Incidence and risk factors of acute kidney injury in COVID-19 patients with and without acute respiratory distress syndrome (ARDS) during the first wave of COVID-19: a systematic review and Meta-Analysis

Abstract Background Acute kidney injury (AKI) is common among patients with COVID-19. However, AKI incidence may increase when COVID-19 patients develop acute respiratory distress syndrome (ARDS). Thus, this systematic review and meta-analysis aimed to assess the incidence and risk factors of AKI, need for kidney replacement therapy (KRT), and mortality rate among COVID-19 patients with and without ARDS from the first wave of COVID-19. Methods The databases MEDLINE and EMBASE were searched using relevant keywords. Only articles available in English published between December 1, 2019, and November 1, 2020, were included. Studies that included AKI in COVID-19 patients with or without ARDS were included. Meta-analyses were conducted using random-effects models. Results Out of 618 studies identified and screened, 31 studies met the inclusion criteria. A total of 27,500 patients with confirmed COVID-19 were included. The overall incidence of AKI in patients with COVID-19 was 26% (95% CI 19% to 33%). The incidence of AKI was significantly higher among COVID-19 patients with ARDS than COVID-19 patients without ARDS (59% vs. 6%, p < 0.001). Comparing ARDS with non-ARDS COVID-19 cohorts, the need for KRT was also higher in ARDS cohorts (20% vs. 1%). The mortality among COVID-19 patients with AKI was significantly higher (Risk ratio = 4.46; 95% CI 3.31–6; p < 0.00001) than patients without AKI. Conclusion This study shows that ARDS development in COVID-19-patients leads to a higher incidence of AKI and increased mortality rate. Therefore, healthcare providers should be aware of kidney dysfunction, especially among elderly patients with multiple comorbidities. Early kidney function assessment and treatments are vital in COVID-19 patients with ARDS.


Background
In December 2019, a novel coronavirus disease (COVID- 19), was identified in Wuhan, China, which has spread quickly worldwide [1]. On March 11, 2020, the World Health Organization declared COVID-19 as a pandemic [2]. At the time of writing this review, over 32.7 million COVID-19 cases and 991,000 deaths have been reported. This rapid spread of COVID-19 across the world brought many challenges to global health care systems. COVID-19 leads to a wide range of clinical manifestations from asymptomatic, mild, moderate to severe health conditions. Symptomatic patients usually present with fever, dry cough, shortness of breath, myalgia, or fatigue [3]. COVID-19 is of clustering onset and mostly affects the respiratory system, with some patients rapidly progressing to acute respiratory distress syndrome (ARDS); these patients frequently require admission to the intensive care unit (ICU) [4]. Furthermore, it has been demonstrated that severe cases are mostly among the elderly and those with an underlying disease, which may progress to respiratory failure and increased risk of mortality [4].
A growing body of literature has exhibited that there are three potential mechanisms through which the kidney becomes involved in COVID-19 disease: cytokine-induced damage, systemic effects, and organ crosstalk [5]. These mechanisms are interrelated and have important implications for extracorporeal therapy [5]. Several risk factors may contribute to developing acute kidney injury (AKI) in severe cases; for example, most COVID-19 patients with AKI are older. Furthermore, they have comorbidities such as diabetes mellitus or hypertension, which are well-known to associate with kidney vulnerability [6]. ARDS has been identified as one of the prominent conditions that could lead to AKI development; this is also one of the primary reasons for admission in individuals with severe COVID-19 [4,7,8]. Additionally, some factors were associated with AKI development in patients with ARDS, such as impairment of gas exchange and severe hypoxemia [9].
Early reports from China suggested that the prevalence of AKI among patients with COVID-19 was low [6]. Guan et al. studied 1,099 Chinese patients with confirmed COVID-19; 93.6% of those patients were hospitalized, but only 5.3% of those were admitted to the ICU [7]. Across the hospital cohort, 3.4% developed ARDS and only 0.5% developed AKI. Notwithstanding, AKI is mostly common in critically ill patients with COVID-19, affecting 20-40% of patients admitted to ICU [10]. It is deemed a marker of disease severity and an adverse prognostic factor of survival [10]. A multicentre retrospective study from Wuhan investigated 239 critically ill COVID-19 patients and reported that the most common complications among those patients were ARDS, acute cardiac injury, and AKI (68.6%, 43.1%, and 49.8%, respectively) [1].
To date, most studies have revealed the incidence of AKI and ARDS in COVID-19 patients. However, none of these studies has compared AKI incidence in COVID-19 patients with and without ARDS. Therefore, this systematic review aimed to assess the incidence and risk factors of AKI among COVID-19 patients with and without ARDS. We also aimed to assess the necessity of kidney replacement therapy (KRT) in addition to the mortality rate.

Materials and methods
The protocol of this review has been registered prospectively in PROSPERO (CRD42020219460). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement (PRISMA) guideline was used to report this systematic review and meta-analysis [11]. The PRISMA checklist is provided in Supplementary 1.

Study selection
Two authors independently screened the abstracts, and any disagreements were resolved by consensus or by a third reviewer. The studies were initially screened by title and abstract to assess articles for relevance. The following populations were excluded: pediatric, endstage kidney disease, kidney transplant, pregnant, or patients who have cancer. Additionally, case reports, reviews, and case series with less than ten patients were excluded. The full texts for the remaining studies were assessed for the following inclusion criteria: observational cohort studies, cross-sectional studies, and case series, with extractable quantitative data on patient demographics and data on AKI, ARDS, need for KRT, and mortality rate. Retrieved full-texts were independently assessed for eligibility by two authors, and disagreements were resolved through consensus.

Data extraction and quality assessment
The following information was extracted from each included study: authors' names, study sites, study design, sample size, cohort demographics (age and sex), the prevalence of comorbidities, and ICU admission. All clinical information related to AKI incidence, ARDS, need for KRT, and the mortality rate was extracted. The primary outcome was the incidence of AKI in COVID-19 patients with and without ARDS. The secondary outcomes were the need for KRT and the mortality rate of COVID-19 patients with AKI and ARDS as well as AKI without ARDS. All studies were qualitatively evaluated based on the National Institute of Health (NIH) quality assessment tools for observational cohort/cross-sectional studies and case series studies with the quality classified as good, fair, and poor. Two independent reviewers assessed the quality of included studies (FA and MAA). Disagreement was resolved through discussion.

Data synthesis and analysis
Random-effects meta-analyses were performed to obtain pooled incidence, Risk Ratio (RR) estimates, and 95% confidence intervals (CIs) for categorical variables. The Metan and Metaprop command tools were used to analyze the incidence of AKI [12]. After the initial analysis, heterogeneity in incidence and mortality outcomes were investigated. All analyses were performed using Stata 16 and Revman 5 (Cochrane). Funnel plot were used to assess publication bias.

Study characteristics
The included studies covered 27,500 patients from hospitals in China, United States (US), United Kingdom (UK), Morocco, Italy, France, and Belgium. All studies included hospitalized patients, except Guan et al. [7], who included both inpatients and outpatients. Seven studies included only patients admitted to the ICU. All included studies reported the incidence of AKI and mortality rate; twenty-four studies reported the need for KRT; twenty-one studies reported the development of ARDS in COVID-19 patients.
Definitions of AKI and ARDS were consistent among the included studies, and the same definitions were used in this review. AKI was defined based on the Kidney Disease: Improving Global Outcomes criteria (KDIGO) [42] and ARDS was defined based on Berlin definition [43].

Quality assessment and risk of bias
As the studies were case-series and cohort/cross-sectional studies, all included articles were evaluated using the NIH quality assessment tools. Domains 5 of the NIH quality assessment tool of case series and also the domain 5, 8, and 12 of the NIH quality assessment tool of observational cohort/cross-sectional studies were not applicable as they relate to the intervention, power calculation, exposure measure overtime and blinding of outcome assessment and these were not applicable to the included studies (Supplementary 3). Out of the 31 studies, 27 were judged as good quality, 3 as fair [15,19,28] and 1 as poor quality [24]. A funnel plot did not indicate major publication bias ( Figure 2).
Furthermore, Cheng et al. [18] studied 1392 COVID-19 patients retrospectively, aiming to assess AKI incidence. The median age of this cohort was 63 years, and they had fewer comorbidities compared to previous studies: hypertension (36%), diabetes (17%) and CKD (1.5%). Only 7% (99 patients) of this cohort developed AKI, and no patient needed KRT. Later, the same authors conducted a prospective cohort study on 701 COVID-19 patients [33]. The median age was 63 years, and 33.4% of those patients had hypertension, 17% had diabetes, and only 2% had CKD. The incidence of AKI among those patients was 5.1%, and the overall mortality was 16.1%.

AKI incidence rate in ARDS COVID-19 patients
Twenty-two studies included COVID-19 patients with ARDS. While these cohorts included >90% of patients with ARDS, the pooled AKI incidence was 67% (95% CI 57-76) [22,23,28,32,39,40]. Grimaldi  were also risk factors for presentation with AKI. Both Arentz et al. [30] and Zheng et al. [40] carried out similar case series that included 55 participants to assess the clinical characteristics and outcomes of confirmed COVID-19 patients admitted to the ICU. The mean age was between 66 to 70 years. Although the majority (96.36%) of participants in both studies had ARDS, AKI incidence was ranged from 19.1% to 20.6%. The most common comorbidities reported by Arentz et al. [30] were CKD (47.6%) and congestive heart failure (42.9%), and the most frequent comorbidity reported by Zheng et al. [40] was hypertension (64.7%).

Need for KRT
Twenty-five studies reported the need for KRT and only 6% of all patients required KRT (Table 2). After analyzing the studies that reported patients with ARDS, we found that 10% of patients required KRT across nineteen studies. Nevertheless, the need for KRT was only 5% among studies that did not report ARDS incidence [14,21,27,[34][35][36]. A higher incidence of AKI (76.7%) and need for KRT (44.3%) were reported by Ch et al. [32] who only included patients who were admitted to the ICU, of which >90% developed ARDS.

Synthesis of results
A total of 27,500 patients from 31 studies were included in the meta-analysis of AKI incidence amongst all COVID-19 cohorts. By using a random-effects model, the pooled incidence of AKI was 26% (95% CI 19-33; I 2 ¼ 99.38%; p < 0.001; Figure 3).
The AKI mortality was analyzed based on the data of 19,888 patients available from 16 studies. Development of AKI was associated with significantly higher mortality among COVID-19 patients, with RR of 4.46 (95% CI 3.31-6; I 2 ¼ 95%; p < 0.00001; Figure 6) in a randomeffects model.

Discussion
To the authors' knowledge, this is the first systematic review to assess and compare the incidence of AKI between COVID-19 patients with and without ARDS. Based on the literature, the incidence of AKI in ARDS patients prior to the emergence of COVID-19 was approximately 35-50%, and was associated with a high mortality rate which significantly increased by almost two-fold in the ICU [44,45]. For instance, a retrospective study examined a cohort of 634 ARDS patients and revealed an AKI incidence rate of 68.3% after ARDS onset [46]. However, AKI incidence may increase significantly in COVID-19 patients after the onset of ARDS, suggesting lung-kidney crosstalk as the dominant cause of kidney injury [47].
This review shows that most of the included studies reported a higher incidence of AKI in COVID-19 patients who developed ARDS. Across all cohorts with confirmed COVID-19 infection from China, US, UK, Morocco, Italy, France, and Belgium, we identified a wide range of AKI incidence which ranged from 0.5% to 76.7%. However, this may reflect the variety of disease severity thresholds for hospitalization all over the world and potential    differences in clinical practice for monitoring kidney dysfunction.
Our findings show that COVID-19 patients admitted to the ICU had a higher incidence of AKI; however, most (68.6%) of those patients also had ARDS. This finding is consistent with that of Xu et al. [1] who studied 239 critically ill COVID-19 patients admitted to the ICU and found that 49.8% of those patients developed AKI, and 68.6% had ARDS.
Although some studies did not enroll COVID-19 patients with ARDS, a few studies demonstrated a high incidence of AKI that ranged from 5.1% to 56.9% [14,18,21,27,[33][34][35][36]. Based on the literature, patients with a history of CKD are more likely to have comorbidities that have been identified as risk factors for poor outcomes, such as older age, hypertension, diabetes mellitus, cardiovascular disease, and this may be explained by our findings [49]. Additionally, Nadim et al. suggested that the advanced age, CKD, and high prevalence of other comorbidities (e.g., hypertension, diabetes, obesity, and heart failure) are all linked to worse outcomes and are risk factors for the development of AKI in patients with COVID-19 [50].
The metabolic syndrome is a cluster of risk factors (diabetes mellitus, hypertension, obesity) for developing cardiovascular disease and kidney abnormalities [51]. The consequences of the metabolic syndrome on kidney functions in COVID-19 patients are poorly descried Lohia et al. [52] investigated the association between metabolic syndrome and severe disease outcomes among 1871 patients with confirmed COVID-19 diagnosis. Author found that patients with metabolic syndrome had significantly increased in the ICU admission (odds ratio [OR], 1.68; 95% CI, 1.36-2.08; p < .001), and need for mechanical ventilation (OR, 1.90; 95% CI, 1.52-2.37; p < .001), and mortality rate (OR, 1.40; 95% CI, 1.11-1.75; p ¼ .004). Patients with CKD were higher in the metabolic syndrome cohort compared to the nonemetabolic syndrome cohort (14.7% vs. 9%; p < .001). Metabolic syndrome cohort may be at higher risk of developing AKI if they contract COVID-19, and this risk may be by background CKD. Chen et al. [53] recruited >6000 participants to examine the association between the metabolic syndrome and kidney impairment, and they found that the multivariate-adjusted risk for both CKD (defined as a GFR of <60 mL/min per 1.73 m2) and microalbuminuria was significantly higher in those with the metabolic syndrome compared to participants without metabolic syndrome (Odds ratio 2.60, 95% CI 1.68-4.03 and 1.89, CI 1.34-2.67, respectively). Also, they revealed that risk increased progressively with the number of the metabolic syndrome's components observed in each patient.
A meta-analysis comparing ARDS with non-ARDS COVID-19 cohorts showed the need for KRT was also higher in ARDS cohorts (20% vs 1%). Across all studies which reported the need for KRT, we observed a higher need for KRT among patients with a high incidence of AKI as reported by Ch et al. [32] and Fominskiy et al. [22]. However, it is unclear from these studies whether KRT was used to treat fluid overload, end-stage kidney disease, or patients with AKI alone.
The mortality rate was significantly increased in critically ill patients if they had a high incidence of AKI or ARDS [23,32]. In our meta-analysis, the overall mortality among patients with AKI was significantly higher than those without AKI (p < 0.00001). In both Ch et al. [32] and Arentz et al. [30], AKI and ARDS development was associated with a high mortality rate. Ch et al. [32] found that the AKI incidence was associated with a 25% increased risk of death (RR: 1.25; 95% CI: (1.02, 1.52)). These findings are consistent with previously published research, which found that severe AKI in patients with COVID-19 is associated with a high mortality rate [54]. Additionally, in the population of Ch et al. [32] high rates of comorbidities such as obesity (85.7% with BMI >25) and hypertension (66.7%) were also linked with increased mortality rate. Both Gasparini et al. [14] and Fominskiy et al. [22] also found that patients with AKI had a higher mortality rate compared to patients without AKI. A possible explanation for the higher mortality rate in the previous two studies is that each study recruited >90% critically ill individuals who were invasively ventilated in ICU.

Limitations
The main limitation of this study is that most of the selected studies were of the retrospective design, making them susceptible to selection bias. There was limited data on AKI patients' characteristics among included studies, such as urine output status, the need for KRT and the grade of AKI. This information is essential to define the severity and recoverability of AKI.
Moreover, most cohorts were from China and the United States, and a small amount from Africa and Europe, which does reduce the generalization to other countries. The heterogeneity between studies was high; consequently, most of the analyses were interpreted based on a random-effects model. Furthermore, given the rapid, constant expansion of COVID-19 research, several studies had limitations in their description of details and relatively short periods of follow-up. Finally, we cannot be certain that all relevant studies were captured, despite a detailed, comprehensive search strategy.

Conclusion
A higher incidence of AKI was most common among COVID-19 cohorts admitted to the ICU and was associated with the development of ARDS. Also, increased age, male sex, and preexisting comorbidities (e.g., hypertension, diabetes, obesity, and CKD) were associated with significantly worse outcomes, which may be considered risk factors for developing AKI and increased mortality among COVID-19 patients. Therefore, healthcare providers should be aware of kidney dysfunction among those patients, especially those who are elderly with multiple comorbidities. This may ultimately lead to using appropriate diagnostic tools and treatment strategies to avoid additional damage to the kidneys and providing adequate kidney support when needed.

Author contributions
FA and JP conceived, planned and analyzed the data, and made a major contribution in writing the manuscript. FA and MAA performed abstract screening, full-text screening and quality assessment. FA and RM planned and performed searching strategy and data synthesis. MAA and RM contributed to data analysis and writing the manuscript. MNB and DRT reviewed the manuscript. All authors have read and approved the manuscript.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This systematic review was part of a funded Ph.D. by King Saud Bin Abdul-Aziz University for Health Sciences, Riyadh, Saudi Arabia under the egis of the Government of Saudi Arabia.

Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.