Patients hospitalized with advanced cirrhosis, a chronic and degenerative disease of the liver, are at increased risk of death. The tools currently used to assess that risk are limited in predicting which patients will need a liver transplant and which will be healthy enough to survive transplantation. A new study from the Liver Center at Beth Israel Deaconess Medical Center (BIDMC) suggests that standard assessments that nurses already use to care for patients can be mined for data that significantly improve the ability to predict survival following transplantation and may help improve patient outcomes.
The study appears online April 8 in the journal Hepatology.
"Liver specialists currently use blood tests to assess risk, and while these are powerful tools, they only capture a small portion of the patient's risk. There's a sense that there are things outside of those blood tests that are meaningful," says lead author Elliot Tapper, MD, Clinical Fellow in Gastroenterology and Hepatology at BIDMC and Harvard Medical School. "We hypothesized that frailty, or decreased functional reserve, may be one of those factors, and it turns out that our nurses have been calculating that frailty all along, and we just haven't noticed."
Tapper and his colleagues looked at data from 734 patients admitted a collective 1,348 times to BIDMC's liver unit between 2010 and 2013. They had access to the patients' MELD (model for end-stage liver disease) scores, an algorithm based on blood tests that is used for transplant planning. The higher the MELD score, the more likely the patient will die without a liver transplant.
The researchers also had access to data collected by nurses and used for general care of the patients but not specifically related to liver disease, including assessments used to determine a patient's risk for skin ulcers or whether the patient needs help transferring out of bed or getting to the bathroom.
"These data points help determine a patient's level of frailty and inform very important nursing activities, but until now no one has actually looked at whether those metrics speak to a patient's overall risk," says Tapper.
When the researchers combined traditional assessment tools like MELD with the data from the nurses' assessments, they were able to predict the 90-day mortality rate with 83 percent accuracy and the rate at which patients were discharged to a rehabilitation facility at 85 percent accuracy.
"This is a very significant improvement in our power to provide prognosis, be it 90-day survival, discharge to a rehabilitation hospital or length of stay in the hospital based on a simple inexpensive clinical tool that is readily available in all hospitals across the country," says senior author Michelle Lai, MD, MPH, a liver specialist in the Division of Gastroenterology and Hepatology at BIDMC and Assistant Professor of Medicine at Harvard Medical School.
And that's where the hope comes in. In harnessing the information in the nursing assessments, they have landed on something that can be acted on.
"Many of the things identified in the nursing assessments are modifiable," says Tapper. "By intensifying their nutritional support, utilizing physical therapy, by identifying the patients who will benefit from rehabilitation, we may be able to improve their frailty and therefore their ability to survive, for example, the liver transplant or at least to stay healthier after discharge, at home and away from the hospital."
Foundation receives $100,000 to advance MS care