Severe SARS-CoV-2 infection has been linked to immune dysregulation and alterations of the microbiome that could increase susceptibility to secondary infections. The longer-term risk of acute gastroenteritis (AGE) following COVID-19, particularly after hospitalisation, is not well characterised. This study aimed to compare AGE incidence after COVID-19 in community-managed and hospitalised patients against test-negative controls and against patients hospitalised with sepsis.
The investigators performed a retrospective cohort study using electronic healthcare records extracted from the Clinical Practice Research Datalink (CPRD) in England. The study period for index SARS-CoV-2 PCR tests ran from 1 March 2020 to 1 March 2022. Ethical approval was provided by the Independent Scientific and Ethical Committee (ISAC) of CPRD (reference 23_002987), and the work adhered to the Declaration of Helsinki. The dataset is available through formal request to CPRD; the authors do not directly share the extracted data.
Patients with a recorded SARS-CoV-2 PCR test during the study interval were allocated to one of three primary cohorts: test-negative controls, community COVID-19 cases (those managed outside hospital), and hospitalised COVID-19 cases. In addition, a hospitalised sepsis cohort was assembled to enable age-, sex-, and time-matched comparisons with hospitalised COVID-19 patients.
The primary outcome was incidence of acute gastroenteritis (AGE). Incidence rates were calculated for the two years preceding enrolment to establish baseline (pre-pandemic) rates and then during the follow-up period after the index test or hospital admission. The manuscript reports crude incidence calculations and adjusted incidence rate ratios (IRRs) derived from regression models.
Negative binomial regression models and conditional Poisson models were used to estimate incidence rate ratios (IRRs), adjusting for demographic and clinical covariates. Matched analyses compared hospitalised COVID-19 patients to community controls and to hospitalised sepsis controls with matching on age, sex, and time to account for confounding by these factors.
The analysis included 788,103 patients with a recorded SARS-CoV-2 PCR test. In the two-year pre-pandemic period, individuals who went on to be hospitalised with COVID-19 or sepsis had higher baseline AGE incidence than those whose COVID-19 was managed in the community. This baseline difference is important for interpreting follow-up comparisons.
During follow-up after the index event, community-managed COVID-19 cases had a slightly lower AGE incidence than test-negative controls. The adjusted IRR for community COVID-19 versus test-negative controls was 0.88 (95% CI 0.85–0.92). The authors interpret this reduction as likely reflecting baseline differences between groups rather than a protective effect of community COVID-19 infection against AGE.
By contrast, hospitalised COVID-19 patients experienced a substantially higher AGE incidence during follow-up. The adjusted IRR comparing hospitalised COVID-19 patients with test-negative controls was 5.80 (95% CI 5.38–6.25), indicating a marked increase in post-discharge AGE risk.
Matched analyses confirmed elevated AGE rates after COVID-19 hospitalisation relative to matched community controls (IRR 4.63, 95% CI 4.15–5.17). When hospitalised COVID-19 patients were compared to hospitalised sepsis controls, AGE incidence remained higher after COVID-19 hospitalisation (IRR 1.41, 95% CI 1.21–1.64), indicating that the excess risk is not fully explained by severe intercurrent illness or hospitalisation alone.
The findings show that COVID-19 hospitalisation is associated with a substantial increase in subsequent AGE risk, exceeding the post-discharge risk observed after sepsis hospitalisation. The authors suggest these results indicate heightened vulnerability to gastrointestinal illness following severe COVID-19 and support the need for enhanced follow-up and surveillance of patients after hospital discharge. The lower AGE rates observed in community-managed COVID-19 cases are likely attributable to baseline group differences rather than a protective effect of mild infection.
Ethical approval for the study was granted by the ISAC of CPRD. The dataset used in the analysis was obtained under a data sharing agreement between the University of Liverpool and CPRD; the authors note that study data are available via formal request to CPRD but are not shareable directly by the authors. Funding for the study was declared from the National Institute for Health Research (NIHR) with identifier PB-PG-NIHR-200910, NIHR205131.
Disclosures reported in the manuscript include grant support and other engagements: one author received a research grant for rotavirus surveillance from Merck & Co.; some authors received grant support related to influenza vaccine evaluation from Seqirus UK Ltd. One author received honoraria and consulting support related to rotavirus surveillance outside this work. All other authors declared no interests.