To our knowledge, this study is the first to provide data on HCRU and overall costs in patients with metastatic lung cancer over time in France, which is of interest during a period of unprecedented improvement in survival of patients with metastatic lung cancer in recent years. These results are particularly interesting because lung cancer is the second most costly cancer after breast cancer. It is primarily diagnosed in the metastatic setting and has been an area of much innovation over the past decade (15, 16). Although the cost of managing metastatic lung cancer has increased globally, the average monthly cost per patient has not increased significantly between 2013 and 2021. Drug acquisition costs were offset by lower overall hospitalization costs (except in 2020), resulting in no significant upward trend.
In France, the National Health Service (NHS – Caisse Nationale de l’Assurance Maladie), which annually publishes reimbursement expenditures by pathology, confirmed that the amount allocated to lung cancer management (all stages combined) during the active treatment phase has increased each year (16). Although these estimates are reported for all stages (approximately €2.3 billion in France in 2019), it is particularly interesting to put them in perspective with the results reporting data on metastatic patients. The increase in overall costs observed in this study may be explained by demographic changes (growing and aging population), increased cancer risk, and improved life expectancy for patients treated for metastatic lung cancer. In light of the available literature, the average monthly cost per patient in the metastatic stage estimated in this study in 2019 was 2.5 times higher than that estimated by the NHS for all patients in the active treatment stage, regardless of stage (€57,360 per year versus €23,100 per year). This is consistent with the fact that the cost of lung cancer care increases with disease stage (17, 18). The NHS also observed a decline in the cost of hospital admissions and an increase in the cost of expensive medicines between 2015 and 2019 (16). However, unlike what was seen in the transition phase in our study, hospitalization costs did not offset patients’ medication costs at any phase. It can be assumed that the evolution of the treatment environment has had a greater impact on the most severely ill patients in the metastatic phase, which may reduce the need for hospitalization more than less severely ill patients. International studies examining the economic burden of lung cancer since the treatment environment has evolved have shown increasing costs of systemic therapy (10, 11, 19). Zhang et al. showed that inpatient, outpatient, and pharmacy costs remained relatively stable (11). Korytowsky et al, who matched two cohorts before and after immunotherapy approval, found lower rates of emergency room visits and hospitalizations in the post-immunotherapy period, which resulted in lower associated costs, which had a favorable impact on total monthly treatment costs per patient ($12,681 vs. $10,758 for the pre- and post-immunotherapy periods, respectively). (19).
Over the past decade, the advent of new treatments, such as immunotherapy and targeted therapies, has led to significant changes in the treatment of metastatic lung cancer, making promising advances that may contribute to reducing total hospitalizations. Compared to existing treatments, these innovative treatments provide better symptom management, limit serious complications, and reduce the need for frequent full hospitalization. Moreover, in a situation characterized by an aging population, the development of chronic pathologies and a strong desire for home care, France needs a deep restructuring of the organization of health services. One of the main strategies chosen is the development of outpatient care, which began in the early 2000s but actually emerged in a report published by the French National Audit Office in 2013 and was strengthened by the three-year plan for 2015-2017, implemented in 2014, and by the French Healthcare System Modernization Act of January 26, 2016 (20,21,22). Therefore, the observed decrease in average monthly costs between 2013 and 2014 may be a result of the implementation of these health policies. One of the two main goals of this plan is to reduce the number of complete admissions (as observed in this study) and favor same-day admissions. Because outpatient surgery is not expected in this population (oncologic surgery is not performed at the metastatic stage), the reliable use of same-day hospitalization described in this study is consistent. Instead, it appears to have been converted to outpatient treatment.
However, the potential contribution of health policies implemented in France since 2013 to modifying health resource utilization has been affected by the COVID-19 pandemic in 2020. This was the only year in this study in which the average monthly cost per patient increased slightly. The COVID-19 pandemic has particularly impacted lung cancer patients, who are at the highest risk of severe COVID-19 disease and mortality (25–30%) after receiving treatment for hematological malignancies (23). Many learned medical societies and collaborative organizations quickly issued recommendations for the management and prevention of COVID-19 in patients with lung cancer (24,25,26–27
SNDS is a national medical claims database covering more than 98% of the French population. Overall, this database is representative of the entire population leaving for France. It also includes comprehensive information on treatment and use of reimbursed medical resources, as well as comprehensive costs. All costs in our study are expressed in inflation-adjusted euros, allowing consistent comparisons over time. This allowed us to provide the most comprehensive view of the burden of metastatic lung cancer in the French NHS among available data sources.
It was not possible to focus on non-small cell lung cancer. That’s because there are no specific ICD-10 codes or treatments to differentiate between these two types of lung cancer. Nevertheless, small cell lung cancer accounts for only 15% of lung cancers, so the bias can be considered limited. Moreover, coding of metastases is not mandatory and may be delayed or unavailable, which may lead to selection bias if coding varies depending on patient severity. In practice, transition coding appears to be performed more frequently on full admissions (to increase hospitalization costs) than on day admissions for therapeutic infusions. This means that some of the milder patients who come to hospital only for therapeutic infusions may be missed and costs overestimated. However, given the severity of metastatic lung cancer, this bias is clearly limited. Additionally, the administration of bevacizumab and pemetrexed, two treatments used in the metastatic phase, were also used as surrogates to reduce this bias. In any case, this bias was similar each year and did not affect the trend test. Lastly, by using only discharge diagnosis and hospital treatment as selection criteria, there is a possibility that some patients who received only outpatient treatment were omitted, resulting in selection bias toward the most seriously ill patients. Although this only concerns a small portion of our population, the burden described in this study probably cannot be generalized to the entire population of patients with metastatic lung cancer, and is likely only to the most severely ill patients. Additionally, our results reflect French healthcare policies and therefore are difficult to generalize to patients with metastatic lung cancer worldwide. However, our results appear to be consistent with those of other studies abroad (10, 19).
Additionally, unfortunately, localization of primary cancer is not possible in the SNDS database. Therefore, based on a 7-year prior study period, we developed an algorithm to identify lung cancer and exclude patients with lung metastases from other cancers. This algorithm was specifically selected and although this may have excluded some patients with metastatic lung cancer, it was able to account for HCRU and cost estimates specific to metastatic lung cancer.
Finally, inpatient reimbursement for conventional chemotherapy (i.e., drugs that are neither expensive nor innovative) is not available in the SNDS database because it is directly included in the cost of hospital admission. Although this did not account for the overall use of chemotherapy, it did not affect average monthly costs per patient, per year. However, it is important to note that the acquisition cost of all anticancer drugs was calculated based on the total amount, not the net price negotiated with the Economic Committee. In particular, oncology drugs have the highest average rebate rate per therapeutic class (34.5% according to the latest 2022 activity report published by the Economic Committee), which leads to an overestimation of acquisition costs in our analysis. Finally, not including some patients who received only outpatient treatment may lead to an underestimation of targeted therapy use and may explain the decline in targeted therapy use. Patients who received only outpatient treatment at the start of the study period are likely to be included in this study because they have more follow-up than patients included in 2020 or 2021 and are therefore more likely to be hospitalized or receive inpatient treatment due to disease progression. Further investigation is needed to investigate whether the addition of patients treated only as outpatients could confirm this and the trends observed in this study.
