Validation of nutritional screening tools in patients undergoing cancer surgery in low- and middle-income countries

• DOI: 10.3389/fnut.2025.1576916
• Publication Date: 2025-06-05T04:10:30Z
• Authors: Jana Sremanakova; Stephen R. Knight; Marie Carmela M. ...; Edwin Yenli; Stephen Tabiri; Dhruva Ghosh; Pamela A. Kingsley; Apple Valparaiso; Ewen M. Harrison; Maria Thomas; Parvez D. Haque; Sreejith K. Veetil; Atul Suroy; Ashish Choudhrie; Rohin Mittal; Rajkumar Kottayas...; Bipradas Roy; Debra Jones; Sorrel T. Burden
• Keywords: Cancer surgery

Approximately one-third of patients are severely malnourished prior to surgery in low- and middle-income countries (LMICs). Identifying the most appropriate tool for detecting malnutrition is a critical first step toward enabling effective treatment interventions. Therefore, this study aimed assess validity reliability nutritional screening tools with cancer scheduled LMICs. Participants included adults undergoing either curative elective or palliative surgeries Ghana, India, Philippines. Nutritional status was assessed using anthropometric measurements, Malnutrition Universal Screening Tool (MUST), Patient-Generated Subjective Global Assessment (PG-SGA). Data were analysed Bland–Altman plots confidence intervals (CIs) intra-class correlation coefficients (ICCs) inter-rater reliability. Sensitivity specificity tests conducted Area Under Receiver Operating Characteristics Curve (AUROC). A total 167 participants recruited, mean age 53.3 years (SD 14.7) body mass index (BMI) 23.0 kg/m 2 4.9). The proportion identified as at risk 53.3% MUST, 47.3% PG-SGA SF, 66% full PG-SGA. When compared PG-SGA, MUST SF had AUROCs 0.78 (95% CI: 0.73–0.87) 0.76 0.68–0.83), respectively. demonstrated sensitivity 85% 25%, while showed 93% 42%. Excellent agreement observed ICC values >0.9 across all assessments. Both good when However, slightly greater than MUST. Based on these findings, recommended preoperative