RESUMO
INTRODUCTION: Recent developments characterizing the pathophysiological basis of infection in the Coronavirus disease - 19 (COVID-19) have stirred great interest in studying this disease outside the purview of respiratory involvement and especially focusing on the gastrointestinal (GI) system. The present study involving a large cohort of COVID-19-infected patients reports on the characteristics of GI manifestations in patients infected with COVID-19 as well as the predictive role in their association with disease severity and adverse outcomes. METHODS: A retrospective cohort study was carried out in a tertiary care hospital in northern India. Descriptive analysis of GI symptoms was carried out followed by predictive analysis assessing COVID-19 severity and with the primary endpoint of 28-day in-hospital all-cause mortality. RESULTS: Of 3842 hospitalized COVID-19 patients, 2113 (55%) were symptomatic. GI symptoms were present in 163 (7.1%) patients. Common GI symptoms were diarrhea 65 (3.1%), anorexia 61 (2.9%) and vomiting 37 (1.8%). Mild and moderate-to-severe disease was seen in 1725 (81.6%) and 388 (18.4%) patients, respectively. Logistic regression showed greater odds of moderate-to-severe disease in patients with any GI symptoms (odds ratio [OR] 1.849, 95% CI 1.289-2.651 [p = 0.001]) and anorexia in particular (OR 2.797, 95% CI 1.647-4.753 [p = 0.001]); however, on multivariable-analysis, this association lost its significance. A total of 172 patients succumbed to illness. In the Cox proportional hazards model for mortality, patients with any GI symptom (HR 2.184, 95 CI 1.439-3.317 [p < 0.001]) and anorexia (HR 3.556, 95% CI 2.155-5.870 [p < 0.001]) had higher risk. In multi-variable analysis after adjustment to age, sex, oxygen saturation and comorbidities, the presence of any GI symptom was a significant predictor of mortality (hazard ratio adjusted [HRadj] 1.758, 95% CI 1.147-2.694 [p = 0.010]). CONCLUSION: GI symptoms were common among patients infected with COVID-19. The presence of any GI symptom was a significant predictor of the risk of mortality after adjustment to respiratory failure, age, sex and pre-existing comorbidities. The clinical and pathophysiological basis of these associations has been explored.
