Is Preoperative Serum Albumin Predictive of Adverse Outcomes in Head and Neck Cancer Surgery?

ABSTRACT

BACKGROUND: Patients with head and neck cancer are at increased risk of malnutrition due to tumor burden and surgical morbidity. PURPOSE: The purpose of this study was to evaluate the association between preoperative serum albumin and 30-day adverse outcomes in patients undergoing head and neck cancer surgery. STUDY DESIGN, SETTING, SAMPLE This was a retrospective cohort study using the American College of Surgeons National Surgical Quality Improvement Program database. Patients undergoing an ablative head and neck cancer procedure were included. Patients who had an unclear tumor location based on coding or missing outcome data were excluded. PREDICTOR VARIABLE The primary predictor variable was preoperative albumin categorized as low (<3.4 g/dL), intermediate (3.4 to 3.9 g/dL), or high (>3.9 g/dL). OUTCOME VARIABLE The primary outcome variable was intensive care unit (ICU)-level complications scored using the Clavien-Dindo classification system. This is a tool used to grade surgical complications, with grade IV and V complications defined as requiring ICU-level care. COVARIATES Covariates were demographic (age, sex, body mass index), medical (smoking, functional status, weight loss), and perioperative (concurrent procedures, tumor location, reconstructive modality). ANALYSES Descriptive, bivariate, and multiple logistic regression with bootstrap resampling statistics were used to evaluate the association between albumin and adverse outcomes. A significance level of P ≤ .05 was significant. RESULTS: A total of 4,491 subjects met inclusion criteria and had a documented albumin. There were 435 subjects with low albumin levels, 1,305 with intermediate levels, and 2,751 with high levels. In bivariate analysis, low albumin levels were associated with an increased risk of ICU-level complications, any complication, extended length of stay, and adverse discharge disposition (all P ≤ .001), while high levels were protective (all P ≤ .001). In bootstrapped multivariate analysis using intermediate albumin as the reference group and adjusting for demographics, tumor location, and reconstructive modality among others, low albumin levels were an independent predictor of ICU-level complications (P = .008, odds ratio, 1.64; 95% confidence interval, 1.14 to 2.40), while high levels were protective (P = .014, odds ratio, 0.689; 95% confidence interval, 0.521 to 0.923). CONCLUSIONS: Preoperative serum albumin was an independent predictor of adverse outcomes following ablative head and neck cancer procedures.