Recent studies have demonstrated that the angiotensin-converting enzyme 2 (ACE2), which serves as the entry receptor for the SARS-CoV-2 virus, is not only present in the respiratory system but also in extrapulmonary systems, such as the kidney [1]. This receptor-mediated internalisation of the virus has been suggested as the underlying cause for the aggressive inflammatory response in various organs, leading to multi-organ dysfunction in patients with COVID-19 [2]. Moreover, post-acute sequelae of COVID-19, commonly referred to as “Long COVID,” have also been observed in patients who survive the acute illness, with manifestations in both pulmonary and extrapulmonary organ systems, including the kidneys [3]. These findings underscore the need for continued research into the pathophysiology of COVID-19 and its potential long-term effects on multiple organ systems.

Defining Long COVID is a challenging task as it encompasses a wide range of conditions with varying aetiologies, including viral persistence, immune dysregulation, autoimmunity, gut dysbiosis, and organ injury [4,5]. Several short- and long-term sequelae of COVID-19 infection have been reported in previous review articles [6,7]. However, the evidence generated from previous studies has remained inconclusive due to the large variability in effect estimates from existing studies, which differ in study design, population, and selection of controls [8]. Moreover, conflicting evidence on the risk association of certain diseases, including long-term kidney injury, has also been reported [9-11]. In order to bridge this gap, a systematic literature review was conducted to inform the development of care strategies aimed at improving the health and well-being of people with Long COVID.

The systematic review followed the guidelines laid out by the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement and the study protocol was formulated using the population, intervention, comparison, outcome, and study (PICOS) framework. The review included all individuals who were infected with COVID-19 as the population of interest, and the primary outcome measure was the incidence of major adverse kidney events (MAKE), defined as a composite of eGFR decline, death, and initiation of dialysis. This review did not include an intervention or a comparator group. This systematic review excluded literature not published in English, case reports, animal trials, and studies that solely concentrated on the analysis of diseases among individuals under the age of 18 years.

A comprehensive literature search was conducted in PubMed, Embase, Cochrane Library, and CINAHL, from their inception to Jul 18, 2022. The search strategy and the specific search terms are listed in Supplementary Table 1. Upon completion of our search, the retrieved records were exported to Covidence®. Subsequently, we screened the titles, abstracts, and full texts to identify studies that met the eligibility criteria. At least two reviewers screened the retrieved studies independently, and any discrepancies in the process of selecting the studies were addressed through collaborative dialogue.

The study characteristics, such as the name of the first author, year of publication, country or region, study design, and participant details including patient selection criteria, age at baseline, data source or database, male ratio, and incidence of outcomes were extracted by two reviewers (VL and KC) in parallel. In case of any discrepancies, a third reviewer was consulted to resolve the matter. The methodological quality of each included study was evaluated using the modified Leboeuf-Yde and Lauritsen tool developed by Hoy et al. [12] by the two reviewers. The overall risk of bias can be categorized as ‘low,’ ‘moderate,’ or ‘high,’ based on ten items from two study domains (external and internal validity) by two reviewers. In case of any discrepancies, a third reviewer was consulted to resolve the matter.

A random-effects meta-analysis was used to pool the incidence across multiple studies with the application of the Hartung-Knapp-Sidik-Jonkman [13] method to adjust the CI of the pooled estimate. All statistical analyses were performed using R version 4.0.5 (R Foundation for Statistical Computing, Vienna, Austria).

Fig. 1 depicts the study selection process using a PRISMA flow diagram. The initial database search resulted in 450 studies; after removing duplicates, 389 papers were screened. In the title and abstract phases, 316 papers were excluded, leaving 61 papers for full-text screening. Fifty studies were excluded at the full-text screening stage, and 11 met the inclusion criteria for the meta-analysis.

Figure 1. PRISMA flow diagram for study inclusion.

Table 1 provides a summary of the features of the 11 studies, with two originating from the United States, two from Turkey, and one each from Spain, Brazil, France, Croatia, the United Kingdom, and China. A single study was conducted across Croatia, Bosnia and Herzegovina, and Montenegro. The risk of bias assessment of the included studies is summarized in Table 2.

Table 1 . Summary of the characteristics of studies.

Authors	Year	Country	Study design	Setting	Sample size	Male (%)	Age
Maestre-Muñiz et al. [15]	2021	Spain	Cross-sectional	Centre-based	766	50.7	65.1
Chand et al. [16]	2022	US	Cohort	Centre-based	123	51.6	54.0
Oto et al. [17]	2022	Turkey	Cohort	Centre-based	944	56.4	46.0
Ozturk et al. [18]	2022	Turkey	Cohort	Centre-based	1,223	26.6	60.0
Basic-Jukic et al. [20]	2021	Croatia, Bosnia and Herzegovina, and Montenegro	Cohort	Centre-based	308	64.9	57.0
Nakayama et al. [21]	2022	Brazil	Cohort	Centre-based	565	56.3	61.1
Chawki et al. [22]	2021	France	Cohort	Centre-based	248	66.5	67.0
Basic-Jukic et al. [19]	2021	Croatia	Cohort	Centre-based	104	66.3	56.0
Cohen et al. [23]	2021	US	Cohort	Population-based	133,366	43.6	75.0
Ayoubkhani et al. [2]	2021	UK	Cohort	Population-based	47,780	54.9	65.0
Huang et al. [24]	2021	China	Cohort	Centre-based	1,733	51.8	57.0

Table 2 . Risk of bias of included studies.

Authors	A	B	C	D	E	F	G	H	I	J	Overall
Maestre-Muñiz et al. [15]	–	+	+	–	+	–	+	+	+	+	Moderate
Chand et al. [16]	–	–	–	+	+	–	+	+	+	+	Moderate
Oto et al. [17]	–	–	+	+	+	+	+	+	+	+	Moderate
Ozturk et al. [18]	–	–	+	–	+	+	+	+	+	–	Moderate
Basic-Jukic et al. [20]	–	+	+	+	+	+	+	+	+	+	Low
Nakayama et al. [21]	–	+	+	–	+	–	+	+	+	+	Moderate
Chawki et al. [22]	–	+	+	+	+	–	+	+	+	–	Moderate
Basic-Jukic et al. [19]	–	+	+	+	+	–	+	+	+	+	Moderate
Cohen et al. [23]	–	+	+	+	–	+	+	+	–	+	Moderate
Ayoubkhani et al. [2]	+	+	+	+	–	+	+	+	–	+	Moderate
Huang et al. [24]	–	+	+	–	+	+	+	+	+	–	Moderate

+, No (Low risk=1 point); –, Yes (High risk=0 point); Overall, Low (>8 points), Moderate (6–8 points), High (≤5 points)..

A=Was the study’s target population a close representation of the national population in relation to relevant variables, e.g., age, sex, occupation?.

B=Was the sampling frame a true or close representation of the target population?.

C=Was some form of random selection used to select the sample, OR, was a census undertaken?.

D=Was the likelihood of non-response bias minimal?.

E=Were data collected directly from the subjects (as opposed to a proxy)?.

F=Was an acceptable case definition used in the study?.

G=Was the study instrument that measured the parameter of interest (e.g., prevalence of low back pain) shown to have reliability and validity (if necessary)?.

H=Was the same mode of data collection used for all subjects?.

I=Was the length of the shortest prevalence period for the parameter of interest appropriate?.

J=Were the numerator(s) and denominator(s) for the parameter of interest appropriate?.

The forest plots in Figs. 2 and 3 show the results of the meta-analysis of the incidence of MAKE, death, kidney failure, initiation of RRT, and worsening of renal function tests (RFT). The random-effects meta-analysis revealed that the pooled incidence of MAKE was 6.16 per 100 persons, with a 95% CI of 3.01–12.21. Death had an incidence of 4.28 (95% CI: 2.18–8.26), kidney failure had an incidence of 1.49 (0.27–7.77), initiation of RRT had an incidence of 3.54 (95% CI: 0.00–98.43), and worsening of RFT had an incidence of 21.82 (0.05–99.42). The analyses showed high heterogeneity among the pooled data, with I² values ranging between 88.5% and 99.8%.

Figure 2. Forest plot showing incidence of MAKE.

Figure 3. Forest plot showing incidence of death, kidney failure, initiation of RRT and worsening of RFT.

The possibility of publication bias was evaluated using a Begg’s funnel plot, as shown in Fig. 4. The results were distributed symmetrically, indicating the absence of publication bias. However, eight studies were identified as falling beyond the pseudo 95% CI.

Figure 4. Begg’s funnel plot of the studies investigating the incidence of MAKEs.

Our meta-analysis included 11 studies, and the incidence of MAKE after COVID-19 was estimated at 6.16% (95% CI: 3.01–12.21). Despite the limited information about CKD incidence [14], such an incidence is higher than expected in the general population considering the relatively short follow-up period in these eleven studies [2,15-24]. A recent study conducted in Southern Denmark reported an annual incidence rate of CKD of 0.49% [25], which is comparable to rates reported in US Veterans [26]. Given the high incidence of MAKE following COVID-19 infection, further research is warranted to investigate the effect of long-term COVID-19 infection on renal outcomes.

However, the extreme heterogeneity observed in the data highlights the limited applicability of calculating incidence rates, as these rates are expected to differ across various populations. Moreover, the limited number of studies included in the analysis may compromise the credibility of funnel plots in detecting publication bias. Therefore, further research is required to determine the occurrence of MAKE after COVID-19 infection in different geographical regions to facilitate the monitoring of COVID-19 survivors.

Based on our findings, it is important to prioritize monitoring of kidney disease during the post-acute care of COVID-19 patients. However, it is equally crucial to ensure that resources are strategically and carefully allocated. The clinical importance when compared to other potential long COVID symptoms, such as diabetes [2,27-29], cardiovascular diseases [30], and respiratory failure, remains uncertain. It is imperative to meticulously plan and manage resources to effectively address the post-acute care needs of patients with long COVID.

This study had several limitations. We exclusively investigated the incidence of MAKE after COVID-19 and did not compare the risk with that of appropriate controls. Therefore, the excess risk of MAKE after COVID-19 remains uncertain compared to that in individuals who have never contracted COVID-19. Further research is warranted to determine the likelihood of MAKE following COVID-19. Second, patient-level data for each study included in the analysis were not accessible. Third, this review was limited by the quality of the included studies. Most studies were considered to be of moderate quality. Extreme heterogeneity demonstrates the limited utility of attempting to calculate global incidence rates of diseases, which are likely to vary in their incidence across different populations. We propose that future epidemiological studies should focus on discrete populations and conduct thorough subgroup analyses to further assess the factors that lead to variation. Finally, our funnel plot analysis may have been insufficient to rule out potential publication bias. While the presence of asymmetry in a funnel plot can be used to identify possible publication bias along with other issues, it can be unreliable when the number of studies is small (n=8 in this case).

In conclusion, it is evident that the impact of COVID-19 extends beyond immediate survival, revealing a landscape characterized by unforeseen health challenges. This underscores the necessity for a re-evaluation of post-acute COVID-19 care pathways, emphasizing the critical importance of incorporating specialized kidney care components for individuals in their post-COVID-19 recovery phase.

None.

None.

No potential conflict of interest relevant to this article was reported.

Supplementary materials can be found via https://doi.org/10.59931/rcp.24.0005.