Self-reported financial difficulties of colorectal cancer patients 1 year after start of treatment.

Background: The purpose of this study is to investigate self-reported financial difficulties (FDs) in patients with colorectal cancer (CRC) and to develop easy-to-implement models to predict FDs 1 year after initiation of CRC treatment.

Methods: In the prospective EDIUM cohort, CRC patients are asked to complete the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 30 (QLQ-C30) and QLQ-CR29 before (T0) and 1 year after initiation of treatment (T1). These patient-reported outcomes are combined with sociodemographic and clinical data. The QLQ-C30 'FD' item is firstly analysed descriptively, together with covariates. Logistic lasso regression was carried out to develop a predictive model (full model) of FDs at T1, and its predictive capacity was assessed. This predictive model was then simplified using easy-to-collect predictors which showed importance in the full model.

Results: Of 9124 CRC patients enrolled between October 2018 and May 2023, 6061 completed follow-up questionnaires (T1). Of those initially without FDs (n = 4989), 25% (1248 patients) reported new FDs at T1. These patients were younger [median age 66 versus 72 years; odds ratio (OR) 2.81 (95% confidence interval [CI] 2.46-3.22)] and had a higher proportion of lower educational attainment [74% versus 69%; OR 1.32 (95% CI 1.14-1.54)]. Clinically, FDs were more frequently reported by patients with higher International Union Against Cancer (UICC) stages [48% in stages III/IV versus 32.5% in stages I/II; OR 1.98 (95% CI 1.74-2.26)] and rectal cancer [37% versus 28%; OR 1.52 (95% CI 1.32-1.73)]. The full model [based on a test and training dataset of n = 5910 patients, including 14 predictors, area under the curve (AUC) 0.75] was reduced to two simplified models with only the predictors age, sex (model 1, AUC 0.72)/localisation of tumour (model 2), and FDs at T0 (AUC = 0.74). On this basis, two risk assessment tables were developed.

Conclusions: A significant proportion of CRC patients report FDs 1 year after treatment initiation, particularly among socioeconomically disadvantaged groups. Together with the predictive models and two ready-to-use risk assessment tables, these findings can be utilised by clinicians and psychosocial counselling services to provide tailored support options.