A Novel Nomogram for Early Identification and Intervention in Colorectal Cancer Patients at Risk for Malnutrition

Abstract

Background

Malnutrition is under-recognized in cancer patients and can lead to poor treatment outcomes. We aim to develop an outpatient-focused score based on the Malnutrition Screening Tool (MST) to help identify colorectal cancer (CRC) profiles at high risk for malnutrition.

Methods

506 CRC patients during initial outpatient oncology consultation at our tertiary referral outpatient oncology clinic completed the MST. Objective and subjective data were collected through chart review. Data gathered are as follows: demographics, anthropometrics, laboratory values, patient-reported symptoms, MST score, cancer history, performance status, socioeconomic status, and Charlson Comorbidity. Predictors of malnutrition were identified by logistic regression. Receiver operating curve (ROC), area under the curve (AUC), and our model’s predictability were determined.

Results

Significant predictors of malnutrition are as follows: younger age (20-39 vs >40 years) (P = .007), normal-to-low body mass index at presentation (P = .019), Eastern Cooperative Oncology Group classification 2-3 (P = .012), metastatic disease (P = .046), albumin <3.0 g/dL (P = .033), fatigue (P < .001), and change in stool/bowel habits (P = .002). In our derived malnutrition score, risk of malnutrition increased from 11% for score 0, to 100% for scores 9-10. Receiver operating curve showed AUC .745 (95% CI, .697-.793).

Discussion

An outpatient clinic-derived malnutrition score obtained from objective and patient-reported variables may facilitate identification of CRC patients at highest risk for malnutrition. Rapid identification and intervention in high-risk patients may improve treatment recovery, therapy tolerance, and quality of life. Our tool requires external validation before application in clinical practice.