HOLLYWOOD, FLA. – A three-parameter scoring system predicts which patients are likely to experience complications from surgery for a small bowel obstruction.
The new tool – dubbed FAS (Functional status, American Society of Anesthesiologists [ASA] classification, and Sepsis) – focuses mostly on preoperative functional status and the presence of preoperative sepsis. It’s as accurate as a time-consuming 10-item Margenthaler system published in 2006, which requires data on blood chemistry, neurologic status, and cardiac and lung function as well as age, sepsis, and preoperative functional measures.
Small bowel obstruction is a common problem, said Dr. Asuzu, who is also a medical student at Yale University, New Haven, Conn. Whether to treat conservatively or surgically can be a complex decision. “Conservative treatment avoids postoperative complications, but there is a higher risk of occurrence and a quicker time to recurrence than with surgery. But surgery carries its own risks. If we could identify patients at high risk for complications, then perhaps we could push those patients more toward conservative treatment.”
Theattempted to do just that. It was retrospectively validated in 2,000 patients included in the Veterans Affairs Surgical Quality Improvement Program database (VASQIP) who underwent surgery for small bowel obstruction. The authors examined about 60 clinical factors associated with postsurgical morbidity and mortality, finally settling on 10 that, when scored, accurately predicted 30-day morbidity and mortality.
These factors were:
• History of congestive heart failure
• Neurological deficit or stroke
• Chronic obstructive pulmonary disease
• Elevated white cell count
• Preoperative functional health status
• Surgery type
• Preoperative creatinine
• Wound classification
• ASA class
Dr. Asuzu and his mentor, Kevin Y. Pei, MD, FACS, wanted to come up with a more user-friendly risk assessment tool for patients undergoing open small bowel adhesiolysis. They focused on two measures of preoperative functional status: dependent vs. independent and ASA classification. Another measure – preoperative sepsis – estimated the impact of the patient’s current medical problem.
The tool was tested retrospectively in two independent cohorts extracted from the ACS National Surgery Quality Improvement Project (NSQIP) database. The initial discovery cohort comprised 6,036 patients; the replication cohort, 9,000. These patients had a mean age of 60 years and were relatively healthy, with low rates of congestive obstructive pulmonary disease, renal failure, cancer, bleeding disorders, and ascites. About half were taking antihypertensive medications and 5%, steroids.
Using multivariable regression, the authors developed a scoring system as follows:
• 6 points for each level of preoperative functional status (1 – independent, 2 – partially dependent, 3 – totally dependent)
• 6 points for each level of ASA classification (1 – no disturbance, 2 – mild disturbance, 3 – severe disturbance, 4 – life-threatening disturbance, and 5 – moribund state)
• 4 points for each level of perioperative sepsis (1 – systemic inflammatory response syndrome [SIRS], 2 – sepsis, 3 – septic shock)
In the discovery cohort, the three-item FAS tool was just as accurate as the Margenthaler tool, with an odds ratio of 1.11 vs 1.10 for any complication. The areas under the curve were 0.69 vs. 0.68. These results were virtually identical in the replication cohort.
With a combined total score of 32 as the cutoff, FAS yielded a specificity of 93% for predicting any complication and 92% for any of the six most common complications (ventilator dependence greater than 48 hours, pneumonia, superficial surgical site infection, postoperative sepsis, urinary tract infection, or unplanned intubation) in the replication cohort. The positive predictive value was 50% for any complication and 45% for the six most common complications, and the negative predictive values were 81% and nearly 85%, respectively.
“We are very pleased with how this performs,” Dr. Asuzu said in an interview. “It’s apparent that these three parameters are sufficient to tell us with a high level of specificity which patients could benefit from a more conservative approach. The next step is to prospectively validate it in a single center dataset.”
He said discriminating the most meaningful risk factors plainly showed that preoperative physical status is the best indicator of how well a patient will handle the surgery.
“It turns out that the biggest predictor of you how do after surgery is how you are doing before surgery. We can look at it as the how big the hit is, and the patient’s ability to take that hit. If their ability is already compromised, it’s a sign they might not do well.”
The “functional status” parameter may seem overly simplistic at first glance, he said. “But it really takes into account everything: the gout, the hypertension, the smoking, heart and respiratory and kidney function. All of this plays a role in functional status. I think this is why some of these more complex scores suffer. They’re not clear because there is so much overlap there.”
Dr. Asuzu had no financial disclosures.