This article is more than 5 years old. Information may no longer be current.

Early warning score accurately predicts mortality risk in liver disease

The National Early Warning Score accurately identified patients with chronic liver diseases at risk for death, admission to the intensive care unit or cardiac arrest within a 24-hour period, according to a recently published study.

“Hospitalized patients with chronic liver disease (CLD) can rapidly deteriorate, particularly when acute decompensation is accompanied by extrahepatic organ dysfunction, a situation associated with high mortality,” Theresa J. Hydes, MBBS, from the University of Southampton, United Kingdom, and colleagues wrote. “Early recognition of clinical deterioration is vital if effective, goal-directed therapies are to be employed before complications develop.”

The National Early Warning Score (NEWS) calculations include pulse, respiratory rate, systolic blood pressure, the Alert-Verbal-Painful-Unresponsive scale, temperature, peripheral oxygen saturations and use of supplemental oxygen.

Hydes and colleagues prospectively analyzed the vital signs of adult inpatients with a primary diagnosis of liver disease (n = 722), a non-primary liver disease diagnosis (n = 2,339), and patients with no liver disease as controls.

Compared with patients without liver disease (AUROC = 0.879; 95% CI, 0.877-0.881), NEWS performed equally well among patients with a primary liver disease diagnosis (AUROC = 0.873; 95% CI, 0.86-0.886) and those with a non-primary diagnosis (AUROC = 0.898; 95% CI, 0.891-0.905).

In subclinical analysis, the researchers observed that NEWS performed well in patients with primary diagnoses of nonalcohol-related liver injury, chronic liver disease, acute alcoholic-induced liver injury or cirrhosis compared with patients without liver disease. Results were similar for those with non-primary diagnoses of the same diseases.

“While the efficiency of NEWS was slightly reduced in cirrhosis compared to other subgroups, its ability to identify acute deterioration remained high,” the researchers wrote.

Additionally, the researchers analyzed a subgroup of 1,136 patients with a primary diagnosis and 4,486 patients with a non-primary diagnosis of alcohol-related liver disease, other liver disease, hepatitis, viral infection, autoimmune disease or drug-induced liver injury.

NEWS performed equally well among subgroup patients with primary diagnoses (AUROC = 0.886; 95% CI, 0.857-0.896) and non-primary diagnoses (AUROC = 0.88; 95% CI, 0.874-0.885), especially those with a primary (AUROC = 0.902; 95% CI, 0.889-0.916) or non-primary diagnosis of alcohol-related liver injury (AUROC = 0.915; 95% CI, 0.903-928).

NEWS performed better among patients with primary or non-primary liver disease diagnosis, those with primary or secondary diagnosis of acute alcohol-related injury or chronic liver disease, and those with alcohol- and nonalcohol-related liver disease compared with 34 other early warning sign scores.

“Our study was designed to test the hypothesis that NEWS might not accurately predict serious events in patients with liver disease due to pre-existing altered physiology associated with the underlying condition,” the researchers concluded. “This hypothesis was disproven and we were encouraged to find NEWS remained a highly accurate discriminator of adverse events in liver disorders, with its performance being highest in [alcohol-related liver disease.]” – by Talitha Bennett

Disclosure: Hydes reports no relevant financial disclosures. Please see the full study for the other authors’ relevant financial disclosures.