Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Background: Glossectomy, in multiple forms, is a common procedure for managing both malignant and benign tongue lesions. Patient frailty is a critical factor influencing postoperative outcomes, especially in elderly patients. Accurately measuring frailty preoperatively can identify high-risk patients and aid in surgical decision-making. The modified frailty index (mFI-5) has been traditionally used for this purpose, however, the risk analysis index (RAI) has demonstrated better predictive performance in other surgical populations. This study compares the predictive utility of mFI-5 and RAI for adverse outcomes in patients undergoing glossectomy, focusing on those categorized as severely frail.
Methods: Patients undergoing glossectomy between 2010 and 2020 were selected from the NSQIP database using specific CPT codes. Frailty scores were calculated using the RAI and mFI-5 indices, and patients were grouped into prefrail (RAI: 21-30, mFI-5: 1), frail (RAI: 31-40, mFI-5: 2), and severely frail (RAI >40, mFI-5: 3-5) categories. Multivariate logistic regression analyses were used to evaluate associations between frailty and postoperative outcomes, including mortality, Clavien-Dindo (CD) grade complications, extended length of stay (eLOS), and non-home discharge. Predictive performance of the two indices was compared using receiver operating characteristic (ROC) curves.
Results: A total of 2949 patients with a mean age of 60.7 years were included. RAI demonstrated superior predictive capacity in several outcomes for severely frail patients, including mortality (RAI: OR 9.449, 95% CI: 1.732-51.545, p=0.009; mFI-5: OR 8.59, 95% CI: 0.977-75.558, p=0.053), CD-II complications (RAI: OR 3.134, 95% CI: 1.431-6.861, p=0.004; mFI-5: OR 2.919, 95% CI: 1.215-7.013, p=0.017), CD-IIIb complications (RAI: OR 2.455, 95% CI: 1.011-5.598, p=0.047; mFI-5: OR 1.94, 95% CI: 0.81-4.648, p=0.137), and eLOS (RAI: OR 3.483, 95% CI: 1.572-7.718, p=0.002; mFI-5: OR 1.754, 95% CI: 1.374-2.24, p<0.001). However, mFI-5 demonstrated a higher predictive value for CD-IV complications (RAI: OR 2.666, 95% CI: 0.964-7.372, p=0.059; mFI-5: OR 4.017, 95% CI: 1.664-9.702, p=0.002) and non-home discharge (RAI: OR 2.762, 95% CI: 0.885-8.624, p=0.08; mFI-5: OR 8.526, 95% CI: 3.848-18.895, p<0.001). While mFI-5 showed higher odds ratios for these outcomes, RAI had superior discrimination for mortality and several other key complications, making it a more reliable tool for predicting severe adverse events in the severely frail glossectomy cohort.
Conclusions: In glossectomy patients, both RAI and mFI-5 frailty indices are effective predictors of adverse outcomes, particularly in severely frail individuals. Although the mFI-5 showed stronger associations with certain outcomes, the RAI demonstrated better overall discrimination and may be a more reliable tool for preoperative risk stratification. These findings suggest that incorporating the RAI into preoperative assessment protocols can improve patient counseling and optimize surgical and post operative planning. Further research is warranted to validate these findings in other head and neck surgeries.