Id |
Subject |
Object |
Predicate |
Lexical cue |
T297 |
0-108 |
Sentence |
denotes |
Limitations of the Evidence on Gastrointestinal and Liver Manifestations in Patients With COVID-19 Infection |
T298 |
109-174 |
Sentence |
denotes |
The individual studies in our analysis were at high risk of bias. |
T299 |
175-316 |
Sentence |
denotes |
The majority of studies reported on cohorts of patients based on inclusion dates and did not specify whether these were consecutive patients. |
T300 |
317-518 |
Sentence |
denotes |
There was an inconsistent assessment of symptoms and/or laboratory tests with missing data, and none of the studies reported whether patients were systematically evaluated for GI symptoms on admission. |
T301 |
519-609 |
Sentence |
denotes |
Most studies did not report on the duration of the GI symptoms preceding the presentation. |
T302 |
610-814 |
Sentence |
denotes |
When GI symptoms were reported, it was difficult to discern whether these were isolated symptoms or whether patients also had concurrent typical COVID-19 symptoms (eg, fever cough or shortness of breath). |
T303 |
815-933 |
Sentence |
denotes |
LFTs were mostly reported as the mean/median value of the entire cohort and without cutoff values for the institution. |
T304 |
934-1012 |
Sentence |
denotes |
Many of the studies did not report on underlying chronic GI or liver diseases. |
T305 |
1013-1152 |
Sentence |
denotes |
There was a lot of heterogeneity in our pooled estimates that could not be explained by our subgroup analysis based on geographic location. |
T306 |
1153-1348 |
Sentence |
denotes |
Lastly, the data on prognosis were especially difficult to analyze due to insufficient follow-up of the patients (the majority of the patients were still hospitalized at the time of publication). |
T307 |
1349-1514 |
Sentence |
denotes |
Finally, there was no stratification of GI-related symptoms and severity of COVID-19 or patient important outcomes, such as need for intensive care unit or survival. |
T308 |
1515-1589 |
Sentence |
denotes |
There may be additional limitations of our findings based on our analysis. |
T309 |
1590-1794 |
Sentence |
denotes |
Due to concerns about overlapping cohorts, we used a hierarchical framework to identify unique cohorts based on the number of patients and the hospitals to analyze the prevalence of GI and liver symptoms. |
T310 |
1795-1964 |
Sentence |
denotes |
It is possible that we excluded relevant studies that provided more granularity regarding the GI and liver manifestations, or had more systematic assessment of outcomes. |
T311 |
1965-2058 |
Sentence |
denotes |
As a result, this may have led to an over- or underestimation of the pooled effect estimates. |
T312 |
2059-2267 |
Sentence |
denotes |
However, we have high confidence that we were able to eliminate the counting of some patients in more than 1 report by using our selection framework, unless they were transferred from one hospital to another. |
T313 |
2268-2369 |
Sentence |
denotes |
An important strength of this study is the appropriate statistical analysis used to pool proportions. |
T314 |
2370-2611 |
Sentence |
denotes |
We also reviewed gray literature from prepublication repositories, which allowed us to include a large number of studies that have not been published yet, with data from a total of 10,890 unique COVID-19 patients being included in this work. |
T315 |
2612-2838 |
Sentence |
denotes |
Lastly, we tried to narratively describe studies that informed us on the type of diarrhea symptoms; whether diarrhea was reported as the only presenting symptom; or diarrhea as the initial symptom that preceded other symptoms. |
T316 |
2839-3124 |
Sentence |
denotes |
Based on our study selection process, we may have missed studies, including smaller case series that reported on this information, and studies that were published after our inclusion period, in light of the exponential number of studies in press, under review, and on preprint servers. |