PROTOCOL Rapid reviews of evidence for WHO scientific briefs on COVID-19 and selected Noncommunicable Diseases (NCDs)

December 8, 2020

General approach to be taken for suite of documents on noncommunicable diseases and COVID-19

Introduction
In the context of the COVID-19 pandemic, WHO and WHO Member States are requesting information and guidance on key topics related to COVID-19 and the virus which causes the disease (SARS-COV-2 in a novel and swiftly-evolving situation it is also necessary to develop de novo reviews on a variety of topics on an expedited timeline. This piece of work is commissioned to address  specific key questions which need to be answered for WHO to provide high-quality, evidence-informed  information products around COVID-19.

The purpose of a scientific brief is to present a summary of the latest research evidence on a specific technical topic. The target audience for scientific briefs is a technical one, including health care providers, researchers and policy-makers interested in a succinct synthesis of the latest research evidence.

Scientific briefs do not include:

1. de novo normative guidance (most suited to technical guidance (usually “interim”));
2. detailed “how to” or implementation information (more suited to technical guidance (usually “interim”) or tools).

Scientific briefs contain a summary and assessment of the research evidence, presented in a consistent order, format and using technical language.

WHO is commissioning a set of Scientific Briefs to cover noncommunicable diseases (NCDs).

Our team, led by Jamie Hartmann-Boyce has been commissioned to review the evidence on asthma and on diabetes as relates to COVID-19. In order to produce the scientific briefs, two rapid reviews will be conducted, details of which are in this document.

Overview of process for identifying evidence
Stage 1. Searches

Search of databases with subject specific terms will be performed, for published literature or literature accepted for publication but not yet published, in any language. (Specific terms/databases to be determined for each review and specified in advance. Searches will not be restricted by language but if a text is found in another language, for which despite best efforts neither the author team nor WHO can find translation support, it will be excluded and this will be transparently reflected in the review)

Stage 2. Review relevant systematic reviews

Screening and data extraction will be done by one reviewer, rather than in duplicate, due to time constraints.

Screen results to identify systematic reviews addressing one or more of the pre-specified questions of interest.

Extract data from relevant systematic reviews.

Appraise systematic reviews using modified version of AMSTAR-2 checklist (https://www.bmj.com/content/358/bmj.j4008), focussing only on critical domains, namely: protocol registered before commencement; adequacy of literature search; justification for excluding individual studies; risk of bias from individual studies; appropriateness of meta-analytical methods; consideration of risk of bias when interpreting results; assessment of presence and likely impact of publication bias). Appraisal will not be used as a basis for excluding reviews but will be used when considering certainty in the findings from said reviews, and whether seeking additional data from primary studies is needed.

Narratively synthesise data from systematic reviews, including a focus on quality of contributing reviews.

Identify critical gaps/critical weaknesses across systematic review literature; if needed move on to stage 3.

Stage 3. Review relevant primary studies

Some primary studies not included in the relevant systematic review  might make a substantial change to the results and conclusions. Such “key” studies will be reviewed. Their quality will be evaluated (using Newcastle Ottawa Scalre), and data will only be included in the scientific brief if no high risk of bias is identified (defined for our purposes as Newcastle Ottawa Scale score >6/9). Narratively summarise each study.

Review output
The rapid reviews will contain:

• Flow diagram (PRISMA)
• List of included reviews, including key characteristics (search date, number of studies, number of participants, countries, study designs)
• Modified critical appraisal of included systematic reviews
• List of excluded systematic reviews (if any), with reasons for exclusion
• Narrative synthesis of included reviews
• Identification of evidence gaps
• Identification of key primary studies not covered in the systematic reviews, where needed (i.e. reviews are out-of-date[1] and/or contain critical weaknesses which justify inclusion of primary studies)
• Short summary (without synthesis) of primary studies, that might make a substantial difference, where needed, with recommendation for updating an existing outdated review if warranted

[1] In normal times, outdatedness may be considered more than a year but during the COVID-19 scenario, with new studies on the topic published extremely quickly, a few months might make a difference. This can be discussed with the specific WHO team for the topic, but up to 3 months might be guide meanwhile.

Subject specific detail
The evidence reviews and development of the scientific briefs will follow the general approach to evidence appraisal and presentation in the WHO scientific brief format described for the suite of scientific briefs on selected noncommunicable diseases.

Specific PI(E)CO questions for diabetes

1. Are people living with diabetes at increased risk of infection with the virus that causes COVID-19, and/or of suffering complications or death?

Population: people diagnosed with any type of diabetes, with no limitations by age, disease severity, or duration. Excluding people with ‘pre-diabetes’ (e.g. impaired glycaemic control which does not meet clinical threshold for diabetes diagnosis) and gestational diabetes.

Exposure: SARS-CoV-2 infection

Comparator: people without diabetes

Outcome: Rates of confirmed SARS-CoV-2 infection; admission to ICU; death.

2. Are there differences in outcomes of SARS-CoV-2 infection within the population of people with diabetes?

Population: people diagnosed with any type of diabetes, with no limitations by age, disease severity, or duration. Excluding people with ‘pre-diabetes’ (e.g. impaired glycaemic control which does not meet clinical threshold for diabetes diagnosis) and gestational diabetes.

Exposure: SARS-CoV-2 infection

Comparator: people with diabetes,  according to the following comparisons:

• Type 1 vs type 2 diabetes
• Controlled vs uncontrolled glycaemia (by HbA1c, whichever definition of control has been used)
• Previously diagnosed diabetes vs diabetes first diagnosed at COVID-19 diagnosis
• People treated with metformin vs people not treated with metformin
• People treated with DPP4-i vs people not treated with DPP4-i
• People treated with insulin vs people not treated with insulin
• People with CVD/hypertension/ chronic kidney disease vs people without
• Low socioeconomic status vs high socioeconomic status

Outcome: Rates of confirmed SARS-CoV-2 infection; admission to ICU; death.

Specific PI(E)CO questions for asthma
Aim: to review the evidence regarding whether people with asthma (PWA) are at increased risk of infection with the virus that causes COVID-19, and/or of suffering complications or death. In particular, we set out to analyse evidence on the following questions:

• Is asthma associated with increased risk of acquiring COVID-19?
• Is asthma associated with hospitalization with COVID-19?
• Is asthma associated with the severity of COVID-19 outcomes?

Inclusion criteria

• Population: people diagnosed with asthma, with no limitations by age, disease severity, or duration.
• Exposure: SARS-COV-2 infection
• Comparator: people without asthma
• Outcome: Rates of COVID-19: confirmed infection; hospitalisation; admission to ICU; death. Where reported, we also extracted information on the specified outcomes where broken down by asthma medication, age, ethnicity, BMI, and co-morbidities.
• Study types: Systematic reviews (see stage 2), with a very broad definition as articles which systematically searched at least one database; and primary studies (see stage 3), including prospective and retrospective cohort studies and case-control studies, but excluding case reports and case series