This study included more than 5000 patients hospitalised from a robust cohort of COVID-19 patients from 35 Brazilian hospitals, with external validation in an independent cohort with over 1000 patients. One in every five patients evolved with AKI and 9.3% required KRT. Among the analysed predictors, four variables were related to progression to AKI and KRT requirement, including the need for mechanical ventilation, sex, creatinine upon hospital presentation and diabetes mellitus. The MMCD score had excellent discrimination in derivation and temporal validation cohorts, with AUROC higher than 0.9, a good overall performance.
Renal involvement in COVID-19 infection is complex and probably occurs due to several factors, including direct injury to the renal endothelium, tubular epithelium and podocytes ; cytokine storm, with the release of several interleukins and cytokines ; cardiorenal syndrome, caused by right ventricular dysfunction secondary to pulmonary infection; hypercoagulable statea; and release of nephrotoxic substances such as creatine phosphokinase secondary to rhabdomyolysis .
The need for mechanical ventilation at any time during hospitalisation was an important predictor of progression to AKI and the need for KRT, being the variable with the highest points in the risk score. Scoring mechanical ventilation by itself changed patients’ category to “high risk” for evolving to AKI and KRT requirement. This finding confirms findings from studies carried out in other countries to assess the risk of progression of AKI to KRT in COVID-19 patients, such as USA (OR 10.7 [95% CI 6.81–16.70])  and UK (HR 4.1 [95% CI 1.61–10.49]) . There is a close relationship between alveolar and tubular damage (lung-kidney axis) in acute respiratory distress syndrome (ARDS), often progressing to different degrees of AKI . This is a complex and not fully understood mechanism, probably multifactorial, in which inflammatory mediators are released by ventilated lungs into the systemic circulation . The relationship between mechanical ventilation (MV) and AKI has been widely recognised before the COVID-19 pandemic. Husain-Syed et al. had demonstrated the occurrence of physiological changes triggered by increased intrathoracic pressure secondary to invasive mechanical ventilation that are harmful to the renal function. These changes can cause reduced renal blood flow, glomerular filtration rate and sodium excretion, with a consequent predisposition to progression to AKI and need for KRT . It is difficult to define the specific role that each mechanism plays in the pathogenesis. They are usually observed simultaneously in critically ill patients, which limits the possibility to develop preventive strategies .
In studies published by Chan L et al. (n=3993) and Fisher M et al. (n=3345) with hospitalised patients with COVID-19 in the USA, male sex was considered an independent predictor of progression to AKI and KRT requirement [31, 32], what is in line with our findings. Male sex has been previously observed to be associated with other adverse outcomes in COVID-19 patients.
Creatinine levels upon hospital presentation may be evidence of previous chronic kidney disease or an early manifestation of AKI caused by COVID-19 infection. Chronic kidney disease is a global health problem and a silent disease . Several risk classifications included serum creatinine levels in mortality scores in patients admitted to an intensive care unit (APACHE II, SAPS 3, Sequential Organ Failure Assessment Score [SOFA]), demonstrating the importance of creatinine levels as a marker of severity [15, 34, 35]. In the present analysis, creatinine levels were categorised according to the SOFA score  to comply with TRIPOD guidelines, which advises not to use a data-driven method, to avoid model overfitting . Our finding is consistent with a recent systematic review and meta-analysis with 22 studies (n=17,391), which observed an increased incidence of AKI in COVID-19 patients hospitalised in the USA who had abnormal baseline serum creatinine levels due to pre-existing chronic kidney disease . Hansrivijit P et al. in their meta-analysis described abnormal basal serum creatinine levels as predictors of progression to AKI . A meta-analysis with 10,335 patients showed that severe cases of COVID-19 had higher serum levels of creatinine and BUN. In severe cases, the risk of progression to need for KRT was 12.99-fold higher compared to non-severe cases, and among patients who died, there was a higher prevalence of AKI, high levels of creatinine and need for KRT .
The association between diabetes mellitus and renal dysfunction is well known, in the form of diabetic nephropathy and non inflammatory glomerular damage [39, 40]. In the present analysis, diabetes proved to be a predictor of risk of progression to AKI and KRT requirement in patients hospitalised with COVID-19, which was in line with a recent meta-analysis (26 studies, n=5497) .
In Brazil, a country severely hit by the pandemic, there is lack of evidence on the association among AKI, need for KRT, mortality and COVID-19. The scarce existing studies are based in small databases. A study published with 200 ICU patients showed a high incidence of AKI (about 50%) and 17% of patients requiring KRT, with significantly higher mortality in patients with AKI and needing KRT, in contrast to patients without AKI and KRT requirement . In our study, the incidence of AKI and need for KRT in ICU patients were lower (about 16 and 9%, respectively), although with higher in-hospital death in this group, similarly to finds in this article. As shown in Table 3, there was a progressive increase in the mortality rate associated with the increase in the score. Patients classified as non-high risk had a mortality of 1.6% in the derivation cohort and 1.9% in the validation cohort, while patients classified as very high risk had a mortality of 80.0% in the derivation cohort and 76.7% in the temporal validation cohort.
The MMCD model retrieved an AUROC of 0.96, which was classified as an excellent discrimination. An American study (n=2256) developed prediction models for mechanical ventilation, KRT and readmission in COVID-19 patients using machine learning techniques. Logistic L1 had the best accuracy, although the discrimination results were inferior than the one observed in the present analysis (0.847 [95% CI, 0.772-0.936]). Additionally, the model uses too many risk predictor variables, hindering its applicability in clinical practice .
External validation was performed with a cohort of patients referred to a tertiary hospital, most of which were critically ill, with a high rate of ICU admission, use of mechanical ventilation and need for KRT and mortality. As the accuracy of a prediction model is always high, whether the model is validated on the development cohort used to derive the model only, the assessment of accuracy in those studies may be overoptimistic .
The criteria for orotracheal intubation evolved over time. Still, we believe it has not affected our findings. The first wave of COVID-19 pandemic in Brazil was in June 2020, late in relation to Europe, which was affected in March 2020. Therefore, when the country faced its first wave, the knowledge about intubation criteria and outcomes had already evolved. The fact that the score’s high accuracy was not reduced in the temporal validation cohort (cut-off on July 21, 2020) is evidence of no significant influence on the results obtained in the temporal validation sample (AUROC 0.927 CI 95% 0.911–0.941).
Strengths and limitations
Our study used a large patients database to develop a risk score to predict the need for KRT in patients admitted with COVID-19. A major strength of the MMCD score is its simplicity; the use of objective parameters, which may reduce the variability; and easy availability, even in under-resourced settings. Then, the MMCD score may help clinicians to make a prompt and reasonable decision to optimise the management of COVID-19 patients with AKI and potentially reduce mortality. Additionally, its development and validation strictly followed the TRIPOD recommendations .
This study has limitations. Indication and timing of initiation of the KRT may differ according to institutional protocols; however, there is a consensus on the criteria on which KRT should be initiated . We did not collect information on patients who did not perform dialysis due to limited resources. Still, this has not affected the accuracy of the score. Additionally, as any other score, MMCD may not be directly generalised to populations from other countries without further validation.
With regard to AKI assessment, it was not possible to use the criterion based on diuresis due to unavailability of this data, as well as the baseline creatinine value to identify AKI due to the lack of data on previous serum creatinine of patients admitted to participating hospitals. Instead, we used the increase of >0.3 mg/dl in creatinine values over 48 h or 1.5-fold increase within 7 days during the hospitalisation, when compared to creatinine at hospital presentation. Therefore, the real incidence of AKI may be underestimated.
Finally, external validation of the MMCD score in other countries should be performed with more recent data on COVID-19 infection, considering the multiple temporal aspects of the pandemic and changes in disease management.
Using predictors available at baseline and within the first hours of the admission, we could objectively predict the probability of KRT of a COVID-19 patient with AKI. With an accurate prediction, it may help to organise resource allocation to patients who are at the highest risk of KRT requirement , in addition to selecting patients who may benefit from renal protection strategies, close assessment and follow-up by a nephrologist .