A predictive model for swallowing dysfunction after oral cancer resection.

Area of Science:

• Oncology
• Surgical Outcomes
• Health Services Research

Background:

• Aggressive oral cancer therapy often leads to significant postoperative morbidity, including feeding difficulties.
• Current reactive referrals for feeding issues increase patient length of stay (LOS) and morbidity.
• A proactive approach is needed to optimize patient flow and outcomes.

Purpose of the Study:

• To develop a predictive model for identifying patients likely to experience postoperative feeding issues.
• To facilitate a preoperative referral pathway for nutritional support.
• To improve patient flow and reduce length of stay.

Main Methods:

• Analysis of 203 oral cancer patients referred to a Head and Neck multidisciplinary team meeting over five years.
• Utilized linear regression to assess the impact of feeding issues multidisciplinary team meeting (FiMDT) referral on LOS.
• Employed logistic regression to predict FiMDT referral based on preoperative factors.

Main Results:

• Inpatient FiMDT referral was an independent predictor of increased LOS.
• Key preoperative predictors for FiMDT referral included tracheostomy, patient age, and alcohol intake.
• The developed model demonstrated 90% sensitivity and 93.8% specificity.

Conclusions:

• Inpatient feeding issue multidisciplinary team meeting referral significantly impacts patient length of stay.
• A robust predictive model for preoperative referral can be established using specific patient factors.
• Shifting to a preoperative referral pathway can enhance patient flow and potentially reduce morbidity.