key: cord-0821482-ik2wlk61
authors: Sterling, Richard K; Oakes, Tavis; Gal, Tamas S; Stevens, Michael P; deWit, Marjolein; Sanyal, Arun J
title: The FIB-4 Index Is Associated with Need for Mechanical Ventilation and 30-day Mortality in Patients Admitted with COVID-19
date: 2020-08-28
journal: J Infect Dis
DOI: 10.1093/infdis/jiaa550
sha: 3ebb72fe41f29d0d0477e7f9e52742e31836dcb9
doc_id: 821482
cord_uid: ik2wlk61

The fibrosis-4 index (FIB-4), developed to predict fibrosis in liver disease, was used to identify patients with COVID-19 who will require ventilator support as well as associated with 30-day mortality. Multivariate analysis found obesity (OR 4.5), diabetes (OR 2.55), and FIB-4 ≥ 2.67 (OR 3.09) independently associated with need for mechanical ventilation. When controlling for ventilator use, gender, and comorbid conditions, FIB-4 ≥ 2.67 was also associated with increased 30-day mortality (OR 8.4; 95% CI 2.23-31.7). While it may not be measuring hepatic fibrosis, its components suggest that increases in FIB-4 may be reflecting systemic inflammation associated with poor outcomes

Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes coronavirus disease 2019 (COVID-19), can be associated with a severe systemic disease leading to respiratory failure and the need for mechanical ventilation [1] . Patients with underlying medical comorbidities, such as respiratory, cardiac, and liver disease, diabetes mellitus (DM), and obesity are at higher risk for respiratory failure [1] [2] [3] [4] [5] . Therefore, prediction factors are needed to help front line providers to identify who might be at higher risk for intensive care and ventilator support for respiratory failure.

COVID-19 has been associated with liver injury [6] . The fibrosis-4 index (FIB-4), developed to predict advanced fibrosis in those with liver disease has four components: age, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and platelet count (PLT) [7] . Recent studies show these parameters may be affected by other systemic diseases, such as COVID-19 through non-hepatic mechanisms, including systemic inflammation and/or bone marrow suppression [8] .

Our aims were to determine if FIB-4, a simple tool available to front line providers, would be associated with the need for mechanical ventilator support, and 30-day mortality among hospitalized patients with COVID-19.

Electronic medical records (EMR) were used to identify all patients admitted to Virginia Commonwealth University Medical Center in Richmond, Virginia, from February-May 2020 with confirmed COVID-19 by polymerase chain reaction (PCR).

Because de-identified EMR data were used, institutional review board (IRB) was not required. Demographic data (age, gender, race), body mass index (BMI), obesity A c c e p t e d M a n u s c r i p t (BMI ≥ 30 kg/m 2 ), existing respiratory, cardiac, liver, and diabetes mellitus (DM) disease by ICD-10 codes, from the index hospitalization, any previous hospital or ambulatory visit, and any previous billing diagnosis, AST, ALT, and PLT count were collected on the day of admission. FIB-4 was calculated (age*AST/PLT*ALT 1/2 ) and categorized as < or ≥ 2.67 or < or ≥ 3.25 [7, 9] . Data were presented by mean and standard deviation (SD) or median and interquartile range (IQR) or frequency and percent. [4, 12] .

Of the components of FIB-4, only AST was significantly higher in those who required ventilator support. However, when the components of FIB-4 were substituted in the model, including AST, none were significantly associated with respiratory failure suggesting that it is the combination in the FIB-4 index and not its components that helps predict COVID-19 severity. The relationship of increased AST to ICU admission or ventilator support was seen in other studies [2, 12, 13] . However, unlike at the case of the Veterans' cohort [11] , we did not observed differences in age, ALT, or platelet count. Regardless of these differences, our data confirm the independent association between increased FIB-4 and COVID-19 severity as measured by respiratory failure and need for ventilator support. Unlike the study by Ibanez-Samamiego that used a FIB-4 cutoff of 2.67 for those with NAFLD [9] , we use both the NAFLD threshold as well as the higher cutoff of 3.25 used for viral hepatitis [7] also used by Rentsch [11] . Although our cohort had a high frequency of A c c e p t e d M a n u s c r i p t increased FIB-4 (52% had FIB-4 > 2.67 and 42% had FIB-4 > 3.25), our cohort had low prevalence of known underlying liver disease (6%) which was not higher in those needing ICU or ventilator support. The choice threshold of any biomarker must balance sensitivity and specificity. Because a threshold of 2.67 had higher sensitivity (70%), with similar PPV and NPV compared to 3.25, we recommend it as a tool for future studies where the prevalence of viral hepatitis and known liver disease is low.

However, the strength of FIB-4 may be in its high NPV to identify those who will not need mechanical ventilation. While FIB-4 is not assessing liver fibrosis, it may be a global score for systemic inflammation that has been associated with COVID-19 that is producing increased AST [6, 8, 10, 12] . In support of FIB-4 associated with nonliver related outcomes, a recent study found that increased FIB-4 was associated with outcomes after intracranial hemorrhage while the NAFLD fibrosis score did not [13] , suggesting that FIB-4 has unique properties to assess outcomes not related to fibrosis or NAFLD. Finally, while some studies have observed higher mortality in AA with COVID-19 [14] we did not observe any difference in race or gender between those who required ventilator use and those that did not.

Our study has several important limitations. Our data and outcomes were determined by de-identified data obtained from our EMR, therefore if past medical conditions were not recorded, our data may have been affected. In addition, we did not have historical laboratory data to determine if increased liver enzymes were new and related to COVID-19 or chronic and due to existing liver disease which may impact COVID-19 severity [4] . Furthermore, we could have underestimated chronic liver disease and clinically silent cirrhosis that was not captured by the ICD-10 codes used. Importantly, we also were not able to capture data on need for high flow oxygen upon presentation to the health system. We also were not able to capture A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t M a n u s c r i p t 

for the China Medical Treatment Expert Group for Covid-19. Clinical Characteristics of Coronavirus Disease

JAMA Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease

Patients with diabetes are at higher risk for severe illness from COVID-19

Risk of severe illness from COVID-19 in patients with metabolic dysfunction-associated fatty liver disease and increased fibrosis scores

Obesity Is a Risk Factor for Greater

Diabetes Care

Characteristics and Mechanism of Liver Injury in 2019 Coronavirus Disease

Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. APRICOT Clinical Investigators

Understanding SARS-CoV-2-Mediated Inflammatory Responses: From Mechanisms to Potential Therapeutic Tools

Comparison of noninvasive markers of fibrosis in patients with nonalcoholic fatty liver disease. Nash Clinical Research Network

Association between elevated liver enzymes and C-reactive protein: possible hepatic contribution to systemic inflammation in the metabolic syndrome

Covid-19 Testing, Hospital Admission, and Intensive Care Among 2,026,227 United States Veterans Aged 54-75 Years. medRxiv

Elevation of liver fibrosis index FIB-4 is associated with poor clinical outcomes in patients with COVID-19

Liver Fibrosis Indices and Outcomes After Primary Intracerebral Hemorrhage

Seoane L Hospitalization and Mortality among Black Patients and White Patients with Covid-19

New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality

M a n u s c r i p t A c c e p t e d M a n u s c r i p t