Gastrointestinal sequelae three and six months after hospitalization for Coronavirus Disease 2019.

Gastrointestinal (GI) symptoms are highly prevalent in coronavirus disease 2019 (COVID-19) ranging from 17.6 % to 53 %.1, 2, 3, 4 The proposed mechanism for GI symptoms involves SARS-CoV-2 virus binding to the host cell's angiotensin-converting enzyme-2 receptor, commonly found in GI tract epithelial cells.

With an increasing population of patients recovering from acute infection, there is now interest in understanding post-COVID-19 sequelae. Our study aims to report GI sequelae 3 and 6 months after hospitalization for COVID-19 infection.

Methods

This is a multicenter, retrospective study of hospitalized adult patients (18 years and older) who tested positive for COVID-19 infection from March 1, 2020 to January 24, 2021 across 12 hospitals comprising the Northwell Health System in New York. Patients were included if they (1) tested positive for SARS-CoV-2 nasal swab polymerase chain reaction, (2) were hospitalized with GI manifestations on initial presentation, and (3) had 3 and/or 6 months follow-up as an outpatient posthospitalization or repeat hospitalization. Further description of methods is included in the Supplementary Methods.

Results

During the study period, 17,462 COVID-19 patients were hospitalized. Of the 17,462 patients, 3229 (18.5 %) also had GI manifestations. A total of 715 (22.1 %) patients had 788 (24.4 %) distinct GI manifestations and also had 3- and/or 6-months outpatient follow-up data post-discharge. Patient characteristics are shown in Table 1 .

Table 1

Results Summary of Patients With COVID-19 Infection With GI Symptoms at 3- and/or 6-Month Follow-Up

Patient demographics	Median (IQR) or n (%)
Patients (+) COVID-19, n	17,462
Patients with GI symptoms, n (%)	3229 (18.5)
Patients with GI symptoms + follow-up, n (%)	715 (22.1)
Distinct GI symptoms in patients	788 (24.4)
Patients with 3-mo follow-up, n (%)	627 (88.7)
Patients with 6-mo follow-up, n (%)	404 (56.5)
Median age, y (IQR)	66 (IQR, 55 to 76)
Female, n (%)	336 (46.9)
Race/ethnicity, n (%)
White	325 (46)
African American	161 (23)
Hispanic	121 (17)
Asian	50 (7)
Native American, Alaskan, Hawaiian	5 (0.5)
Length of stay, d (IQR)	9 (5 to 17)
Mechanical ventilation	101 (14.1)
BIPAP	36 (5.0)
Vasopressors	106 (14.8)
Irritable bowel syndrome	7 (1.0)
Inflammatory bowel disease	14 (2.0)
GI symptoms
Gastroenteritis, n (%)	414 (52.5)
Resolution at 3 mo	323 (90.5)
Resolution at 6 mo	210 (89.5)
GI bleeding, n (%)	161 (20.4)
Resolution at 3 mo	138 (92.0)
Resolution at 6 mo	89 (94.7)
Malnutrition, n (%)	181 (23.0)
Present at 3 mo	81 (53.6)
Present at 6 mo	33 (32.4)
Pancreatitis, n (%)	4 (0.5)
Resolution at 3 mo	4 (100)
Resolution at 6 mo	4 (100)
Endoscopy for GI bleeding
Underwent endoscopy	19 (11.8)
Upper endoscopy	15 (9.3)
Lower endoscopy	4 (2.5)
Underwent intervention	4 (2.5)
Findings
Gastric and duodenal ulcers	4 (21.1)
Gastritis, duodenitis	4 (21.1)
Angioectasias	3 (15.8)
Rectosigmoid ulcers	1 (5.3)
Diverticulosis	1 (5.3)
Hemorrhoids	2 (10.5)
Normal	4 (21.1)
Weight measurements
Median weight on admission, lb (IQR)	156.4 (131.7 to 193)
Median weight loss 3 mo, lb (IQR)	-4.9 (-16.7 to +6.6)
Median weight loss 6 mo, lb (IQR)	-2.2 (-19.6 to 11.8)
Patients unable to gain weight
3 mo, n (%)	81 (59.1)
6 mo, n (%)	44 (56.4)
Median weight loss 3 mo, lb (IQR)	-14.7 (-26.6 to -7.9)
Median weight loss 6 mo, lb (IQR)	-17.8 (-35.2 to -6.5)
Patients with malnutrition
Malnutrition + mechanical ventilation	36 (19.9)
Malnutrition + BIPAP	11 (6.1)
Malnutrition + vasopressors	41 (22.7)

BIPAP, bilevel positive airway pressure; GI, gastrointestinal; IQR, interquartile range.

Initial GI symptoms were as follows: gastroenteritis in 414 (52.5 %), GI bleeding in 161 (20.4 %), malnutrition in 181 (23.0 %), and idiopathic pancreatitis in 4 (0.5 %) patients. Gastroenteritis resolved in 323 (90.5 %) patients at 3 months and 210 (89.4 %) at 6 months. GI bleeding resolved in 138 (92.0 %) patients at 3 months and 89 (94.7 %) at 6 months. Malnutrition remains in 81 (53.6 %) patients at 3 months and 33 (32.4 %) at 6 months. Pancreatitis attributed to a viral etiology resolved in all patients at 3 months and 6 months.

For patients with malnutrition, median weight on admission was 156.4 pounds (interquartile range [IQR], 131.7–193.0). Median weight loss for patients at 3 months was -4.9 pounds (IQR, -16.7 to +6.6) and at 6 months -2.2 pounds (IQR, -19.6 to 11.8), respectively. Interestingly, 81 (59.1 %) patients were unable to gain weight at 3 months and 44 (56.4 %) were unable to gain weight at 6 months. For patients with malnutrition unable to gain weight at follow-up, median weight loss was -14.7 pounds (IQR, -26.6 to -7.9) at 3 months and -17.8 pounds (IQR, -35.2 to -6.5) at 6 months.

Discussion

We report the following important findings from 3- and 6-month follow-up regarding GI symptoms and COVID-19 disease:

GI manifestations of COVID-19 disease are common (18.5 % in our cohort).

Most initial presentations of GI bleeding, gastroenteritis, and pancreatitis associated with COVID-19 infection resolve by 3-months follow-up.

Malnutrition is the most persistent GI sequelae without resolution at 3- and 6-months follow-up intervals. A significant portion of patients with these complaints may have difficulty gaining weight long term (ie, a median 17.8-pound weight loss remained for these patients at 6-months follow-up).

Our study provides reassurance that most individuals with the onset of GI bleeding or gastroenteritis associated with COVID-19 infection will have resolution of these symptoms. Our study highlights the need to pay particular attention to COVID-19 patients who suffer from malnutrition during their initial hospitalization. Many of these patients may have ongoing malnutrition or weight loss despite resolution of COVID-19 infection. Therefore, it may be imperative to establish malnutrition screening practices in post-COVID 19 patients who have recovered from acute infection. ,

The strengths of our study include the largest cohort of patients with follow-up data of COVID-19-associated GI symptoms to date; a diverse patient population across multiple hospitals in Long Island, Manhattan, Queens, and Staten Island making results more generalizable; and manual chart review to confirm GI findings on admission and follow-up visits. Moreover, patients were diagnosed with malnutrition by a dedicated nutrition service. Limitations of our study lie inherently in the retrospective design. Other considerations include our follow-up data being restricted to our network facilities (although they are quite expansive spanning across Long Island and New York City) and 6-month follow-up is not yet available for several patients who developed COVID-19 infection later in the study interval.

In conclusion, GI symptoms of malnutrition, weight loss, and anorexia may persist for several months after COVID-19 infection and may require further medical attention, whereas GI bleeding, gastroenteritis, and pancreatitis are likely to resolve after initial presentation.