Lung cancer treatment pathways in the largest private health insurance in Brazil: a real-world data study.

Introduction
Lung cancer remains the leading cause of cancer-related mortality worldwide, accounting for more than 1.8 million deaths annually [1]. In Brazil, it is the most lethal malignancy, with an estimated 30,200 new cases each year and a 5-year survival rate of only 18% [2, 3]. Despite advances in diagnosis and therapy, late-stage presentation and limited access to precision oncology remain significant barriers [4].
Real-world evidence (RWE) provides insights into how patients are diagnosed and treated outside the controlled settings of clinical trials, and can guide both clinical practice and health policy [5–7]. In Brazil, previous efforts to characterize lung cancer care have been conducted. Mathias et al. described national epidemiology and treatment trends, while Baldotto et al. [8] analyzed treatment patterns and costs in advanced non–small cell lung cancer within private institutions [3, 9]. However, there is still a lack of large-scale analyses focusing specifically on the private supplementary health sector, which covers approximately 25% of Brazilians and plays a central role in cancer care delivery.
This study aims to describe patient characteristics, diagnostic pathways, treatment patterns, survival outcomes, healthcare utilization, and direct medical costs among lung cancer patients treated in the largest private health insurance system in Brazil. By focusing on the supplementary sector, our findings provide complementary evidence to existing national reports and highlight opportunities for improving access, efficiency, and outcomes in real-world oncology care.

Materials and methods

Study design and setting
We conducted a retrospective cohort study using data from the Hapvida NotreDame Intermédica system, Brazil’s largest private health insurance network, which provides coverage to more than 8.8 million individuals across all five geographic regions. The study protocol was approved by the National Research Ethics Commission (CONEP), National Health Council, Ministry of Health, Brazil (CAAE: 82712024.1.0000.0229). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and the Brazilian National Health Council Resolution No. 466/2012. Given the retrospective design and the use of anonymized secondary data, the requirement for informed consent was waived by the ethics committee.
In parallel, the Hapvida NotreDame Intermédica system provides coverage across all five geographic regions of Brazil, with a higher concentration of beneficiaries in the Southeast and Northeast regions. Patients from all states represented within the insurer’s network were eligible for inclusion, allowing for nationwide representation within the private supplementary sector.

Population
Eligible patients were adults (≥ 18 years) with a histologically confirmed diagnosis of primary lung cancer between January 2020 and December 2024. The date of the first positive biopsy was defined as the index event. Patients were excluded if diagnostic confirmation was not available in structured records.

Data sources and variables
We used anonymized insurance claims, procedure authorizations, ICD-10 coded encounters, and institutional clinical data warehouses. Variables included:Demographic: age, sex, geographic region.

Clinical: smoking history, comorbidities, histology, tumor location, stage.

Treatment: surgery, chemotherapy, radiotherapy, targeted therapies, immunotherapies.

Healthcare utilization: outpatient visits, emergency visits, hospitalizations, ICU admissions.

Costs: inpatient and outpatient care, diagnostics, procedures, and systemic therapies.

Stage information followed AJCC criteria.

Biomarker data (EGFR, ALK, ROS1, KRAS, PD-L1) were extracted from structured reports; due to incomplete testing, results represent only patients who underwent molecular profiling.

We must note that molecular profiling was not systematically performed in all patients and it reflects the practice in clinical routine within the supplementary health sector. Biomarker testing was available for a subset of patients, primarily those with non–small cell lung cancer and was based on tissue-derived assays reported in structured pathology or molecular reports. EGFR, ALK, ROS1, KRAS mutations and PD-L1 expression were assessed using institutionally available techniques, including immunohistochemistry and polymerase chain reaction–based methods. As testing was performed at the discretion of the treating physician and according to availability over time, the proportion of patients undergoing molecular profiling varied across clinical stages and histologic subtypes.

Cost analysis
Direct medical costs were obtained from the insurer’s financial databases for the period 2019–2024. Categories included hospitalizations, surgeries, imaging, laboratory tests, outpatient consultations, and systemic therapies. Costs were adjusted for inflation using the Brazilian consumer price index (IPCA) and expressed in 2024 Brazilian reais (BRL). Values in US dollars (USD, conversion July 31, 2025; 1 USD = R$5.46) are provided in Supplementary material.
Additionaly, the cost analysis was conducted from the private payer perspective and included direct medical costs accrued from the date of diagnostic confirmation until the end of follow-up or death. Costs therefore reflect cumulative healthcare utilization over variable survival times and were not adjusted for person-time exposure.

Statistical analysis
Descriptive statistics were reported as medians with interquartile ranges (IQR) for continuous variables and frequencies for categorical variables. Group comparisons were performed using chi-square, Kruskal–Wallis, or Mann–Whitney U tests as appropriate. Kaplan–Meier survival curves were estimated for overall survival, with log-rank tests for comparisons by stage, histology, smoking status, and surgical treatment. Statistical significance was set at p < 0.05. Analyses were performed using SPSS 30.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 10.0 (GraphPad Software, San Diego, CA, USA).

Results

Cohort characteristics
A total of 690 patients with lung cancer were identified between 2020 and 2024. The median age was 67 years (IQR 60–74), and 56.4% were women. Stage IV disease was most frequent (35.7%), while 32.5% had no documented stage. Adenocarcinoma predominated (64.5%), followed by squamous cell carcinoma (17.1%) and small-cell lung cancer (4.9%).
Hypertension (62.5%), diabetes (39.0%), and COPD (24.3%) were the most common comorbidities (Supplementary Table S1). A history of smoking was reported in 61.4%, with a median burden of 45 pack-years (Supplementary Table S2).

Diagnosis and staging
Most patients were diagnosed via image-guided biopsy (26.8%) or bronchoscopy (21.7%). Chest CT was the most frequently used imaging modality (68.6%) (Supplementary Table S3). Tumors were more often located in the upper lobes (43.5%) and in the right lung (50.3%) (Supplementary Table S3).
Furthermore, descriptive statistics were calculated based on available information for each variable. Percentages for diagnostic modalities and clinical characteristics therefore refer only to patients with documented data, while missing values were reported separately and not included in the denominators for descriptive comparisons.

Treatment patterns
Surgical resection was performed in 37.5% of patients, most commonly lobectomy (69.1%) and segmentectomy (26.3%). Chemotherapy was administered in 64.9% of patients, predominantly with palliative intent (51.0%). Adjuvant (8.3%) and neoadjuvant (1.3%) chemotherapy were less common. Radiotherapy was indicated in 48.8% of cases.
Immunotherapy was administered to 132 of 690 patients (19.1%), while targeted therapies were used in 94 of 690 patients (13.6%), reflecting selective use based on clinical and molecular eligibility.

Healthcare resources utilization
Hospitalization occurred in 78.5% of patients, with a median length of stay of 3 days (IQR 1–7). ICU admissions were recorded for 36.4% of patients, with a median stay of 2.6 days (IQR 1.2–4.0). The number of outpatient consultations and emergency visits was high, reflecting the complex care needs of this population.

Survival outcomes
At the time of analysis, 147 deaths (21.3%) were documented, with lung cancer as the main cause. Overall survival differed significantly by stage (log-rank p < 0.0001), histology (p = 0.0033), and smoking status (p = 0.0152). Patients who underwent surgical resection had superior survival compared with non-surgical patients (p = 0.0165). Survival analyses should be interpreted with caution due to 32.5% missing stage data.
Exploratory survival analyses according to molecular status and use of targeted therapies were limited by the non-systematic nature of biomarker testing and incomplete documentation of treatment regimens. Among patients with identified actionable alterations who received targeted therapy, overall survival tended to be longer compared with patients without targeted treatment; however, formal comparative analyses were underpowered and should be interpreted with caution.

Costs of care
Median direct healthcare costs increased with advancing stage:

Stage I: R$19,544

Stage II: R$26,213

Stage III: R$31,528

Stage IV: R$75,837

Hospitalizations and surgeries were the largest cost drivers across all stages, followed by imaging and laboratory testing. Patients with adenocarcinoma and squamous cell carcinoma had the highest median costs, reflecting both advanced-stage presentation and the use of systemic therapies. Detailed cost breakdowns are presented in Supplementary tables.

Patient demographics and tumor characteristics
A total of 690 patients diagnosed with lung cancer between 2020 and 2024 were included in the cohort (Table 1). The majority resided in the Southeast region of Brazil, though all five geographic regions were represented. Median age at diagnosis was 67 years (IQR 60–74), and 56.4% of the patients were female. Clinical staging at diagnosis was heterogeneous: stage IV accounted for 35.7% of cases, while stages I, II, and III represented 13.4%, 8.1%, and 10.1%, respectively (Table 2). Clinical stage was not reported in 32.5% of patients. Adenocarcinoma was the predominant histologic type (64.5%), followed by squamous cell carcinoma (17.1%) and small-cell lung cancer (4.9%) (Table 3).

Tumor grading was available in 48.8% of patients, with 16.2% presenting well-differentiated tumors and 16.2% poorly or undifferentiated lesions. Tumor size increased proportionally with clinical stage, from a median of 2.2 cm in stage I to 4.3 cm in stage IV. Tumors were most frequently located in the upper lobes (43.5%) and exhibited right-side predominance (50.3%).

Genetic mutations and comorbidities
Molecular profiling was performed in a subset of patients, revealing EGFR positivity in 10.7%, ALK rearrangement in 2.9%, ROS1 in 0.6%, and PD-L1 positivity in 18.8%. Among those with PD-L1 expression, 47 patients had high expression and 68 low expression. KRAS mutations were less frequently tested, with a positivity rate of 1.3% (Table 4). The most common comorbidities were hypertension (62.5%), diabetes (39.0%), and COPD (24.3%) (Supplementary Table S1).

Smoking history
A history of smoking was reported in 61.4% of patients, with 13.6% current smokers and 48.0% former smokers. The median smoking burden was 45 pack-years (IQR 30–60), and squamous cell carcinoma was more frequently associated with heavier smoking history compared to adenocarcinoma and carcinoid tumors (Supplementary Table S2).

Diagnostic workup and biopsy methods
Most tumors were diagnosed via image-guided percutaneous biopsy (26.8%) or bronchoscopy (21.7%). Less frequent methods included surgical resection or thoracoscopy. Imaging modalities used for initial tumor identification included chest CT (most frequent), MRI, and PET-CT. Detailed distribution of diagnostic modalities is provided in Supplementary Table S3.

Treatment patterns
Surgical treatment was performed in 37.5% of patients (Table 5), with lobectomy (69.1%) and segmentectomy (26.3%) being the most common procedures. Chemotherapy was administered to 64.9% of patients, most commonly in the palliative setting (51.0%), followed by adjuvant (8.3%), salvage (4.5%), and neoadjuvant (1.3%) (Table 6). Immunotherapy agents (e.g., pembrolizumab, durvalumab, atezolizumab) and targeted therapies (e.g., gefitinib, osimertinib, afatinib, crizotinib) were included in several treatment regimens.

Healthcare utilization
The median number of outpatient visits prior to biopsy was 13 (IQR 5–43), and 9 (IQR 5–16) post-biopsy. Emergency department visits were also frequent, with a median of 5 (IQR 1.5–8.5) prior to and 4 (IQR 0–8.4) following diagnosis. Hospitalization occurred in 78.5% of patients, with a median length of stay of 3 days (IQR 1–7). ICU admissions occurred in 36.4% of cases, with a median stay of 2.6 days (IQR 1.2–4.0) (Table 7).

Survival analysis
At the time of analysis, 147 deaths had been recorded (21.3%), with lung cancer being the leading cause. Overall survival varied significantly by stage (log-rank p < 0.0001), histologic subtype (p = 0.0033), and smoking status (p = 0.0152 for smokers vs. never-smokers). Patients undergoing surgical resection had superior survival compared to those managed with non-surgical approaches (p = 0.0165).

Cost analysis
The total median cost of lung cancer care increased with advancing stage, ranging from R$19,544 (US$3,579) in stage I to R$75,837 (US$13,890) in stage IV (Table 8). Hospitalizations and surgical procedures were the primary cost drivers:

Stage I: Hospitalization R$12,673 (US$2,321), Imaging R$3,261 (US$597), Laboratory R$2,915 (US$534), Consultations R$3,745 (US$686)

Stage II: Hospitalization R$16,690 (US$3,056), Imaging R$3,302 (US$605), Laboratory R$3,091 (US$566), Consultations R$4,483 (US$821)

Stage III: Hospitalization R$13,915 (US$2,549), Imaging R$4,141 (US$758), Laboratory R$5,741 (US$1,051), Consultations R$4,039 (US$740)

Stage IV: Hospitalization R$12,061 (US$2,210), Imaging R$4,284 (US$785), Laboratory R$6,911 (US$1,266), Consultations R$3,527 (US$646)

Systemic anticancer treatments, including chemotherapy, immunotherapy, and targeted therapies, accounted for a substantial proportion of total costs, particularly in advanced-stage disease. Hospitalizations and surgical procedures remained the dominant cost components overall, while systemic therapies represented a growing share of expenditures in stage III and IV patients.
Values were converted to US dollars using the official exchange rate of July 31, 2025 (1 USD = R$ 5.46; source: Brazil Central Bank) [10]. While comprehensive, these cost data will be further analyzed in a dedicated cost-effectiveness and budget impact study.
Costs include hospitalizations, surgeries, imaging, outpatient consultations, and systemic therapies. Values are adjusted to 2024 Brazilian reais using IPCA inflation rates.

Discussion
This study provides a large-scale real-world overview of lung cancer care within Brazil’s private supplementary health sector, analyzing 690 patients diagnosed between 2020 and 2024. To our knowledge, this represents one of the most comprehensive characterizations of treatment patterns, healthcare utilization, survival outcomes, and costs in the supplementary system, which covers nearly one-quarter of the Brazilian population [11].
Our results highlight that more than one-third of patients presented with stage IV disease, and an additional one-third lacked stage documentation. This mirrors the challenges of delayed diagnosis and incomplete staging previously described in Brazilian series, and underscores the persistent barriers to early detection. Similar to international reports, adenocarcinoma was the predominant histology, and smoking remained strongly associated with worse outcomes [12].
When compared to prior national studies, our findings add new perspectives. Mathias et al. described lung cancer epidemiology and treatment trends in Brazil but did not include detailed cost analyses or stratification within the private sector. Baldotto et al. examined treatment patterns and the economic burden of advanced NSCLC in selected private institutions, focusing exclusively on late-stage disease. In contrast, our study encompasses all stages, includes patients from across the country, and provides stage-stratified costs, hospitalization rates, and ICU utilization. These complementary insights reveal both the opportunities and inefficiencies of cancer care in supplementary health [3, 8].
The relatively low proportion of patients undergoing surgery (37.5%), despite nearly 22% being diagnosed with stage I–II disease, raises questions about referral patterns, surgical capacity, and documentation accuracy. Similarly, while chemotherapy was the most common modality, systemic therapy intent was predominantly palliative, reflecting the advanced disease burden at diagnosis. The use of immunotherapy and targeted therapies was observed, but limited molecular testing likely restricted broader adoption, a gap already emphasized in Latin American studies [11]. Expanding access to biomarker testing remains a critical step for aligning practice with international guidelines.
Although molecular alterations such as EGFR and ALK were identified in a subset of patients, molecular profiling was not universally available, limiting the ability to fully assess survival differences according to biomarker status or targeted treatment. Nevertheless, patients who received targeted therapies appeared to have favorable survival trends, consistent with international real-world and clinical trial data. These findings reinforce the importance of expanding access to molecular testing in the Brazilian supplementary sector to enable more precise therapeutic decision-making.
Healthcare utilization was substantial: nearly 80% of patients required hospitalization and over one-third ICU care, generating high direct costs. Stage IV disease was associated with more than a two-fold increase in expenditures compared with stage I, largely driven by hospitalizations and systemic therapy. This pattern is consistent with international series [8, 13], but the relative weight of inpatient costs in our data was higher than reported in high-income countries [10, 14]. This suggests that optimization of hospitalization and perioperative pathways may be an important strategy for improving efficiency in the Brazilian private sector.
Cost estimates of our analyses should be interpreted in light of the study’s analytic perspective and follow-up structure. As expected, since costs were accumulated over the observed disease course, longer survival may be associated with higher cumulative expenditures in earlier stages, whereas advanced-stage disease was characterized by more intensive resource use over shorter survival periods, largely driven by hospitalizations and systemic therapies.
Limitations of this study include reliance on administrative and EMR-based data, which are prone to underreporting or miscoding. A major limitation of this study is certainly the absence of documented clinical stage for approximately one-third of patients. Missing staging information may reflect limitations of administrative and EMR-based data sources, as well as fragmented diagnostic pathways across institutions. This limitation restricted stage-stratified survival analyses and may have introduced residual confounding, particularly in comparisons involving treatment patterns and outcomes. Additionaly, performance status, quality of life, and detailed systemic regimens were not consistently available. Finally, as data derive from a single private insurer, generalizability to the Brazilian public health system (SUS) is limited. Nevertheless, given the supplementary sector’s central role in oncology care, these findings remain highly relevant for policy and planning.

Conclusion
In summary, this real-world analysis of lung cancer care within Brazil’s largest private health insurance network reveals a predominance of late-stage diagnoses, incomplete staging documentation, limited access to molecular testing, and high rates of hospitalization and ICU use. Direct costs of care increased markedly with advancing stage, with hospitalizations and surgeries as the major cost drivers.
By including all disease stages and patients from across the country, this study expands upon prior Brazilian reports and offers new evidence to inform resource allocation and access policies in the supplementary health system. Efforts to promote earlier detection, broaden molecular testing, and optimize hospital resource utilization may help improve outcomes and efficiency of lung cancer care in Brazil.

Supplementary Information