National Early Warning Score discriminates deterioration of inpatients with liver disease
, within 24 hours in hospitalized patients, according to a study published in the journal Clinical Gastroenterology and Hepatology.
Although existing predictive models estimate medium- and long-term prognosis in patients with liver disease, none are validated for short-term outcomes in inpatient settings. Despite the potential concern that patients with liver disease present with chronic physiological derangements affecting National Early Warning Score (NEWS) parameters, the score accurately discriminated patients at risk of death, admission to the ICU, or cardiac arrest within a 24-hour period for a range of liver-related diagnoses achieving an area under receiver operating characteristics (AUROC) curve of 0.894 (95% confidence intervals, 0.887-0.902), 0.857 (95% CI, 0.847-0.868), and 0.722 (95% CI, 0.685-0.759), respectively.
NEWS was launched in 2012 to reduce variations among existing early warning scores by the Royal College of Physicians for use in all adults except pregnant women. Theresa J. Hydes MD, PhD, and her colleagues at the department of gastroenterology & hepatology, Portsmouth (England) Hospitals NHS Trust, sought to validate NEWS in 35,585 unselected medical patients.
The NEWS allocates weighted points, based on derangement of vital signs from defined normal ranges including pulse, respiratory rate, systolic blood pressure, an alert-verbal-painful-unresponsive scale, temperature, peripheral oxygen saturations, and use of supplemental oxygen. The summation of allocated points to directs changes in the level of care, for example, more frequent monitoring, involving senior staff, or calling a rapid response team.
The study was performed in a large, acute care hospital in southern England and analyzed a database of electronically captured vital signs recorded in real time from completed consecutive admissions of patients aged 16 years or older, during 2010-2014. International Classification of Diseases-10 codes were used to categorize patients for any finished consultant episode. The categories included patients with primary diagnosis of liver disease, those with nonprimary liver diagnosis (comorbidity), and patients not allocated any liver disease codes (controls). The NEWS performance was examined according to whether liver disease was acute or chronic, alcohol induced, or associated with portal hypertension, with four clinical subgroups: acute alcohol-induced liver injury, other acute injury, chronic liver disease without cirrhosis, or cirrhosis. The final dataset comprised 722 patients (1,112 episodes) with a primary liver diagnosis, and 2,339 patients (3,658 episodes) with a nonprimary liver diagnosis.
The primary study endpoint was any of the following events occurring within 24 hours of an observation set: in-hospital mortality, unanticipated ICU admission, or cardiac arrest. “The NEWS identified patients with primary, nonprimary, and no diagnoses of liver disease with AUROC values of 0.873 (95% CI, 0.860-0.886), 0.898 (95% CI, 0.891-0.905), and 0.879 (95% CI, 0.877-0.881), respectively. High AUROC values also were obtained for all clinical subgroups; the NEWS identified patients with alcohol-related liver disease with an AUROC value of 0.927 (95% CI, 0.912-0.941). The NEWS identified patients with liver diseases with higher AUROC values than [did] other early warning scoring systems,” according to Dr. Hydes and her colleagues.
“Because of its widespread use, the NEWS serves a ready-made, easy-to-use option for identifying patients with liver disease who require early assessment and intervention, without the need to modify parameters, weightings or escalation criteria,” wrote the authors.
The study was supported by VitalPAC, a collaborative development of the Learning Clinic and Portsmouth Hospitals NHS Trust. Dr. Schmidt, Dr. Aspinall, and Dr. Meredith are employed by PHT. Dr. Hydes had no conflicts of interest.
SOURCE: Hydes TJ et al. Clin Gastroenterol Hepatol. 2017 Dec 22. doi: 10.1016/j.cgh.2017.12.035.
