Frailty Predicts Postoperative Complications and Length of Stay in Oral Squamous Cell Carcinoma: A 250-Patient Retrospective Cohort Study Using the Modified Frailty Index

Frailty reflects diminished physiologic reserve and has been linked to adverse surgical outcomes, but data in oral squamous cell carcinoma (OSCC) are limited.

We conducted a retrospective cohort study of 250 consecutive patients with OSCC who underwent curative-intent surgery at a tertiary cancer centre in India (2021–2024). Pre-operative frailty was assessed using the 11-item Modified Frailty Index (mFI-11), and patients were classified as robust, pre-frail or frail. The primary outcome was 30-day postoperative complications graded by the Clavien–Dindo system. Secondary outcomes included length of hospital stay (LOS), adjuvant therapy utilisation and 6-month mortality. Multivariable logistic and linear regression models adjusted for age, stage, reconstruction and operative factors were used.

Of 250 patients (mean age 52 years; 67% male), 17% were robust, 36% pre-frail and 47% frail. Complications occurred in 36% overall, rising stepwise from 10% in robust to 25% in pre-frail and 56% in frail patients (P < 0.001). Major complications (Clavien–Dindo ≥ IIIa) followed an even steeper gradient (0% in robust, 6.6% in pre-frail and 22.2% in frail; P < 0.001). On adjusted analysis, frailty independently predicted complications (odds ratio 8.1; 95% CI 3.2–20.3). Median LOS rose from 9 days in robust to 11 in pre-frail and 14 in frail patients (P < 0.001); frailty remained an independent determinant of prolonged stay. Adjuvant therapy completion rates were similar across groups, though interruptions were more common among frail patients. Six-month mortality was 2.3% in robust, 4.4% in pre-frail and 6.8% in frail patients (log-rank P = 0.21).

Frailty assessed with the mFI-11 is a strong independent predictor of postoperative complications and hospital stay in OSCC surgery, supporting routine frailty screening to guide perioperative planning in resource-constrained, high-volume oncology settings.