) centers (23 in the United States, three in Europe/Russia) had critically ill children who were COVID-19 positive during our catchment days. Of 106 included children, almost half (n=47; 44%; 95% CI, 35% to 54%) developed AKI: 47% (n=22) had stage 1, 23% (n=11) had stage 2, and 30% (n=14) had stage 3. No child received dialysis. Even though 71 children lacked baseline creatinine, AKI estimates were similar between those with and without a baseline creatinine (46%; 95% CI, 29% to 62% and 44%; 95% CI, 32% to 55%, respectively). Table 1 displays demographic and clinical characteristics of children with and without AKI. Most children had at least one comorbidity (n=71; 67%), and almost 6% died (n=6). Vasopressor support and admission diagnosis of shock/hemodynamic instability were the only factors associated with higher risk for AKI (RR, 1.6; 95% CI, 1.1 to 2.5 and RR, 2.3; 95% CI, 1.5 to 3.5, respectively). Table 1. - Epidemiology of AKI development among critically ill children who are coronavirus disease 2019 positive Characteristics Total Coronavirus Disease 2019 Positive AKI No AKI n (row %) 106 47 (44) 59 (56) Date of enrollment Week 1 25 (24) 10 (21) 15 (25) Week 2 16 (15) 9 (19) 7 (12) Week 3 9 (9) 4 (9) 5 (9) Week 4 15 (14) 6 (13) 9 (15) Week 5 21 (20) 12 (26) 9 (15) Week 6 20 (19) 6 (13) 14 (24) Age, yr, median (range) 11.0 (0.1–17.8) 10.5 (0.2–17.7) 11.0 (0.1–17.8) Age categories Infants, 0–1 yr 13 (12) 4 (9) 9 (15) Toddlers, 1–5 yr 17 (16) 9 (19) 8 (14) Children, 5–13 yr 30 (28) 15 (32) 15 (25) Adolescents, ≥13 yr 46 (43) 19 (40) 27 (46) Sex Females 52 (49) 26 (55) 26 (44) Males 54 (51) 21 (45) 33 (56) Race White 37 (35) 17 (36) 20 (34) Black 34 (32) 17 (36) 17 (29) Asian 2 (2) 2 (4) 0 (0) Unknown 33 (31) 11 (23) 22 (37) Ethnicity Non-Hispanic, non-Latino, non-Spanish 62 (59) 31 (66) 31 (53) Hispanic, Latino, Spanish 26 (25) 10 (21) 16 (27) Unknown 18 (17) 6 (13) 12 (20) Locationa United States 100 (94) 43 (92) 57 (97) Western Europe 1 (0.9) 0 (0) 1 (2) Eastern Europe/Russia 5 (5) 4 (9) 1 (2) Baseline eGFR, ml/min per 1.73 m2, median (range)b 112 (24–458) 128 (24–458) 105 (71–203) Baseline serum creatinine, mg/dl, median (range)b 0.5 (0.1–1.7) 0.4 (0.1–1.7) 0.5 (0.1–1.1) Admission serum creatinine, mg/dL, median (range)c 0.6 (0.1–5.6) 0.8 (0.1–5.6) 0.4 (0.1–1.1) Peak serum creatinine, mg/dL, median (range)d 0.6 (0.1–5.6) 0.8 (0.3–5.6) 0.4 (0.1–1.1) Any chronic condition 71 (67) 28 (60) 43 (73) No chronic condition 35 (33) 19 (40) 16 (27) Common chronic conditions Seizures/epilepsy 16 (15) 8 (17) 8 (14) Congenital heart disease (corrected and uncorrected) 11 (10) 6 (13) 5 (8) Asthma 11 (10) 5 (11) 6 (10) Admission reasons, multiple allowed Shock/hemodynamic instability 39 (37) 27 (58) 12 (20) Sepsis/infection 30 (28) 14 (30) 16 (27) Respiratory distress 52 (49) 22 (47) 30 (51) CNS symptoms 10 (9) 3 (6) 7 (12) Other 26 (25) 7 (15) 19 (32) Nephrotoxin medication exposuree 43 (41) 20 (43) 23 (39) Maximum percent fluid overload, median (range)f 3 (−10, 110) 2 (−10, 110) 5 (−5, 100) Respiratory support None 48 (45) 19 (40) 29 (49) Noninvasive 28 (26) 13 (28) 15 (25) Invasive 30 (28) 15 (32) 15 (25) Vasopressor support 31 (29) 19 (40) 12 (20) ECMO 2 (2) 2 (4) 0 (0) Mortalityg 6 (6) 3 (6) 3 (5) Length of hospitalization <14 d 84 (79) 36 (77) 48 (81) Length of hospitalization ≥14 d 12 (11) 8 (17) 4 (7) Data are presented as N (column percentages) except where indicated. CNS, central nervous system; ECMO, extracorporeal membrane oxygenation.aOnly regional location provided to maintain anonymity of centers/patients.bBaseline creatinine was missing for 71 patients, and height was missing for four; so, accurate baseline eGFR is only available for 31 (29.2%) patients. The remainder of baseline creatinine values were estimated by standard estimating equations (Materials and Methods).cMissing n=1.dPeak serum creatinine on the basis of weekly ascertainment. Therefore, true maximum peak creatinine may have been missed, and this may underestimate the true peak creatinine during the child’s hospitalization.eNephrotoxic medications are defined on the basis of the large pediatric quality improvement initiative, Nephrotoxic Injury Negated by Just in time Action (5).fFluid overload = (net fluid in − net fluid out)/intensive care unit admit weight.gThe 28-d hospital mortality (missing n=1). We found a high prevalence of AKI among critically ill children with COVID-19 (44%). This matches what has been seen in adult ICU AKI studies on COVID-19 (1). Compared with previous pediatric ICU-associated AKI research, the AKI prevalence with COVID-19 was much higher (44% versus 26%), although differences exist in inclusion criteria and centers (3). We did not see the degree of AKI severity as seen in adult COVID-19 studies as no child in this analysis received dialysis. It should be noted that our cohort comes primarily from the United States during the earlier part of the COVID-19 pandemic, when testing was limited and children were more isolated. Subsequent waves of the pandemic may appear differently. Despite general findings of less severe disease in children compared with adults, 6% of our study population died. This raises concerns given the rising cases of COVID-19 globally and in the United States specifically, and a recent prediction model suggests that if COVID-19 remains unabated, pediatric ICUs could be overwhelmed (4). Our findings should be confirmed with additional in-depth pediatric studies. Only half of participating centers had patients with COVID-19 to include during the analysis period. It could be argued that children are less likely to contract COVID-19, but newer reports suggest otherwise, with rising pediatric cases. Our 6% mortality rate suggests that pediatric severe disease is certainly possible. Thus, we feel low inclusion rates at some centers may primarily be driven by minimized transmission to children as a downstream effect of good population-level control. It is important to note that our AKI estimate is likely an underestimation as we only used creatinine change to define AKI and not urine output. We observed a high prevalence of AKI in critically ill children with COVID-19. With the rising cases worldwide, particularly in the United States, it is likely that AKI rates in critically ill children will also rise. With the complexity that AKI adds to the medical care of children, clinicians need to be even more diligent with early AKI identification, careful attention to fluid balance, and minimization of nephrotoxic medications, so that these patients do not suffer potentially avoidable consequences. Disclosures D. Askenazi reports consultancy agreements with AKI Foundation, Baxter, Bioporto, Inc., CHF Solutions, and Medtronic and receiving research funding from Baxter Renal Products and CHF Solutions. R.K. Basu reports consultancy agreements with Baxter Healthcare Solutions, BD, Biomerieux, BioPorto Diagnostics, and Potrero and speakers bureau for Baxter and BioPorto Diagnostics. E.C. Bjornstad reports that Bioporto, Inc. provided free supplies for a past research project in Malawi (using their NGALds dipstick). S.L. Goldstein reports consultancy agreements with Baxter Healthcare, Bioporto, Inc., CHF Solutions, Fresenius, Kaneka, Inc., La Jolla Pharmaceuticals, MediBeacon, Medtronic, Otsuka, Reata, and Renibus; ownership interest in MediBeacon; receiving research funding from Baxter Healthcare, Bioporto, Inc., and CHF Solutions; receiving honoraria from Baxter Healthcare and Fresenius; patents and inventions with Vigilanz; serving as a scientific advisor or member of MediBeacon; and speakers bureau for Baxter Healthcare and Fresenius. M. Zappitelli reports consultancy agreements with Bioporto, Inc., CytoPheryx Inc., and Eloxx Pharmaceuticals; receiving honoraria from Bioporto, Inc. and Eloxx Pharmaceuticals; and other interests/relationships with the Canadian Pediatric Nephrologists Association, the Canadian Society of Nephrology, and the Kidney Foundation of Canada. The remaining author has nothing to disclose. Funding Grant support was provided by National Institutes of Health grant 2UL1TR001425-05A1.

Topics & Concepts