Gram Research analysis shows that gut bacteria patterns can predict which liver disease patients will need hospitalization within 90 days with 84% accuracy when combined with standard medical tests. A 2026 multinational study of 679 cirrhosis patients found that those hospitalized had significantly lower bacterial diversity and specific bacterial changes, including enrichment of harmful bacteria like Enterococcus faecium and loss of beneficial bacteria that produce short-chain fatty acids. When doctors combined gut bacteria information with clinical tests, they could predict hospitalizations much better than using either method alone, suggesting that microbiome testing could become a valuable tool for identifying high-risk patients
Researchers studied the gut bacteria of 679 liver disease patients across seven countries to see if bacteria patterns could predict who would need hospitalization. They found that patients who ended up in the hospital within 90 days had different types and amounts of bacteria in their guts compared to those who stayed healthy. By combining information about gut bacteria with standard medical tests, doctors could predict hospitalizations more accurately than using either method alone. This discovery suggests that checking someone’s gut bacteria might help doctors identify which liver disease patients are at highest risk and need closer monitoring.
Key Statistics
A 2026 multinational study of 679 cirrhosis patients found that 25% required hospitalization within 90 days, with hospitalized patients showing significantly lower gut bacterial diversity across all seven countries studied
According to research reviewed by Gram, combining gut bacteria information with standard medical tests improved hospitalization prediction accuracy to 84%, compared to 79% using clinical information alone and 74% using microbiome data alone
A 2026 analysis of 679 liver disease patients identified 19 bacterial species independently linked to hospitalization risk, with hospitalized patients showing enrichment of harmful bacteria and depletion of beneficial short-chain fatty acid-producing taxa
Research shows that hospitalized cirrhosis patients had consistent patterns of reduced bacterial diversity and functional shifts affecting carbohydrate degradation and antibiotic resistance genes across all seven countries, despite marked variation in specific bacterial composition by region
The Quick Take
- What they studied: Whether the types and amounts of bacteria living in the gut of liver disease patients could predict who would need to go to the hospital in the next 90 days
- Who participated: 679 people with advanced liver disease (cirrhosis) from seven different countries who were not currently in the hospital when the study started
- Key finding: One quarter of patients needed hospitalization within 90 days. Those who were hospitalized had fewer types of bacteria in their guts and specific bacterial changes. When doctors combined gut bacteria information with standard medical tests, they could predict hospitalizations with 84% accuracy, which was better than using either method alone
- What it means for you: If you have advanced liver disease, a simple stool test checking your gut bacteria might help your doctor figure out if you’re at high risk for needing hospital care soon. This could lead to earlier treatment and closer monitoring. However, this is still new research and not yet used in regular medical practice
The Research Details
This was a multinational study, meaning researchers worked together across seven countries to collect information from 679 patients with advanced liver disease. Each patient provided a stool sample so scientists could analyze their gut bacteria using advanced genetic testing. The researchers looked at three things: which bacteria were present and in what amounts, what functions those bacteria performed, and whether the bacteria carried genes for antibiotic resistance. Patients were then followed for 90 days to see who ended up needing hospitalization.
The researchers used machine learning, a type of artificial intelligence, to build prediction models. They created three different models: one using only standard medical information, one using only gut bacteria information, and one combining both. This allowed them to see which approach worked best for predicting who would need hospital care.
The study was careful to account for differences between countries, the severity of each patient’s liver disease, what caused their liver disease, and what treatments they were receiving. This helped ensure that any bacteria patterns they found were truly linked to hospitalization risk and not just due to these other factors.
Current methods for predicting which liver disease patients will need hospitalization rely mainly on blood tests and physical exams, but these don’t always work well. By studying gut bacteria across multiple countries, researchers could see if bacteria patterns were a reliable sign of hospitalization risk everywhere, not just in one place. This approach is important because it suggests a new tool doctors could use to identify high-risk patients earlier, potentially allowing for preventive treatment before a crisis happens
This study has several strengths that make its findings reliable. It included a large number of patients (679) from seven different countries, which means the results likely apply to many different populations. The researchers carefully controlled for other factors that might affect the results. The study was published in Gut, a highly respected medical journal. However, the study shows which bacteria patterns are linked to hospitalization but doesn’t prove that the bacteria actually cause the hospitalizations. Also, the study was done in outpatient clinics, so results might differ for sicker patients already in hospitals
What the Results Show
Among the 679 patients studied, 25% required hospitalization within 90 days. The most important finding was that hospitalized patients had lower microbial diversity, meaning they had fewer different types of bacteria in their guts. This pattern held true across all seven countries, even though the specific bacteria present varied by country. This suggests that losing bacterial diversity is a universal sign of increased hospitalization risk in liver disease.
The researchers identified 19 specific bacterial species that were independently linked to hospitalization risk after accounting for disease severity and other factors. Two bacteria were more common in hospitalized patients: Enterococcus faecium and Veillonella rogosae. Meanwhile, many helpful bacteria that normally live in healthy guts were depleted in hospitalized patients. These helpful bacteria, called commensals, normally produce short-chain fatty acids, which are important for gut health.
When researchers looked at what functions the bacteria were performing, they found major changes in how hospitalized patients’ bacteria handled carbohydrates and fats. The bacteria in hospitalized patients also carried more genes related to antibiotic resistance. These functional changes suggest that hospitalization is linked to a fundamental shift in how the gut ecosystem works, not just which bacteria are present.
The most striking result came from the prediction models. A model using only standard medical information correctly predicted hospitalization 79% of the time. A model using only gut bacteria information was slightly less accurate at 74%. But when researchers combined both types of information, the accuracy jumped to 84%, showing that gut bacteria add important new information that doctors don’t get from standard tests alone
Diet had only a modest influence on the microbiome structure, suggesting that the bacterial changes seen in hospitalized patients were primarily driven by the disease itself rather than what people were eating. The study also found that the specific bacteria present varied significantly between countries, likely due to differences in diet, genetics, and environment. However, despite these regional differences in which bacteria were present, the pattern of reduced diversity and functional changes was consistent everywhere. This consistency across countries strengthens the idea that these bacterial patterns are a reliable sign of hospitalization risk globally
Previous research has shown that gut bacteria play a role in liver disease progression, but most studies were done in single countries or smaller groups. This multinational study confirms and expands on those findings by showing that the bacterial patterns are consistent across different populations. The finding that combining bacterial and clinical information improves prediction is consistent with recent trends in medicine toward using multiple types of data together. The specific bacteria identified, particularly the loss of short-chain fatty acid-producing bacteria, aligns with what researchers know about how the gut microbiome affects liver health
While this study is large and well-designed, it has some important limitations. First, it shows that certain bacteria are linked to hospitalization but doesn’t prove that these bacteria actually cause hospitalizations. The study followed patients for only 90 days, so it doesn’t tell us about longer-term outcomes. The prediction models were developed using the same patients who provided the data, so they might not work as well when applied to completely new patients. Additionally, the study was done in outpatient clinics with patients stable enough to be treated outside the hospital, so the results might not apply to sicker patients already hospitalized. Finally, while the study included seven countries, most were in Europe and North America, so results might differ in other parts of the world
The Bottom Line
Based on this research, doctors might consider stool testing to check gut bacteria composition in liver disease patients, particularly those at higher risk. However, this is still experimental and not yet standard medical practice. The research suggests that patients with reduced bacterial diversity or specific bacterial patterns might benefit from closer monitoring and earlier intervention. Maintaining a healthy diet and avoiding unnecessary antibiotics may help preserve beneficial gut bacteria, though the study showed diet has only modest effects. Confidence in these recommendations is moderate because the research is new and needs to be tested in clinical practice
This research is most relevant to people with advanced liver disease (cirrhosis) and their doctors. Gastroenterologists and hepatologists, the specialists who treat liver disease, should be aware of these findings as they develop new ways to predict and prevent hospitalizations. Patients with liver disease who want to understand their hospitalization risk should discuss this research with their doctors. However, this research doesn’t yet apply to people without liver disease or those with early-stage liver disease. The findings are also most applicable to adults, as the study didn’t include children
The study measured hospitalization within 90 days, so if gut bacteria testing were used clinically, doctors would expect to see its predictive value within this timeframe. However, if interventions were developed based on these findings, it might take weeks to months to see improvements in bacterial composition. Longer-term studies would be needed to determine if changing the microbiome actually prevents hospitalizations
Frequently Asked Questions
Can gut bacteria tests predict if I’ll need to go to the hospital if I have liver disease?
Research shows that gut bacteria patterns combined with standard medical tests can predict hospitalization risk within 90 days with 84% accuracy in cirrhosis patients. However, this is still new research not yet used in routine medical practice. Talk to your doctor about whether microbiome testing might be appropriate for your situation
What bacteria changes happen in liver disease patients who get hospitalized?
Hospitalized cirrhosis patients have fewer types of bacteria overall and specific changes including more Enterococcus faecium and Veillonella rogosae, plus loss of beneficial bacteria that produce short-chain fatty acids. These bacteria help with carbohydrate breakdown and fat metabolism, so their loss affects how the gut functions
Does what I eat affect my gut bacteria if I have cirrhosis?
Diet has only a modest influence on gut bacteria composition in cirrhosis patients, according to a 2026 study of 679 patients. This suggests that disease severity drives most bacterial changes, though maintaining good nutrition is still important for overall health
Can antibiotics harm my gut bacteria if I have liver disease?
The study found that hospitalized cirrhosis patients carried more antibiotic resistance genes, suggesting unnecessary antibiotics may harm beneficial bacteria. Discuss with your doctor whether each prescribed antibiotic is truly necessary, as preserving good bacteria may help prevent hospitalization
Is this gut bacteria test available for liver disease patients now?
Microbiome testing for predicting hospitalization in cirrhosis is still experimental and not yet standard medical practice. These findings are from 2026 research that needs further development before becoming widely available. Ask your hepatologist if they’re aware of any clinical trials testing this approach
Want to Apply This Research?
- Users with liver disease could track their hospitalization risk factors weekly, including symptoms like fatigue or swelling, medication adherence, and diet quality. If they had access to microbiome test results, they could log the date of testing and results to monitor changes over time
- Users could use the app to set reminders for maintaining a fiber-rich diet, which supports beneficial gut bacteria. They could also track antibiotic use and discuss with their doctor whether each antibiotic is truly necessary, since unnecessary antibiotics can harm beneficial bacteria. The app could provide education about foods that support short-chain fatty acid-producing bacteria
- For long-term tracking, users could log quarterly or semi-annual microbiome test results if their doctor orders them, along with hospitalization events and emergency room visits. The app could help identify patterns between microbiome changes and health events, providing data to share with their healthcare team
This research describes associations between gut bacteria and hospitalization in liver disease patients but does not establish that bacteria cause hospitalizations. Microbiome testing for predicting hospitalization is not yet standard medical practice and should not replace evaluation by your hepatologist or gastroenterologist. If you have cirrhosis or advanced liver disease, discuss these findings with your doctor to determine if they apply to your situation. This article is for educational purposes and should not be considered medical advice. Always consult with qualified healthcare providers before making decisions about your care
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.