Desenvolvimento e validação de escore de risco de evolução para necessidade de terapia renal substitutiva em pacientes hospitalizados por COVID-19

Abstract

Background: Acute kidney injury (AKI) is frequently associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the need for kidney replacement therapy (KRT) is considered an indicator of disease severity. This study aimed to develop and to validate a prognostic score for predicting the need for KRT in hospitalized with coronavirus disease (COVID-19) patients. Methods: This study is part of a multicenter cohort, the National multicenter hospital registry of patients with disease caused by SARS-CoV-2 (COVID-19). A total of 5,212 adult patients hospitalized with COVID-19 laboratory confirmed diagnosis were included, from March/2020 to September/2020. The sample was divided into derivation (N=3,680) and validation (N=1,532) according to the period of hospitalization: March/2020 to July 21/2020 and July 22/2020 to September/2020, respectively. Additionally, a sample of 1,378 patients hospitalized with COVID-19 laboratory diagnosis at the University Hospital of São Paulo, between March/2020 and July/2020, were used for geographic validation. Variable selection was performed using generalized additive models (GAM) and the absolute minimum contraction and selection operator regression (LASSO) method were used for score derivation. Discrimination was evaluated using the area under the receiver operating characteristic (AUROC) curve. Results: The median age of the model derivation cohort was 59 years (CI 47-70), 54.5% were men, 34.3% were admitted to the intensive care unit (ICU), 20.9% evolved with AKI, 9.3% required KRT and 15.1% died during hospitalization. The temporal validation cohort showed similar age, gender, ICU admission, AKI, distribution of KRT and hospital mortality. The geographic validation cohort was similar in age and gender; however, this cohort had higher rates of ICU admission (65.9%), need for KRT (20.2%) and hospital mortality (33.1%). Four predictors of the need for RRT were identified, creating the score's title: M: mechanical ventilation (need for mechanical ventilation); M: male (male gender); C: creatinine (serum creatinine level at hospital presentation); and D: diabetes (presence of diabetes mellitus). The MMCD score had excellent discrimination in derivation cohort (AUROC 0.929, 95% CI 0.918–0.939), temporal (AUROC 0.927, 95% CI 0.911–0.941) and geographic (AUROC 0.819, 95% CI 0.792–0.845), with good overall performance (Brier score: 0.057, 0.056 and 0.122, respectively). The score was implemented in a freely available online risk calculator (https://www.mmcdscore.com/). Conclusions: The MMCD score demonstrated four risk factors for predicting the need for KRT in adult patients hospitalized with COVID-19. Its use can help healthcare professionals to identify hospitalized patients with COVID-19 who may require more intensive monitoring and be useful for resource allocation.