Association of Glucose-Lowering Drugs With Outcomes in Patients With Diabetes Before Hospitalization for COVID-19

Key Points Question What is the difference in the association between COVID-19–related adverse outcomes and 8 routine glucose-lowering therapies in hospitalized patients with diabetes? Findings In this network meta-analysis of 31 observational studies with more than 3.6 million patients, sodium-glucose cotransporter-2 inhibitors were associated with lower risk of COVID-19–related adverse outcomes in diabetes, followed by glucagon-like peptide-1 receptor agonists and metformin, compared with insulin, dipeptidyl peptidase-4 inhibitors, secretagogues, and glucosidase inhibitors. Meaning The findings of this meta-analysis provide information regarding routine antihyperglycemic medications and COVID-19–related adverse outcomes.


Rationale
3 Describe the rationale for the review in the context of what is 3 already known, including mention of why a network metaanalysis has been conducted. Objectives 4 Provide an explicit statement of questions being addressed, 3 with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

Protocol and 5
Indicate whether a review protocol exists and if and where it None registration can be accessed (e.g., Web address); and, if available, provide registration information, including registration number. Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) 3-4 and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. Clearly describe eligible treatments included in the treatment network, and note whether any have been clustered or merged into the same node (with justification). Information sources 7 Describe all information sources (e.g., databases with dates of 3 coverage, contact with study authors to identify additional studies) in the search and date last searched. Search 8 Present full electronic search strategy for at least one database, eTable 1 including any limits used, such that it could be repeated. Study selection 9 State the process for selecting studies (i.e., screening, Figure 1 eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Data collection 10
Describe method of data extraction from reports (e.g., piloted 3-4 and process forms, independently, in duplicate) and any processes for eAppendix2 obtaining and confirming data from investigators. Data items 11 List and define all variables for which data were sought (e.g., 3 PICOS, funding sources) and any assumptions and simplifications made. Geometry of the S1 Describe methods used to explore the geometry of the 4-5 network treatment network under study and potential biases related to it. This should include how the evidence base has been graphically summarized for presentation, and what characteristics were compiled and used to describe the evidence base to readers. Risk of bias within 12 Describe methods used for assessing risk of bias of individual 4 individual studies studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis. Summary measures 13 State the principal summary measures (e.g., risk ratio, 4-5 difference in means). Also describe the use of additional summary measures assessed, such as treatment rankings and surface under the cumulative ranking curve (SUCRA) values, as well as modified approaches used to present summary findings from meta-analyses.

Planned methods of 14
Describe the methods of handling data and combining results of 5 analysis studies for each network meta-analysis. This should include, but not be limited to: • Handling of multi-arm trials; • Selection of variance structure; • Selection of prior distributions in Bayesian analyses; and •Assessment of model fit.

Assessment of S2
Describe the statistical methods used to evaluate the agreement 5 Inconsistency of direct and indirect evidence in the treatment network(s) studied. Describe efforts taken to address its presence when found.

Risk of bias across 15
Specify any assessment of risk of bias that may affect the 7 5 studies cumulative evidence (e.g., publication bias, selective reporting within studies).

Additional analyses 16
Describe methods of additional analyses if done, indicating 5 which were pre-specified. This may include, but not be limited to, the following: • Sensitivity or subgroup analyses; • Meta-regression analyses; • Alternative formulations of the treatment network; and • Use of alternative prior distributions for Bayesian analyses (if applicable).

Study selection 17
Give numbers of studies screened, assessed for eligibility, and 5, Figure 1 included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

Presentation of S3
Provide a network graph of the included studies to enable 5, Figure 2 network structure visualization of the geometry of the treatment network.

Summary of S4
Provide a brief overview of characteristics of the treatment 5 network geometry network. This may include commentary on the abundance of trials and randomized patients for the different interventions and pairwise comparisons in the network, gaps of evidence in the treatment network, and potential biases reflected by the network structure. Study 18 For each study, present characteristics for which data were 5, individual studies each study: 1) simple summary data for each intervention group, and 2) effect estimates and confidence intervals.
Modified approaches may be needed to deal with information from larger networks.

Synthesis of results 21
Present results of each meta-analysis done, including 6, Figure3 and confidence/credible intervals. In larger networks, authors may e Figure 1 focus on comparisons versus a particular comparator (e.g. placebo or standard care), with full findings presented in an appendix. League tables and forest plots may be considered to summarize pairwise comparisons. If additional summary measures were explored (such as treatment rankings), these should also be presented.

Exploration for S5
Describe results from investigations of inconsistency. This may 6, inconsistency include such information as measures of model fit to compare eTable 3 and consistency and inconsistency models, P values from statistical eTable 4 tests, or summary of inconsistency estimates from different parts of the treatment network.

Risk of bias across 22
Present results of any assessment of risk of bias across studies 6, eTable 5 studies for the evidence base being studied.

Results of 23
Give results of additional analyses, if done (e.g., sensitivity or 7, eTable 6 additional analyses subgroup analyses, meta-regression analyses, alternative network geometries studied, alternative choice of prior distributions for Bayesian analyses, and so forth).

Summary of 24
Summarize the main findings, including the strength of 7-8 evidence evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policymakers). Limitations 25 Discuss limitations at study and outcome level (e.g., risk of 8 bias), and at review level (e.g., incomplete retrieval of identified research, reporting bias). Comment on the validity of the assumptions, such as transitivity and consistency. Comment on any concerns regarding network geometry (e.g., avoidance of certain comparisons).

Conclusions 26
Provide a general interpretation of the results in the context of 8 other evidence, and implications for future research.

Funding 27
Describe sources of funding for the systematic review and other 9 support (e.g., supply of data); role of funders for the systematic review. This should also include information regarding whether funding has been received from manufacturers of treatments in the network and/or whether some of the authors are content experts with professional conflicts of interest that could affect use of treatments in the network. PICOS = population, intervention, comparators, outcomes, study design. * Text in italics indicates wording specific to reporting of network meta-analyses that has been added to guidance from the PRISMA statement. † Authors may wish to plan for use of appendices to present all relevant information in full detail for items in this section.

eAppendix 2. Research data collection process
Item of collection process Specific measures 1. Eligibility criteria 1.1 Types of studies All articles reporting on observational studies and randomized controlled trials were included. Reviews, comments, editorials, letters to the editor were excluded. Letters needed to be carefully assessed for available data.

Publication restrictions
Without language limitation (published in English, Chinese, German, etc). Without any regional and publication restrictions.

Types of participants
All diabetes hospitalized for COVID-19, either male or female, and across all ages. Pregnant women and those younger than 14 years were excluded. In the setting of the epidemic, the analysis on diabetes type, diabetes severity and diabetes duration was lacking.

Types of outcome measures
A composite adverse outcome including the need for ICU admission, invasive and non-invasive mechanical ventilation, or in-hospital death. If all three outcomes were described, we chose the data of in-hospital death.

Search strategy
2.1 Databases PubMed, Embase, Cochrane Central Register, Web of Science, and ClinicalTrials. Gov. Reference lists of included articles, review articles and meta-analyses were also screened. Grey literature without peer review was also carefully considered.

Retrieval time
From inception to 5 September 2022 2.3 Search query eTable 1 in the Supplement.

Study quality assessment
3.1 Application method Blinded review by 2 independent reviewers (Z.Z. and Q.Y.Z.) and consultations of disagreement by a third independent reviewer.

Quality evaluation
The Newcastle Ottawa scale was used (range from 0 to 9). Studies with a score ≥7 were considered to be included. 3.3 Selection flowchart Figure 1.

Data acquisition
Data were obtained from the original articles and supplementary materials. If the same population was in different studies, in order to avoid bias, the most complete data was selected. Besides, we obtained raw data by contacting the corresponding author.

Data extraction
5.1 Application method A unified and independent data extraction table was employed. Blinded review by 2 independent reviewers (Z.Z. and Q.Y.Z.) and consultations of disagreement by a third independent reviewer.

Extract content
Study information(year, author, country), participant characteristics at baseline (age, sex, BMI, HbA1c, diabetes duration), total number and death number of people using each glucose-lowering therapy.

Retrieval time From inception to 5 September 2022 5.4 Unified standard
We expressed the data using mean and percentage. When the mean and percentage of baseline characteristics were not provided, the percentage of BMI ≥30kg/m2, HbA1c≥ 7.5%, and diabetes duration≥10 years were calculated.

Included in NMA
Total number and death number of people using each glucose-lowering therapy were included in the network meta-analysis.  / 7(4%) Note: Sensitivity 1: the lowest SUCRA value (SGLT-2i) was removed. Sensitivity 2: the highest SUCRA value (insulin) was removed. eFigure 2. The evaluation of publication bias by funnel plots eFigure 3. Brooks-Gelman-Rubin Diagnosis Plot(potential scale reduction factor, PSRF=1)