Financial toxicity in insured cancer patients: association with global health status in a prospective Brazilian Cohort

Abstract

Purpose

Financial toxicity is increasingly recognized as a clinically relevant patient-reported outcome in oncology, yet prospective data from Latin America remain limited, particularly among privately insured populations. We evaluated the prevalence and predictors of Q28-based financial burden (FB) and financial toxicity (FT), and their association with global health status (GHS), in Brazilian cancer patients treated in the supplementary healthcare sector.

Methods

This secondary analysis pooled data from four prospective observational cohorts of patients with breast, prostate, colorectal, and lung cancer treated at Américas Oncologia (Rio de Janeiro and São Paulo, Brazil). FB was assessed using item 28 (Q28) of the EORTC QLQ-C30 at baseline and 6 months and defined as any self-reported financial difficulty (scores 2–4). FT was operationalized pragmatically as new or worsening financial difficulty at 6 months compared with baseline. GHS (items 29–30) was linearly transformed to a 0–100 scale. Logistic regression and exploratory Cox proportional hazards models were used to evaluate factors associated with FB/FT and overall survival (OS).

Results

Among 1,343 patients, 23% reported FB at baseline and 25% at 6 months; 16% developed Q28-based FT. In multivariable models, younger age independently predicted both FB and FT, and baseline FB strongly predicted FB at 6 months (OR 5.86; 95% CI 4.31–8.01). FB and FT were independently associated with lower GHS at 6 months. FT was not independently associated with OS in multivariable analysis (adjusted HR 0.97; 95% CI 0.54–1.74), whereas higher GHS at 6 months was independently associated with improved OS.

Conclusion

Q28-based FB and FT are common even among privately insured cancer patients in Brazil and are associated with worse GHS. Routine screening for financial distress may help identify vulnerable patients and inform supportive care strategies, particularly for younger individuals.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00520-026-11014-0.

Introduction

According to the World Health Organization, monitoring health inequalities is essential for developing evidence-based policies to reduce disparities and promote health equity [1]. In oncology, the financial consequences of treatment have emerged as an important dimension of patient-centered outcomes [2]. In this context, understanding the financial impact of cancer on patients' lives is crucial, particularly in low- and middle-income countries (LMIC), where the burden of out-of-pocket costs can undermine treatment access. Financial toxicity (FT) refers to the detrimental effects resulting from excessive financial burden (FB) related to diagnosis and treatment for patients and their families [2–5].

FT encompasses direct medical costs (e.g., diagnostic tests, medications, hospitalizations), direct non-medical costs (e.g., transportation, dietary needs, caregiver support), and indirect costs, including loss of income due to reduced work capacity of patients or caregivers [3–5]. In oncology practice, FT is increasingly recognized as a treatment-related burden associated with psychological distress, reduced treatment adherence, and potentially worse survival outcomes [5–12].

Brazil, a middle-income country marked by significant social and economic inequalities, presents a relevant setting to examine FT. The Brazilian healthcare system is highly segmented, comprising a public universal system (Sistema Único de Saúde – SUS) and a private sector accessible primarily through employer-sponsored or out-of-pocket insurance plans [13, 14]. Although private insurance may mitigate some direct medical expenses, patients may still experience indirect and productivity-related costs during active cancer treatment. Despite this dual structure, data on FT remain scarce [15–17], particularly in the private, insurance-based sector.

Several tools have been developed or recommended to assess FT, including the COmprehensive Score for Financial Toxicity (COST) questionnaire [18, 19], and question 28 (Q28) of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire—Core 30 (EORTC QLQ-C30) [12, 20]. Q28 provides a brief, scalable measure that can be incorporated into routine quality-of-life assessment in clinical oncology.

Prospective longitudinal data evaluating FT and its association with global health status (GHS) in Latin America are limited, particularly among privately insured populations. In this analysis, we investigated the prevalence and predictors of FB and FT during treatment and examined their association with GHS and overall survival among cancer patients treated in a private oncology facility in Brazil.

Methods

Study design

This study is a secondary analysis of four prospective observational cohort studies of patients with breast (BC), prostate (PC), colorectal (CRC), and lung cancer (LC) treated at Américas Oncologia, a private oncology network in Rio de Janeiro and São Paulo, Brazil. All cohorts used standardized prospective patient-reported outcome (PRO) collection procedures. The enrollment period varied across cohorts: BC, August 2020 to December 2024; PC, June 2015 to December 2024; CRC, March 2023 to December 2024; and LC, March 2015 to December 2024. While EORTC QLQ-C30 questionnaires were collected prospectively, clinical and sociodemographic information was retrospectively extracted from medical records. Therefore, some degree of missing data is expected, reflecting the inherent limitations of retrospective data acquisition.

Although the cohorts differed in tumor type, prognosis, recruitment period, and availability of selected clinical and sociodemographic variables, they were pooled to evaluate financial distress across a real-world insured oncology population. Cancer type was included as an adjustment variable in the main multivariable models whenever appropriate.

For the 6-month assessment, the questionnaire closest to the target time point within a 4.5- to 7.5-month window was selected. When duplicate questionnaires were identified on the same date with discrepant responses, the response indicating the worse financial outcome was retained.

Baseline and 6-month EORTC QLQ-C30 databases were screened for unique patient identifiers in each tumor-specific cohort. The number of unique baseline and post-baseline EORTC QLQ-C30 records screened and the final paired analytic sample differed by tumor cohort: CRC, 119 baseline and 74 post-baseline records screened, with 56 patients included; BC, 596 and 440 records screened, with 387 included; PC, 899 and 806 records screened, with 638 included; and LC, 451 and 354 records screened, with 262 included. The final analytic dataset included 1,343 observations after linkage of baseline and 6-month Q28 responses with available demographic and clinical variables from the respective tumor-specific cohorts. (Supplementary Fig. 1).

Outcomes and definitions

FB and FT were assessed using Q28 of the EORTC QLQ-C30 [21]. Patients completed the questionnaire at baseline, prior to treatment initiation in the corresponding cohort, and at approximately 6 months after treatment start. Q28 asks whether the patient’s physical condition or medical treatment caused financial difficulties, with responses ranging from 1 (“not at all”) to 4 (“very much”).

FB was defined as any reported financial difficulty (scores 2–4) at each time point, with score 1 (“not at all”) used as the reference category. This dichotomization was chosen to identify any degree of self-reported financial difficulty in a screening-oriented framework. FT was defined as newly reported or worsening financial difficulty at 6 months compared with baseline. These operational definitions were adapted from Perrone et al [12] and have been used in subsequent prospective analyses evaluating financial hardship using PROs [20, 22, 23]. Both outcomes were analyzed as binary variables.

GHS was calculated from items 29 and 30 of the EORTC QLQ-C30 [21] and linearly transformed to a 0–100 scale according to EORTC scoring guidelines, with higher scores indicating better overall quality of life.

Statistical analysis

Associations between FB/FT and clinical or epidemiological characteristics were evaluated using univariable and multivariable logistic regression models. Available baseline variables included age, sex, race/ethnicity, body mass index (BMI), smoking status, marital status, comorbidities, metastatic disease, cancer type, baseline FB, and GHS. Variable availability differed across tumor-specific cohorts. Variables with p < 0.05 in univariable analyses and/or considered clinically relevant were evaluated for inclusion in multivariable models.

To preserve sample size and account for tumor-type confounding, final multivariable models included cancer type whenever appropriate. Variables with substantial cohort-specific missingness or tumor-specific distribution were not retained in final models. The main FB model included age, grouped ethnicity, cancer type, and baseline FB. The main FT model included age, grouped ethnicity, and cancer type. Odds ratios (ORs), 95% confidence intervals (CIs), and p-values were reported.

The association between FB/FT and GHS at 6 months was assessed using logistic regression models adjusted for variables significant in univariable analyses.

Overall survival (OS) was defined as the time from baseline questionnaire completion to death from any cause or last follow-up. To minimize immortal-time bias, OS analyses were conducted using a pre-specified 6-month landmark approach, including only patients alive and evaluable at 6 months. Kaplan–Meier curves were generated and compared using the log-rank test. Cox proportional hazards models were used to estimate hazard ratios (HRs), 95% CIs, and p-values. Survival analyses were considered exploratory because of limited follow-up, uneven event distribution across tumor types, and low event rates in some cohorts.

The proportional hazards assumption was assessed using Schoenfeld residuals, and no significant violations were observed.

No imputation procedures were applied. Analyses were conducted using complete-case datasets. Patients with missing QLQ-C30 data at 6 months were excluded from 6-month FB/FT analyses but remained eligible for baseline assessments. The final sample size for each model is reported in the Results section and supplementary tables. Complete-case analysis may have introduced selection bias, particularly for variables not collected across all cohorts or with substantial missingness.

No formal sample size calculation was performed, as this study represents a secondary analysis of prospectively collected cohorts.

All statistical tests were two-sided, and p < 0.05 was considered statistically significant. Analyses were performed using R software, version 4.4.1. This study is reported in accordance with the STROBE guidelines.

Ethical considerations

The study protocol was reviewed and approved by the local Institutional Review Board (IRB; reference number 12747119.5.0000.5533) and was conducted in accordance with the ethical standards of the institutional research committee and with the 1964 Declaration of Helsinki and its later amendments. All participants provided written and oral informed consent. The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Results

A total of 1,343 patients were included: 56 with CRC, 387 with BC, 638 with PC, and 262 with LC. The mean age was 62.2 years (SD 12.5). Most patients were male (60.1%). Regarding race, 68.5% were White and 22.7% Black/Brown, with 10.9% missing race data. Only 12.2% had metastatic disease at baseline (overall missing 18.2%). Missingness varied across variables because selected sociodemographic and clinical variables were not uniformly collected in all tumor-specific cohorts. Baseline characteristics are presented in Table 1.

Table 1.

Baseline characteristics overall and by cancer type

Variable	Total (N = 1,343)	CRC (N = 56)	Breast (N = 387)	Prostate (N = 638)	Lung (N = 262)
Ethnicity
White	916 (68.5%)	36 (69.2%)	273 (78.7%)	424 (74.8%)	183 (79.6%)
Black or Brown	272 (22.7%)	15 (28.8%)	71 (20.5%)	141 (24.9%)	45 (19.6%)
Asian	7 (0.6%)	1 (1.8%)	3 (0.8%)	1 (0.2%)	2 (0.8%)
Indigenous	1 (< 0.1%)	0 (0%)	0 (0%)	1 (0.2%)	0 (0%)
Missing	147 (10.9%)	4 (7.1%)	40 (10.3%)	71 (11.1%)	32 (12.2%)
Age, mean (SD)	62.2 (12.5)	57.1 (13.2)	53.4 (12.5)	66.6 (9.2)	66.6 (11.2)
Sex
Female	535 (39.9%)	23 (41.1%)	386 (99.7%)	0 (0%)	126 (48.1%)
Male	807 (60.1%)	33 (58.9%)	0 (0%)	638 (100%)	136 (51.9%)
Missing	1 (0.1%)	0 (0%)	1 (0.3%)	0 (0%)	0 (0%)
BMI, mean (SD)	27.7 (5.2)	27.1 (5.4)	28.0 (5.7)	27.4 (4.2)	NA
Missing	393 (29.3%)	1 (1.8%)	20 (5.2%)	372 (58.3%)	NA
NA	262 (19.5%)	0 (0%)	0 (0%)	0 (0%)	262 (100%)
Marital status
Married/stable relationship	270 (20.1%)	39 (69.6%)	NA	231 (36.2%)	NA
Divorced/separated	22 (1.6%)	6 (10.7%)	NA	16 (2.5%)	NA
Single	18 (1.3%)	9 (16.1%)	NA	9 (1.4%)	NA
Widowed	11 (0.8%)	1 (1.8%)	NA	10 (1.6%)	NA
Missing	35 (2.6%)	1 (1.8%)	NA	34 (5.3%)	NA
NA	987 (73.5%)	0 (0%)	387 (100%)	0 (0%)	262 (100%)
Smoking status
No	374 (27.8%)	36 (64.3%)	295 (76.2%)	NA	43 (16.4%)
Yes	228 (17.0%)	20 (35.7%)	90 (23.3%)	NA	118 (45.0%)
Missing	103 (7.7%)	0 (0%)	2 (0.5%)	NA	101 (38.5%)
NA	638 (47.5%)	0 (0%)	0 (0%)	638 (100%)	0 (0%)
Comorbidities
No	360 (26.8%)	NA	177 (45.7%)	136 (21.3%)	47 (17.9%)
Yes	917 (68.3%)	NA	203 (52.5%)	499 (78.2%)	215 (82.1%)
Missing	66 (4.9%)	NA	7 (1.8%)	3 (0.5%)	0 (0%)
NA	56 (4.2%)	56 (100%)	0 (0%)	0 (0%)	0 (0%)
Metastatic disease
No	935 (69.6%)	37 (75.5%)	360 (94.7%)	469 (92.1%)	69 (42.9%)
Yes	164 (12.2%)	12 (24.5%)	20 (5.3%)	40 (7.9%)	92 (57.1%)
Missing	244 (18.2%)	7 (12.5%)	7 (1.8%)	129 (20.2%)	101 (38.5%)

Percentages are based on the total number of patients within each tumor subtype. Patients were classified according to the Brazilian Institute of Geography and Statistics (IBGE) racial categories. Missing = variable collected, but data not available for some patients. NA = variable not collected for that specific tumor subtype. CRC  colorectal cancer, BMI   body mass index, SD standard deviation

Baseline characteristics stratified by baseline FB status are shown in Supplementary Table 1. Patients with baseline FB were younger, more frequently female, and more commonly represented in the BC and LC cohorts than in the PC cohort.

The full distribution of Q28 response categories is shown for the overall population and by tumor group in Fig. 1, whereas the regression analyses used the prespecified binary definitions described above.

Fig. 1.

Distribution of responses to item 28 at baseline and after six months, for the total sample and stratified by cancer type. Responses to item Q28 from the EORTC QLQ-C30 questionnaire — “Has your physical condition or medical treatment caused you financial difficulties?” — are shown at two timepoints: baseline and 6-month follow-up

At baseline, approximately 23% of patients reported FB. Baseline FB was observed in 28% of patients with BC and LC, 21% of those with CRC, and 19% with PC. At 6 months, 1,342 patients were evaluable for FB and FT after exclusion of one patient who died before follow-up. In the cohort, 25% of patients reported FB, and 16% developed Q28-based FT.

In univariable logistic regression, FB at 6 months was significantly associated with Black or Brown ethnicity (OR 1.42; 95% CI 1.04–1.91; p = 0.024), younger age (OR 0.96; 95% CI 0.95–0.97; p < 0.001), cancer type, and male sex. Compared with PC, both BC (OR 2.05; 95% CI 1.54–2.73; p < 0.001) and LC (OR 1.44; 95% CI 1.02–2.01; p = 0.036) were associated with higher odds of FB. Male sex was associated with lower odds of FB (OR 0.55; 95% CI 0.43–0.70; p < 0.001). In the final multivariable model, adjusted for age, grouped ethnicity, cancer type, and baseline FB, only age (OR 0.97; 95% CI 0.96–0.98; p < 0.001) and baseline FB (OR 5.86; 95% CI 4.31–8.01; p < 0.001) remained independently associated with FB at 6 months (Table 2 and Supplementary Tables 2 and 3).

Table 2.

Univariable and multivariable associations with financial burden and financial toxicity

	FB – Univariable OR (95% CI)	p-value	FB – Multivariable OR (95% CI)	p-value	FT – Univariable OR (95% CI)	p-value	FT – Multivariable OR (95% CI)	p-value
Ethnicity: White (ref)	—	—	—	—	—	—	—	—
Ethnicity: Black/Brown	1.42 (1.04–1.91)	0.024	1.29 (0.90–1.84)	0.166	1.39 (0.97–1.97)	0.068	1.36 (0.92–1.99)	0.120
Ethnicity: Other*	2.55 (0.50–11.7)	0.223	1.16 (0.74–1.81)	0.511	4.41 (0.86–20.2)	0.054	1.25 (0.76–2.00)	0.357
Age (per year)	0.96 (0.95–0.97)	< 0.001	0.97 (0.96–0.98)	< 0.001	0.97 (0.96–0.98)	< 0.001	0.98 (0.96–0.99)	< 0.001
Male gender	0.55 (0.43–0.70)	< 0.001	—	—	0.63 (0.47–0.85)	0.002	—	—
Comorbidities	0.81 (0.61–1.06)	0.124	—	—	0.71 (0.52–0.98)	0.034	—	—
Cancer type: Prostate (ref)	—	—	—	—	—	—	—	—
Cancer type: Colorectal	1.64 (0.87–2.97)	0.114	1.11 (0.51–2.28)	0.787	1.28 (0.57–2.59)	0.521	1.01 (0.41–2.19)	0.986
Cancer type: Breast	2.05 (1.54–2.73)	< 0.001	1.18 (0.80–1.75)	0.409	1.71 (1.22–2.40)	0.002	1.28 (0.83–1.97)	0.258
Cancer type: Lung	1.44 (1.02–2.01)	0.036	1.23 (0.83–1.81)	0.303	1.42 (0.95–2.10)	0.080	1.50 (0.98–2.28)	0.057
Baseline FB positive	6.29 (4.77–8.32)	< 0.001	5.86 (4.31–8.01)	< 0.001	—	—	—	—

Reference category for ethnicity = White; reference category for cancer type = Prostate. OR  Odds Ratio, CI  Confidence Interval. *For univariable analyses, the “Other” row displays the Asian category; the Indigenous category was not estimable because of very small numbers. For multivariable analyses, Asian and Indigenous patients were grouped as Other. Sex, smoking status, and comorbidities were evaluated in univariable analyses but were not retained in the final multivariable models because of tumor-specific distribution, cohort-specific missingness, and model parsimony. Patients were classified according to the Brazilian Institute of Geography and Statistics (IBGE) racial categories

Regarding FT, univariable analysis showed significant associations with younger age (OR 0.97; 95% CI 0.96–0.98; p < 0.001), male sex (OR 0.63; 95% CI 0.47–0.85; p = 0.002), presence of comorbidities (OR 0.71; 95% CI 0.52–0.98; p = 0.034), and BC compared with PC (OR 1.71; 95% CI 1.22–2.40; p = 0.002). In the final multivariable model, adjusted for age, grouped ethnicity, and cancer type, only younger age remained significantly associated with FT (OR 0.98; 95% CI 0.96–0.99; p < 0.001) (Table 2 and Supplementary Tables 4 and 5).

GHS was lower among patients reporting FB or FT. At six months, mean GHS was 70.9 among patients with FB compared with 81.1 among those without FB. Among patients with FT, mean GHS was 71.7 compared with 79.9 among those without FT. In the multivariable logistic models, each 1-point increase in GHS at six months was associated with a 2% reduction in the odds of experiencing FB (OR 0.98; 95% CI 0.97–0.99; p = 0.004) and FT (OR 0.98; 95% CI 0.97–0.99; p < 0.001) (Table 3, Supplementary Table 6 and Supplementary Fig. 1).

Table 3.

Association between global health status (GHS) and financial burden (FB)/financial toxicity (FT)

Variable	Model	OR	95% CI	p-value
GHS (baseline)	Association with FB (6 m)	1.00	0.98–1.01	0.365
	Association with FT (6 m)	1.00	0.99–1.01	0.576
GHS (6 months)	Association with FB (6 m)	0.98	0.97–0.99	0.004
	Association with FT (6 m)	0.98	0.97–0.99	< 0.001

OR  Odds Ratio, CI Confidence Interval, GHS Global Health Status, FB Financial Burden, FT Financial Toxicity. Multivariable logistic regression models evaluating the association between GHS and FB/FT were adjusted for age, cancer type, grouped ethnicity, and baseline FB when applicable. GHS was modeled as a continuous variable on a 0–100 scale

At the six-month landmark, 1,342 patients were included in exploratory survival analyses. In Kaplan–Meier analyses, patients reporting FT had numerically lower survival, although the difference did not reach statistical significance (HR 1.36; 95% CI 0.95–1.93; p = 0.089). Median OS was not reached among patients without FT, whereas those reporting FT had an estimated median OS of 99.5 months (Fig. 2).

Fig. 2.

Overall survival by presence of financial toxicity at six months. Overall survival by presence of Q28-based financial toxicity at 6 months. Kaplan–Meier curves are shown according to the presence or absence of FT at the 6-month landmark

At a median follow-up of 22.6 months (11.7 months for CRC, 12.3 months for BC, 27.3 months for PC, and 17.6 months for LC), median overall survival was 99.5 months in the group reporting financial toxicity versus not reached in the group without financial toxicity, with an HR of 1.36 (95% CI 0.95–1.93; p = 0.089). Event rates were 3.6%, 4.7%, 11%, and 45% for CRC, BC, PC, and LC, respectively.

In the multivariable Cox model adjusted for cancer type, neither FB at six months nor FT was independently associated with OS. The adjusted HRs were 1.00 (95% CI 0.60–1.68; p = 0.993) for FB and 0.97 (95% CI 0.54–1.74; p = 0.923) for FT (Supplementary Table 7).

GHS at 6 months was independently associated with OS: for every 10-point increase in GHS, the risk of death decreased by 10% (HR 0.90; 95% CI 0.84–0.97; p = 0.007), while baseline GHS was not associated with survival (HR 1.01; 95% CI 0.94–1.08; p = 0.826) (Supplementary Table 8).

Discussion

This study demonstrates that self-reported financial difficulty is common and clinically relevant among patients receiving cancer care within Brazil’s private health system. Even in an insured population, one in four patients reported FB and, using a Q28-based longitudinal definition, one in six developed worsening or new financial difficulty during treatment. While expanded health coverage in another LMIC settings has been associated with reductions in FT and improvements in quality of life [24], our results underscore that formal insurance coverage may not fully mitigate the broader socioeconomic dimensions of financial hardship. Importantly, they suggest that even within comparatively privileged healthcare environments, underlying social and structural inequities may shape the distribution and intensity of financial consequences during cancer treatment. In our analysis, younger patients were particularly vulnerable to both FB and FT, consistent with prior reports showing that younger individuals experience greater income loss, limited financial reserves, and employment disruption following a cancer diagnosis. [5, 6, 20]

FB and FT were associated with lower GHS at 6 months, suggesting that financial difficulty may coexist with broader impairments in perceived health and quality of life. Six-month GHS was also independently associated with OS, supporting the clinical relevance of this global patient-reported measure. However, GHS may capture multiple dimensions of vulnerability, including functional decline, treatment-related toxicity, symptom burden, and socioeconomic stressors. The relationship between financial hardship and GHS is likely multifactorial and potentially bidirectional. Because FB, FT, and GHS were measured using the same EORTC QLQ-C30 instrument, conceptual and measurement overlap should be considered when interpreting these associations. Given the observational design and exploratory nature of the survival analyses, these findings should not be interpreted as establishing a causal relationship between GHS, financial distress, and OS. Overall, our findings are consistent with previous literature linking FT to impaired quality of life. [7, 10, 12, 25]

Although FT has been associated with inferior oncologic outcomes in other cohorts [7, 12], our analysis did not demonstrate a statistically significant independent association between FB or FT and OS. This null adjusted finding should be interpreted cautiously. In unadjusted analysis, FT was associated with numerically worse OS, but the association did not reach statistical significance. The absence of an independent association may reflect measurement limitations of a single Q28-based proxy, limited follow-up, low event rates in BC and CRC, heterogeneity in prognosis across tumor types, and residual confounding.

Methodologically, our approach aligns with Perrone et al., who operationalized FB and FT using item Q28 of the EORTC QLQ-C30 [12]. Subsequent studies have also used Q28-based definitions to evaluate FB and FT in different oncology contexts [22, 23]. The 2024 ESMO Expert Consensus on FT supports the use of structured screening for financial distress in oncology and includes Q28 among the tools that may be used for this purpose [20]. These data support the use of Q28 as a pragmatic, screening-oriented measure that can be incorporated into prospective PRO collection, including in a middle-income country context. However, multidimensional instruments such as COST and PROFFIT provide a more comprehensive assessment of FT [19, 26], including material, behavioral, and psychosocial dimensions of financial hardship. Therefore, Q28 should not be interpreted as a substitute for dedicated FT instruments.

Recent evidence from LMICs and broader reviews of FT measurement further contextualize our findings [2, 27]. Literature focused on LMICs has reported a high burden of objective FT among patients with cancer, although with substantial heterogeneity across settings, populations, definitions, and measurement approaches [2]. Many LMIC studies have relied on objective cost-related measures, such as catastrophic health expenditure or household impoverishment, while subjective dimensions of FT, including material, behavioral, and psychosocial consequences, have been less frequently assessed [2]. In addition, a broader scoping review of FT assessment in oncology has shown that most studies have been conducted in high-income countries, with limited representation from LMIC settings [28]. In this context, our study provides prospective real-world evidence from Brazil, a large middle-income country with a complex dual healthcare system, contributing to a more geographically diverse understanding of FT in oncology.

These data support the need for routine assessment and screening of FT, particularly among working-age patients who may face substantial indirect costs and income instability during treatment. Our study did not evaluate the causes of financial hardship or the effectiveness of specific interventions. Future studies in the Brazilian supplementary health sector should assess culturally appropriate strategies that are feasible in routine practice and responsive to the financial, logistical, and social challenges that may vary across care settings. Evidence from intervention studies suggests that approaches such as financial navigation, counseling, assistance programs, and cost-conversation tools may be considered, but their applicability and impact in this context require prospective evaluation [20, 28].

Several limitations warrant consideration. This was a secondary analysis of multiple prospective cohorts, introducing potential heterogeneity in inclusion periods, treatment contexts, tumor prognosis, follow-up schedules, and variable availability. The long recruitment period may have introduced temporal heterogeneity related to changes in treatment patterns, drug access, costs, insurance coverage, and supportive care practices. Although all cohorts were conducted within the same private oncology network and used standardized PRO collection, formal heterogeneity testing across tumor types was not performed. Analyses of six-month FB and FT were restricted to patients with available follow-up data. Patients with missing PRO data at 6 months were not included in these analyses, which may have influenced the estimates of FT.

Although Q28 is a pragmatic screening measure, it captures a single dimension of financial hardship and does not provide the multidimensional assessment offered by instruments such as COST or PROFFIT. Therefore, FT in this manuscript should be interpreted as a longitudinal approximation of worsening self-reported financial difficulty rather than a comprehensive measure of FT. This approach may have led to misclassification and should be considered when interpreting the absence of an independent association between FT and OS. Future studies incorporating multidimensional measures may allow a more comprehensive characterization of FT and its determinants in this setting.

Socioeconomic variables such as income, education, employment status, insurance plan type, household composition, direct out-of-pocket expenses, indirect costs, and caregiver burden were not available, limiting our ability to characterize the social determinants and specific mechanisms of FT more precisely. In addition, selected sociodemographic and clinical variables were collected only in specific cohorts, resulting in missing data for race/ethnicity, BMI, smoking status, marital status, and metastatic disease. Detailed treatment modality and disease stage were also not uniformly available. These factors should be considered when interpreting subgroup analyses and regression models, and residual confounding cannot be excluded.

Generalizability is another important consideration. All patients were treated in private oncology centers located in the Southeast region of Brazil, the country’s most economically developed area [13]. Privately insured patients in Brazil likely represent a socioeconomically advantaged subgroup compared with the overall cancer population. Therefore, these findings may underestimate the magnitude of financial hardship experienced by patients treated within the public healthcare system (Sistema Único de Saúde—SUS) or in less affluent regions. Nonetheless, the persistence of FB and FT in this comparatively privileged context underscores the magnitude of financial hardship in oncology and suggests that private insurance coverage does not fully mitigate indirect and non-medical costs, such as transportation, caregiving, and income loss. As FT may be underestimated across different clinical settings, social support interventions, including the use of screening tools such as Q28 of PRO assessments, should be considered to better identify vulnerable patients and provide support aimed at improving their quality of life.

To our knowledge, this represents the first prospective longitudinal evaluation of Q28-based FT across multiple tumor types in Brazil. Our findings indicate that financial distress remains relevant even among privately insured patients and support incorporating FT screening into routine oncology care.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Natalia Cristina Cardoso Nunes and Mariana Ribeiro Monteiro contributed to the study conception and design. Material preparation and data collection were performed by Natalia Cristina Cardoso Nunes, Mariana Ribeiro Monteiro, Giselle de Souza Carvalho, Juliana Pompeu Pecoraro, Thamirez de Almeida Vieira Ferreira, Paola Kelly Martins dos Santos, Debora Cristina Victorino Azevedo, and Ana Paula Victorino. Statistical analysis was performed by Natalia Cristina Cardoso Nunes and Mariana Ribeiro Monteiro. Luiz Henrique de Lima Araujo contributed to critical revision of the manuscript and provided overall methodological supervision. The first draft of the manuscript was written by Natalia Cristina Cardoso Nunes, and all authors commented on previous versions. All authors read and approved the final manuscript.

Funding

The authors declare that no specific funding was received for this study.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval

The study protocol was reviewed and approved by the local Institutional Review Board (IRB; reference number 12747119.5.0000.5533) and was conducted in accordance with the ethical standards of the institutional research committee and with the 1964 Declaration of Helsinki and its later amendments.

Consent to participate

All participants provided written and oral informed consent prior to study inclusion.

Consent to publish

The manuscript does not contain any individual person’s identifiable data. All participants consented to the use of anonymized data for research purposes.

Competing interests

The authors declare no competing interests.