Optimizing care of older patients with cancer in India: Comparative study of risk prediction of three assessment tools.

INTRODUCTION

Geriatric assessment is recommended for all older patients with cancer planned for treatment, but time and expertise required limit its use in routine oncology clinics. Screening tools help identify patients requiring complete geriatric assessment and also predict adverse outcomes. Our study aims to compare a culturally appropriate tool developed in India: SCreening for Older PErson with Cancer (SCOPE-C), Version 1, which helps triaging patients for treatment with other frequently used screening tools.

MATERIALS AND METHODS

This single centre study was conducted in treatment-naïve older patients with cancer (aged ≥60 years) at a tertiary care hospital in North India from November 2019 to May 2022. Frequently used screening tools like the Geriatric 8(G-8) and Vulnerable Elders Survey (VES-13), and SCOPE-C were administered on older patients with cancer. Agreement between the scales was analyzed using Cohen's Kappa statistics and mortality risk prediction using logistic regression analysis. Cut offs implying impaired performance for each scale were: G8 ≤ 14, VES 13 ≥ 3, and SCOPE C < 64. Further, G8- cutoff of <12 previously proposed for the Indian population was also explored.

RESULTS

Maximum agreement was observed between VES-13 and G-8 (73.13% [kappa = 0.081]), while SCOPE-C and VES-13 demonstrated a moderate level of concordance (62.19% [kappa = 0.321]). Agreement between SCOPE-C and G-8 was lower (43.28% [kappa = 0.056]). Propensity score matched (PSM) analysis revealed that an unfavorable score on SCOPE-C was significantly associated with higher mortality ratio (OR: 2.07 [95% CI: 1.07-4.03]), while VES-13 (OR: 1.31 [95% CI:0.57-3.04]),G8 cut off ≤14(OR: 0.78 [95% CI: 0.20-3.06]) and G8 cut off <12 (OR:1.49 [95%CI:0.62-3.63]) had no statistically significant association with mortality.

DISCUSSION

SCOPE-C, Version 1 is a validated tool for the Indian population that considers the low literacy rate and socio-cultural milieu of the country. Though perceptible agreement was not seen between the scales, SCOPE-C had better risk predictive potential for mortality. Hence, we recommend using SCOPE-C for prognostication and decision making in resource constrained settings like India and other low- and middle-income countries with similar background.