Evidence Gaps and Guidance for Research
There is insufficient evidence on the impact of COVID-19 on subgroups of patients, such as patients with inflammatory bowel disease, chronic liver disease, or liver transplant recipients on chronic immunosuppression. Early data do not indicate excess risk among patients with inflammatory bowel disease.94, 95, 96, 97, 98 A number of international registries have been established that will provide extremely valuable information about COVID-19 in these potentially vulnerable populations (www.covidibd.org; covidcirrhosis.web.unc.edu; www.gi-covid19.org). Other clinical decisions, including optimal medication management and treatment decisions, are still under investigation. We encourage clinicians to contribute to these registries to further enhance understanding in these subpopulations. Table 4 provides guidance for future studies of GI manifestations in patients with COVID-19 or other similar pathogens.
Table 4 Guidance and Research Considerations for Future Studies of COVID-19a
Study design A prospective inception cohort study is a favorable study design.Another study design that is informative especially when there is a need for rapid data evaluation is a retrospective inception cohort study.
Participants Enrollment of consecutive patients beginning at pandemic onset.Specific set of symptoms that are predictive of COVID-19 infection, all symptoms should be systematically collected on presentation and before COVID-19 diagnosis is established.• Elicit typical upper respiratory infection symptoms (eg, cough, shortness of breath, chest pain, and fever)
• Inquire about less typical symptoms, such as GI-specific symptoms: diarrhea, nausea, vomiting, and abdominal pain, and also other symptoms, such as anosmia, dysguesia
• Describe the GI symptoms in detail, including initial vs late, concurrent vs isolated, duration and frequency, history, and medication initiation relating to the onset of symptoms.
Investigators should avoid:• Undefined sampling (convenience sampling), including undefined time periods.
• Overlap of the same population with other publications, which can be done by coordinating efforts between the different departments within the institution.
Investigators should consider stratification for GI comorbidities, such as inflammatory bowel disease and cirrhosisInvestigators should consider stratification by outpatients vs inpatients
Laboratory Standardized laboratory confirmation should be based on nucleic acid amplification testing for SARS-CoV-2 on respiratory specimen rather than relying on radiologic suspicion on imaging studies, which are less specificLFTs should be obtained on admission and followed throughout the hospitalization.Changes in LFTs should be reported as normal/abnormal and the cutoff for abnormal should be specified, rather than mean and median at the individual patient levelPattern of LFTs abnormalities, hepatocellular vs cholestatic, should be reported as well as the evaluation performed to work up the abnormalitiesBaseline LFTs (prior to developing COVID-19), changes during the duration of the disease, and after resolution should be reported.Report stool RNA testing, when available, and presence of GI symptoms at the time of testing
Disease severity Use of standardized disease severity definitions, for example, as per World Health Organization–China Joint Mission100:• mild-to-moderate: non-pneumonia and mild pneumonia
• severe defined as tachypnea, oxygen saturation ≤93% at rest, or PaO2/FiO2 ratio <300 mm Hg
• critical respiratory failure requiring mechanical ventilation, septic shock, or other organ dysfunction or failure that requires intensive care
Patients can be stratified by:• Disease severity and presence of GI symptoms
• Disease severity and LFTs
Symptoms and their duration before development of a severe stage of the disease should be reported
Outcomes Outcomes should focus on patient-important outcomes, such as death, clinical improvement or disease worsening/progression, hospital discharge; include clinical definitions (eg, threshold reached for intubation); select sufficient follow-up time to ensure outcome is obtainable.
Analysis Analysis should attempt to control for confounding variables; analysis of risk factors should include univariate followed by multivariate analyses to identify independent risk factors predicting more severe disease and poor outcomes
a aIn the table, we specifically refer to COVID-19, but this guidance applies to any future pathogen similar to COVID-19 that presents as a viral illness with potential GI and liver manifestations.
Finally, peer-review remains critical to the process of disseminating information. Journals should add resources to expedite reviews by increasing the number of editors and reviewers to shorten the review process; maintain accuracy, high quality, and details of the data reported; as well as to avoid overlap in patients between studies or multiple studies being published on the same cohort.99