BOSTON – Phosphate levels following colorectal surgery are predictive of risk for intra-abdominal infections, and can be helpful in identifying low-risk patients who may be suitable for early discharge after surgery, investigators say.
“The association between hypophosphatemia and intra-abdominal infections after colorectal resection is a novel finding,” said Dr. Eran Sadot, a visiting surgical oncology fellow at Memorial Sloan Kettering Cancer Center in New York City.
He described an intra-abdominal infection (IAI) risk-prediction tool at the annual Society of Surgical Oncology Cancer Symposium.
An estimated 5%-15% of patients who undergo colorectal surgery develop an IAI, largely from anastomotic leak, fistula, or intra-abdominal abscess. IAIs are associated with prolonged length of stay and higher costs, as well as an estimated 20% increase in short-term mortality, and reduced long-term survival, he said.
“These complications typically become clinically evident beyond postop day 5, and early identification of intra-abdominal infections can potentially lead to early intervention and limit sepsis, “ he said.
Several studies have shown that hypophosphatemia is associated with poor outcomes in various clinical settings – among hospitalized patients in general, in surgical and cardiac intensive care units, following open heart surgery, and after hepatic resections, he noted.
The investigators hypothesized that patients who develop IAIs have an intense acute-phase response accompanied by hypophosphatemia, and that early measurement of postoperative phosphate levels could serve as a marker for systemic response and early IAI. They conducted a retrospective study of data on patients who underwent first colorectal resection at their center from 2005 through 2015. They looked at postoperative hypophosphatemia, defined as serum levels less than 2.5 mg/dL, and they used logistic regression to create a risk model.
The sample included 7,423 patients with a median age of 61 years, including 42% who underwent resection for colon cancer, 26% for rectal cancer, and the remainder for various diagnoses.
In all, 399 patients (5%) developed IAIs, and two of these patients (0.5%) died.
The authors looked at the course of perioperative serum phosphate levels and saw that all patients had a slight rise in phosphate levels on the day of surgery, which then dropped rapidly to a nadir on postoperative day, and began to recover on day 3. They found that patients who did not have IAIs had more rapid recovery of phosphate levels than did patients who developed infections. In addition, they found that hypophosphatemia on postoperative day 3 was associated with a 50% increased risk of IAI (P = .001).
In a multivariable model of the cohort characteristics stratified by IAI, the researchers saw that risk factors independently associated with IAI included body mass index greater than 30 kg.m2 (odds ratio [OR] 1.4, P = .04), combined liver resection (OR 1.9, P less than .001), estimated blood loss greater than 400 mL (OR 1.7, P = .01), hypophosphatemia on postoperative day 3 (OR 1.4, P = .03), and abnormal white blood cell count on postoperative day 3 (OR 1.9, P less than .001).
They then created an IAI risk score assigning 1 point each to the risk factors just mentioned and tested it as a risk prediction tool. A score of 0 or 1 was associated with a low, 5% predicted risk of IAI, giving the score a negative predictive value of 95%. Scores of 2-3 were associated with moderate risk of IAI (9%-15%), and scores of 4 or 5 were associated with high risk (17%-21%) of infection.
The area under the receiver operating characteristic curve was 0.66 (95% confidence interval 0.61-0.7), indicating a test with a good mix of sensitivity and specificity.
“This model may be used in conjunction with the recently developed enhanced recovery or ERAS pathways to safely reduce the hospital stay,” Dr. Sadot concluded.
The study was internally supported. Dr. Sadot and colleagues reported no conflicts of interest.