Test Predicts Mortality Risk of Bariatric Surgery
Results from a multicenter study prove that a standard preoperative test can accurately assess bariatric surgery patients’ risk for death.
Investigators reported at the 2007 annual meeting of the American Surgical Association, held in Colorado Springs, Colo., that the Obesity Surgery Mortality Risk Score (OS-MRS) identifies patients most likely to die in the 30 days or three months after surgery. The score ranks patients into three risk categories based on their preoperative risk factors.
“This is the first system validated by multicenters that appears to differentiate high- and low-risk patients,” said the study’s lead author Eric J. DeMaria, MD, professor of surgery and director of endosurgery and bariatric surgery, Duke University, Durham, N.C.
The system is simple and can be performed easily at the patient’s bedside, he said. Surgeons welcomed the score system, saying it will help them explain risks to patients.
But the score’s most important use may be as a tool for assessing outcomes for surgeons and hospitals, several bariatric specialists predicted.
“There is a great deal of variability in the risks of bariatric patients. Some centers may have higher mortality rates because they take on higher risk patients. This is a way to show that,” said Edward Livingston, MD, professor and chairman of gastrointestinal and endocrine surgery, University of Texas Southwestern Medical School, Dallas.
The OS-MRS calculates a patient’s mortality risk based on five factors—male gender; age older than 45 years of age; BMI>/=50; a diagnosis of hypertension; or the presence of risk factors for pulmonary embolism, defined as venous stasis or ulcers, obesity hypoventilation, pulmonary hypertension, or a previous deep vein thrombosis, pulmonary embolism or inferior vena cava filter.
Patients are scored one point for every risk factor and ranked into three risk categories based on their score: 0 to 1 points is low risk or Class A, 2 to 3 points is medium risk or Class B, and 4 to 5 points is high risk or Class C. The score system predicts only mortality and not morbidity.
The OS-MRS was described in detail in a report earlier this year (Surg Obes Relat Dis2007;3:134-140). However, the current study was the first to test the score in patients outside of the Medical College of Virginia, where the OS-MRS was created.
The investigators studied data from bariatric centers at four hospitals. The group consisted of 4,431 gastric bypass patients. Of these, the researchers grouped 2,166 as low-risk patients, 2,140 as mid-risk patients, and 125 as high-risk patients based on OS-MRS scores.
Results showed deaths rose incrementally among patients with a greater number of risk factors. Mortality was 0.2% for low-risk patients, 1.2% for mid-risk patients and doubled to 2.4% for the highest-risk patients (P<0.05).
Several surgeons pointed out that the study showed high-risk patients and some mid-risk patients die at significantly higher rates than expected based on published reports. A 2004 study reported the risk of death from bariatric surgical treatment at 0.5% (JAMA 2004;292:1724-1737).
“This may be an issue. The mortality rate is much higher than the rate usually quoted by the bariatric society,” said Carson D. Liu, MD, the medical director for the Surgical Weight Control Center in Los Angeles.
The study revealed a wide disparity in mortality rates although most hospitals reported rates higher than the 0.5% benchmark. One hospital where no high-risk patients were treated reported the lowest mortality rate of 0.4% (5 of 1,320 patients). Another center that treated a majority of mid- to high-risk patients reported a 2.0% mortality rate (17 of 816). Deaths at this hospital were highest in the mid-risk patients (3.3%), which was slightly more than the high-risk patients (3.2%) and significantly higher than the low-risk (0.8%). The surgeons may have been early in their learning curve at that hospital, although that was not clear from the study, said Dr. Liu.
Pulmonary embolism was the No. 1 cause of death, accounting for 10 of the 33 deaths reported. Other common causes of mortality were cardiac events (9) and GI leaks (7).
Most studies of bariatric patients report an average mortality rate that includes all patients, Dr. DeMaria noted. His study looked at mortality rates across different risk groups, which accounts for the higher mortality rate among high-risk patients. “The high-risk group of patients is small overall, but they are there,” he said. “As bariatric surgeons, we need to look at what happens to patients other than the average-risk patients.”
Dr. DeMaria added that high-risk patients should still be considered for bariatric procedures. The scoring system could lead to strategies to reduce mortality among high-risk patients, he said. Risk-reduction strategies could mean that patients are treated when they are younger and healthier, or that patients must lose weight before surgery in order to reduce their risk. The OS-MRS score could cause surgeons to consider a different procedure for high-risk patients. Some high-risk patients may be better suited for a sleeve gastrectomy or gastric band than a bypass, he said.
The score system could eventually be rolled into the process for assessing and credentialing centers of excellence. Today, both the American College of Surgeons Bariatric Network and ASBS Surgical Review Corporation rely heavily on volume data and mortality outcomes to grant centers of excellence status. (Several of the study authors are members of the ASBS Surgical Review Corporation.) Many surgeons have argued that credentialing bodies for bariatric surgery need to give more consideration to the types of patients that surgeons are operating on.
Michael Sarr, MD, professor of surgery, Mayo Clinic, Rochester, Minn. specializes in open surgery for patients considered high risk. Most patients who come to him are heavier than the average bariatric patient and have complicating factors such as hernias or previous bariatric surgery. Their higher risk is reflected in his outcomes—his patients tend to have higher mortality and morbidity than those treated by other surgeons in his hospital.
“If you looked only at my outcomes, you would, say, Ôgeez, he’s a terrible surgeon,’ whereas if you look at a small hospital where they take people with no major life-threatening problems, the mortality rate may be very low. So this tool is very, very important,” he said.
“Prior to this, we have not had anything that we can use to predict the mortality of an individual, or the mortality experience at an individual institution.”
He added that the score system will help communication with patients, payers, and credentialing bodies. It could be added to the informed consent process, for example.
The full results of Dr. DeMaria’s study will be published this fall in the Annals of Surgery.