The protocol is established based on the PRISMA protocols (16). This systematic review can be found in the PROSPERO database (CRD42021268701, on http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID = CRD42021268701). Synchronous updates for amendments in the protocol or review process were performed in PROSPERO registration. Enforcement and reporting for the systematic review will observe the Cochrane Handbook for Systematic Reviews of Interventions and Suggestions, and the extended PRISMA statement for reporting systematic reviews containing interventions based on meta-analyses (17, 18).
Criteria for considering studies for this review
Types of studies
Only RCTs (random, non-blinded, placebo) were considered. Due to false-negative (type II) errors, RCTs do not possess sufficient statistical capacity to assess safety outcomes that are atypical or rare (19). Other research types, such as observational research, case series, or case reports, were omitted(20).
Types of participants
Patients diagnosed with CKD and COVID-19, regardless of whether they had CKD before COVID-19 infection or as a result of infection, are included. Patients with CKD or COVID-19 alone are excluded, with no restriction on age, sex, or course of the disease. No clinical-stage restrictions are applied to mild, moderate, or severe/critical cases. Moreover, there are no restrictions on comorbidities.
Types of interventions
Medications with ACEIs and ARBs are given to the intervention group using a random dose. There were no restrictions on standard therapy in the review (such as supportive therapy, antiviral, dialysis treatment).
Types of comparator(s)/control
The control group received the same treatment during the original research. No restrictions were proposed for standard therapy.
Types of outcome measures
Primary outcomes
Primary endpoints
– Time of clinical recovery
– Blood pressure
– Blood creatinine fluctuation,
– Urine protein fluctuation;
– eGFR fluctuation.
Secondary outcomes
Secondary endpoints.
– Discharged date.
– All-cause mortality.
– Frequency of respiratory.
– Oxygen saturation
– Treatment of emergency adverse events
Search methods for identification of studies
E-databases, including PubMed, Embase, CENTRAL, the Chinese Biomedical Literature Database, CNKI, VIP, and Wanfang, are searched between December 1, 2019 to August 31, 2021, in English and Chinese languages. Additionally, Google and Baidu Scholar are employed for seeking potential missing articles. For providing high-quality evidence, RCTs papers were included exclusively.
Reviewers will use the combinations of the following terms for electronic searches: severe acute respiratory syndrome coronavirus 2, novel coronavirus, SARS-CoV-2, COVID-19, COVID19, 2019-nCoV, chronic kidney chronic disease, chronic renal disease, chronic kidney failure, chronic renal failure, chronic kidney injury, randomized controlled trial, randomized, randomly, trials. Table 1 demonstrates the PubMed search strategy, which may be modified by reviewers based on the requirements of other databases(Table 1).
Data collection and analysis
Selection of studies
Screening, trial selection, and data extraction are carried out separately by two reviewers. All searched studies were imported into EndNote X10 to screen titles and abstracts and exclude duplicates and studies inconsistent with inclusion criteria. Finally, the inclusion of studies is decided when reviewers read the whole text of the remaining articles. Making contact with the original corresponding authors is required when access to text is unavailable. The third reviewer is responsible for arbitrating the discrepancies. For summarizing the research selection process, the PRISMA flow chart is displayed (Figure 1).
Data extraction and management
Two reviewers implement the pre-designed data collection process to retrieve information from chosen studies, including reference ID independently; author information; publication year; research type; research design; research setting; sample size; participant characteristics (age, sex, disease duration and severity, laboratory test, CT scan); ACEIs and ARBs intervention and control groups (details about randomization, blinding, allocation, intervention methods, and durations) and primary and secondary outcomes at reporting time points. The third reviewer is responsible for solving the inconsistency. Data are cross-checked before transfer.
Assessment of risk of bias in included studies
The Cochrane Risk of Bias Assessment Tool is applied by two assessors for evaluating publication bias risk in enrolled studies from eight research fields, including generation of randomization sequences, random allocation concealment, participant/personnel/outcome assessor blinding, insufficient outcome data, the bias of selective reporting, as well as other biases. Each field was categorized as high, unclear, or low bias risk. As to unclear division, contact with original corresponding authors is necessarily required. The third scrutator is responsible for solving the inconsistency(21).
Measures of treatment effect
Synthesis and statistical processing of efficiency-related data are accomplished with RevMan V.5.3 independently. As to binary data, a risk/odds ratio with 95% confidence intervals (CIs) was used, while in the case of continuous data, a mean difference/standardized mean difference (SMD) plus 95% CI were adopted. Different assessment tools were applied in conjunction with SMD.
Dealing with missing data
Making a contact with original corresponding authors through mail or telephonically was necessarily required in case of any unclarity or non-availability of data. If the problem remains pending, the studies are excluded. For addressing the underlying influences of missing data, the sensitivity analysis is performed(22, 23).
Assessment of clinical and methodological heterogeneity
For statistical heterogeneity assessment, the χ2 and I2 statistic tests are performed. In the case of low heterogeneity (I2<50%), the fixed-effects model is adopted, whereas the random-effects model is employed for moderate heterogeneity (50%<I2<75%). Meta-analysis is not required in the condition of significantly high heterogeneity (I2>75%).
Assessment of reporting biases
Where 10~ eligible trials are enrolled, the underlying bias of reporting is assessed by funnel plotting. In addition, Egger regression and Begg correlation tests are required to recognize asymmetry in the funnel chart (24).
Data synthesis
Following Cochrane guidelines, in case of low heterogeneity of the pooled data, the fixed-effects model is adopted. Where the heterogeneity was moderate, the random-effects model was employed. Meta-regression or subgroup analysis-based determination of underlying sources is done, when the heterogeneity is high. A p-value of <0.05 represented statistical significance. The results are presented as diagrams when the meta-analysis is impossible (25).
Subgroup analysis and investigation of heterogeneity
Subgroup analyses were conducted where feasible concerning the severity/duration of disease, administration route, as well as source and dosage of ACEIs/ARBs, for elucidating the heterogeneity.
Sensitivity analysis
Sensitivity analysis was performed where feasible for robustness evaluation of the joint effects in enrolled trials, subject to variables, including the sample size, methodological quality, data deficiency, and high bias risk.
Summary of evidence
The present review classified the evidence quality into very low, low, moderate, and high levels in accordance with the GRADE approach by the Cochrane Collaboration Network (26, 27). Evidence quality in specific studies was evaluated by the bias risk, indirectness, inconsistency, publication bias, imprecision, dose-response relation, as well as effect size. The survey results are compiled in the “Summary of surveying table.” Any inconsistency was addressed through negotiation or arbitration by the third examiner.
Patient and public involvement
Since this research was performed through a systematic review of published literature, no patient or the public was involved in the design and implementation of research, data analysis, or result dissemination.
Ethics and dissemination
The present systematic review was based on published data, so ethical approval was exempted. The results are scheduled to be published at international conferences, which are also available in peer-reviewed journals.