Age and American Society of Anesthesiologists (ASA) status may be useful in predicting which patients will not survive an anastomotic leak after colectomy for colorectal cancer (CRC), but the addition of other variables associated with frailty do not appear to be of much use, according to new research.
“As we all know, predicting which parents will experience an anastomotic leak is notoriously unreliable. Perhaps it’s better to ask then how likely it is that a patient will survive one,” said Richard Spence, MD, PhD, a surgical oncology fellow at the University of Toronto.
“This falls under the realm of discussion with regard to the frailty of my patient. There are many definitions of frailty in the literature, but one, of course, is the capacity to overcome the physiological insult of a surgical complication such as an anastomotic leak,” Dr. Spence said in a presentation during the Society for Surgical Oncology’s 2020 virtual meeting.
The American College of Surgeons (ACS) National Quality Improvement Program (NSQIP) modified frailty index-5 (mFI-5), which considers functional status, diabetes, hypertension requiring treatment, and a history of chronic obstructive pulmonary disease and congestive heart disease, has not been evaluated in determining failure to rescue in the event of an anastomotic leak after colectomy for CRC.
“We wanted to determine the predictive performance of the ACS-NSQIP mFI-5 in determining failure to rescue anastomotic leak following colectomy for CRC,” Dr. Spence said.
To do so, he and his colleagues compared the predictive performance of three models (age and ASA status; age, ASA status and mFI-5; and the ACS-NSQIP mortality prediction, which includes more than 150 variables) by analyzing the area under the receiver operating characteristics (AUROC) for each model.
Reviewing data on 50,944 patients in the NSQIP database who underwent colectomy for CRC between 2012 and 2016, Dr. Spence and his colleagues identified 1,755 patients (3.46%) who experienced an anastomotic leak, among whom 113 (6.44%) were characterized as failure to rescue (FTR).
In this unfortunate group, older age and a median ASA III classification were strongly associated with FTR. About 20% of these patients had diabetes, and more than half were on hypertensive treatment—characteristics associated with frailty according to the mFI-5. But analyzing the AUROC, Dr. Spence and his colleagues found no statistical differences in the curves of the three models.
“Age and ASA status appear to be the most reliable predictors of FTR and anastomotic leak after colectomy for CRC. The addition of the mFI-5 and all the variables collected by NSQIP don’t significantly improve predictive performance,” he said.
Dr. Spence acknowledged some limitations of the research, including its retrospective nature and the fact that they could not confirm that every fatality in patients with an anastomotic leak was due to the leak itself. He plans further research of prospectively validated risk calculators and variables to better predict FTR anastomotic leaks, including day of postoperative leaks.
Please log in to post a comment