Real-world economic burden of disease recurrence in patients with muscle-invasive bladder cancer: A population-level claims-based analysis

Abstract

BACKGROUND:

Bladder cancer is a common cancer with significant morbidity, mortality, and economic cost. Muscle-invasive bladder cancer (MIBC) is typically managed with radical cystectomy (RC). Despite its curative intent, a considerable proportion of patients experience recurrence after RC. The economic impact of recurrence among patients with surgically resected MIBC has not been described.

OBJECTIVE:

To assess health care resource utilization (HCRU) and costs among patients with surgically resected MIBC in the United States, including the impact of disease recurrence.

METHODS:

In this retrospective, observational study, the Surveillance, Epidemiology, and End Results–Medicare database (2007-2020) was used to identify patients diagnosed with T2-T4aN0M0 or T1-T4aN1M0 MIBC who underwent RC in the United States. Index date was the date of RC. Patients were stratified by whether they experienced recurrence following surgical resection. The index date for patients with recurrence was defined as 30 days prior to recurrence, and for patients without recurrence, the index date was drawn from a distribution to match the time window between surgical resection and the index date in the recurrence cohort. Patients were followed from the index date until the end of data availability, continuous enrollment, or death. Rates of HCRU per patient per year (PPPY) and mean health care costs per patient per month (PPPM; in 2022 USD) were summarized and compared between cohorts.

RESULTS:

A total of 1,149 patients met selection criteria. Patients had a median of 2.6 years of follow-up. Demographic and clinical characteristics were generally similar between patients with (n = 503) and without recurrence (n = 602), with few exceptions. Patients with recurrence (compared with those without) were more likely to have had stage IIIA disease (47.9% vs 32.7%) and a proxy for cisplatin contraindications (54.1% vs 47.5%, both P < 0.05), which included renal insufficiency, peripheral neuropathy, sensorineural hearing loss, or cardiac disease. Following index, patients with surgically resected MIBC had 3.5 all-cause inpatient admissions, 1.0 all-cause emergency department (ED) visits, and 25.8 all-cause outpatient visits PPPY. Patients with recurrence had higher rates of all-cause HCRU than patients without recurrence after index, including inpatient admissions (adjusted incidence rate ratio: 2.4), ED visits (2.7), and outpatient visits (2.0; all P < 0.001). The total all-cause medical costs PPPM were $11,250 and were higher for patients with vs without recurrence ($10,030 vs $3,343; adjusted cost difference: $7,191), largely because of higher inpatient admissions costs ($6,654 vs $2,102; adjusted cost difference: $4,542; both P < 0.001).

CONCLUSIONS:

Surgically resected MIBC was associated with a substantial economic burden with disease recurrence experiencing higher HCRU and health care costs. These findings underscore the need for novel and effective therapies that can prevent or delay disease recurrence for patients with MIBC.

Plain language summary

This study showed that patients with muscle-invasive bladder cancer (MIBC) who had their bladder removed had substantial health care resource use and costs, with a mean monthly total medical cost of $11,250 per patient. After bladder removal, patients whose disease came back (ie, recurred) had a higher economic burden compared with those who remained tumor free, including $7,191 higher mean monthly medical costs per patient after adjusting for patient and clinical characteristics. These findings highlight the importance of preventing or delaying disease recurrence to potentially alleviate the economic burden on patients and the health care system.

Implications for managed care pharmacy

Surgically resected MIBC and its recurrence imposed a substantial economic burden on patients with MIBC and the health care system. Patients with MIBC had a mean monthly total all-cause medical cost of $11,250 per patient. Patients with recurrence had higher adjusted rates of health care resource use, including a 2.4 times higher rate of hospitalization, 2.7 times higher rate of emergency department visits, and 2.0 times higher rate of outpatient visits. The results of this real-world study highlight the need for developing novel treatments that effectively prevent or delay recurrence and alleviate economic burden in patients with MIBC.

Bladder cancer (BC) is a common cancer with approximately 83,190 new cases and 16,840 deaths estimated in the United States in 2024. ¹ Approximately 25% to 30% of patients with BC present with muscle-invasive BC (MIBC). ² MIBC is a lethal cancer given its high propensity to spread to lymph nodes and other organs. ³

Despite curative-intent treatment with surgery (radical cystectomy [RC] and pelvic lymph node dissection [PLND]), approximately 50% of patients experience recurrence within 2 years. Recurrent metastatic urothelial cancer is associated with a historically poor prognosis (5-year overall survival [OS] rate <15%), although recent advances in treatment are likely to increase OS. ^{4 , 5}

Several treatment strategies have been developed to reduce risk of recurrence, including preoperative (“neoadjuvant”) or postoperative (“adjuvant”) chemotherapy with cisplatin-based treatment regimens. In addition, postoperative adjuvant immunotherapy with nivolumab was also recently approved in the United States in 2021, and most recently, preoperative durvalumab in combination with chemotherapy was approved in the United States in 2025. ^{6 , 7}

Notably, treatment of recurrent MIBC depends on prior therapies received and type of recurrence, such locoregional recurrence (LRR) vs distant metastasis (DM), but may include the use of chemotherapy, immunotherapy, and/or targeted therapy. ⁸

Much of the research to date has focused on health care resource utilization (HCRU) and health care costs associated with MIBC after treatment with RC, noting a high economic burden. A 2025 US study found that patients with BC who received RC had a median of 56 health care visits in 12 months with a median cost of $103,579. ⁹ A 2023 study found that patients with MIBC in the United States who had cystectomy had mean all-cause total costs of $136,104 per patient per year (PPPY). ¹⁰ However, studies on the economic burden of disease recurrence compared with no recurrence in patients with radically resected MIBC remain limited. Given the high recurrence rates associated with MIBC, quantifying the economic burden of disease recurrence after surgical resection is an important component in evaluating the potential value of preventing or delaying disease recurrence in this patient population. For example, patients with recurrent MIBC may experience a higher burden of health care visits relative to patients without recurrence owing to an increased number of scans and/or visits to diagnose the recurrence, the treatment of the recurrent disease, and increased monitoring of patients after recurrence has been detected.

This retrospective observational study was conducted using Surveillance, Epidemiology, and End Results (SEER)–Medicare linked data to quantify the overall economic burden since the time of surgery and further to assess the real-world impact of disease recurrence on HCRU and health care costs among patients with surgically resected MIBC at the population level.

Methods

STUDY DESIGN AND DATA SOURCE

Patients with MIBC who received RC with or without PLND were identified from the SEER-Medicare database (2007-2020). The database links data from 2 sources—namely, SEER cancer registries, which contains clinical and demographic information and cause of death, and Medicare insurance claims for health care services used by beneficiaries from the time of eligibility until death. The linked SEER-Medicare database included all Medicare-eligible people in the SEER registry diagnosed with cancer through 2019 and their Medicare claims through 2020. Information on Medicare drug coverage (Part D) was available starting from 2007. All data were limited and adhered to the Health Insurance Portability and Accountability Act and Declaration of Helsinki; therefore, no reviews by an institutional review board were required.

This study used a retrospective longitudinal observational cohort design. To describe the overall economic burden, the date of surgical resection was designated as the index date, irrespective of recurrence status. To assess the impact of disease recurrence on HCRU and health care costs, the index date for patients with recurrence was defined as the date 30 days before the date of recurrence to capture claims in the study period that were likely related to recurrence but occurred before the claims related to recurrence was observed. For patients without recurrence, the index date was assigned based on the beta distribution of time between surgical resection and recurrence in patients who experienced a recurrence. For all analyses, the baseline period was defined as 12 months before the index date, and the follow-up period was the time from the index date until the earliest of end of data availability, end of Medicare coverage, or death (Supplementary Figure 1 ^{(250.8KB, pdf)} , available in online article).

STUDY POPULATION

The sample selection flowchart is shown in Figure 1. Inclusion criteria were as follows: (1) record of a diagnosis of urothelial carcinoma between 2007 and 2019 based on the International Classification of Diseases for Oncology, Third Edition code C67 and relevant histology codes (8120-8131); (2) diagnosed with T2-T4aN0M0 or T1-T4aN1M0 stage or summary stage 2 of MIBC (if complete Tumor Node Metastasis Classification of Malignant Tumors information was not available) based on American Joint Committee on Cancer 7th edition tumor node metastasis staging criteria ¹¹ ; (3) aged 65 years or older at the time of initial diagnosis of MIBC; (4) underwent RC with or without PLND within 12 months after initial MIBC diagnosis, and with continuous enrollment for Medicare Parts A, B, and D for at least 12 months prior to and 1 month after surgical resection; (5) had not received partial cystectomy or radiation therapy for MIBC prior to RC with or without PLND and after MIBC diagnosis; (6) had no secondary malignancies any time prior to or within 60 days after RC with or without PLND; and (7) had no other non-MIBC cancers within 3 years prior to the initial diagnosis of MIBC.

FIGURE 1.

Sample Selection and Patient Cohorts

AJCC = American Joint Committee on Cancer; MIBC = muscle-invasive bladder cancer.

Patients who met the inclusion criteria were included in the study cohort and were further stratified by the recurrence status after surgical resection. An algorithm was developed through clinical input to identify disease recurrence, including LRR and DM, based on the diagnosis and treatments received. LRR was defined as a diagnosis of secondary malignancy in the urinary bladder or pelvic lymph node at least 60 days after the index date or initiation of radiation therapy at least 180 days after the end of the primary treatment episode, whichever occurred first. DM was defined as a diagnosis of metastatic disease at least 60 days after the index date or initiation of systemic treatment for metastatic disease, whichever occurred first. For patients who did not receive adjuvant treatment, the following events were considered as initiation of systemic treatment: (1) initiation of systemic treatment for metastatic disease that could also be used in the adjuvant setting at least 180 days after the index date or (2) initiation of systemic treatment intended to treat metastatic disease after the index date. For patients who received adjuvant treatment, any of the following events were considered as initiation of systemic treatment: (1) initiation of a new treatment for metastatic disease that could also be used in the adjuvant setting at least 30 days after the start of the adjuvant regimen; (2) reinitiation of the same treatment as received for adjuvant treatment at least 120 days after the end of the adjuvant regimen; or (3) initiation of systemic treatment intended to treat metastatic disease after the start of adjuvant therapy. The adjuvant treatment list was based on approved or recommended drugs for adjuvant use, whereas treatments for metastatic disease included approved or recommended drugs for recurrence (Supplementary Table 1 ^{(250.8KB, pdf)} ).

STUDY MEASURES AND OUTCOMES

Baseline demographic and clinical characteristics were described. All-cause and BC-related HCRU was assessed, including inpatient admissions, emergency department (ED) visits, skilled nursing facility visits, outpatient visits, and inpatient length of stay. BC-related HCRU was defined as claims for medical services associated with a diagnosis for BC or for the administration of antineoplastic treatments for BC. Specifically, a claim was considered BC related if there was a primary diagnosis of BC for an outpatient. For other types of visits, a primary or secondary diagnosis of BC was used to determine whether the claim was BC related. All-cause and BC-related HCRU outcomes were summarized as incidence rates PPPY, defined on the patient level as the ratio between total number of events and years of follow-up to account for variable lengths of follow-up in the study period across individual patients. Per patient per month (PPPM) all-cause and BC-related health care costs, which were amounts paid by Medicare, were summarized to account for different follow-up time across individual patients. BC-related health care costs were costs associated with BC-related HCRU and included medical costs and BC-related Medicare Part D pharmacy costs. All costs were inflated to 2022 US dollars using the personal consumption expenditures medical care component. ¹²

STATISTICAL ANALYSIS

Baseline characteristics were described and compared between patients with and without recurrence. Means and SDs were reported for continuous variables; frequency counts and percentages were reported for categorical variables. Statistical comparisons were conducted between the 2 cohorts with and without recurrence using t-tests for continuous variables and chi-square tests for categorical variables. Rates of all-cause and BC-related HCRU were compared between patients with and without recurrence using generalized linear models with a Poisson distribution and log-link function. Variable lengths of follow-up time across individual patients were accounted for by including an offset term. Incidence rate ratios (IRRs) were reported with corresponding 95% CIs and P values. All-cause and BC-related inpatient length of stay and mean monthly health care costs were compared between patients with and without recurrence using generalized linear models with a Tweedie distribution and log-link function. An offset term was included in the model to account for different lengths of follow-up. Cost differences and corresponding P values were reported. For both HCRU and health care costs, the covariates included in regression models included age at MIBC diagnosis, time from initial MIBC diagnosis to surgical resection, sex, race and ethnicity, region, disease stage at MIBC diagnosis (American Joint Committee on Cancer 7th edition), year of index, Charlson Comorbidity Index (CCI) score, presence of hypertension, any proxy for cisplatin contraindications, any inpatient admissions during the baseline period, and any ED visits during the baseline period.

SAS Enterprise 7.1 software (SAS Institute) was used for all statistical analyses. P values less than 0.05 were considered statistically significant.

Results

Of the 281,476 patients with urothelial carcinoma in the SEER-Medicare data from 2007 and 2019, 1,149 met all selection criteria and were included in the overall radically resected MIBC study population. The median follow-up duration after surgical resection was 2.6 years (IQR, 1.1-5.4).

Among the patients who met the additional eligibility criteria for stratification by recurrence, 1,105 were included in the recurrence analysis, with 503 (45.5%) assigned to the recurrence cohort and 602 (54.5%) to the nonrecurrence cohort (Figure 1). Most (76.7%) patients with recurrence were identified based on diagnosis-based indicators. The median follow-up durations from the time of recurrence or shifted index were 1.0 years (IQR, 0.3-2.7) and 1.6 years (0.6-3.3) for the recurrence and nonrecurrence cohorts, respectively.

BASELINE CHARACTERISTICS

Among the overall population, the mean age at surgical resection was 74.2 years, with 74.8% being male and 86.1% White (Supplementary Table 2 ^{(250.8KB, pdf)} ). Clinical characteristics have been previously published, with common comorbidities including hypertension (82.3%), chronic pulmonary disease (37.2%), and peripheral vascular disease (32.2%). ¹³ In terms of HCRU during the baseline period, patients had 1.2 all-cause and 0.9 BC-related inpatient admissions and 30.3 all-cause and 10.7 BC-related outpatient visits PPPY. The all-cause and BC-related total medical costs were $2,451 and $1,329 PPPM, respectively, during the baseline period.

Clinical and demographic characteristics of the cohort by recurrence status are shown in Table 1. The mean age for patients with and without recurrence was 75.0 years and 75.8 years, respectively. Among patients with recurrence, the majority had a DM (91.1% had DM only, 5.8% had LRR only, and 3.2% had both LRR and DM). The average CCI score was approximately 2.0 in both cohorts. Patients with recurrence were more likely to have high-risk disease (stage IIIA disease: 47.9% vs 32.7%; P < 0.001) and have a proxy for cisplatin contraindication (ie, diagnosis of renal insufficiency, peripheral neuropathy, sensorineural hearing loss, or cardiac disease) (54.1% vs 47.5%; P < 0.05). Among the specific cisplatin contraindication indications, patients with recurrence were more likely to have a diagnosis of renal insufficiency than patients without recurrence and had a similar proportion of diagnoses of peripheral neuropathy, sensorineural hearing loss, and cardiac disease.

TABLE 1.

Baseline Characteristics of Patients With MIBC by Recurrence Status

	Patients with recurrence (n = 503)	Patients without recurrence (n = 602)	P value
Demographic characteristics
Age at recurrence or matched index (years)	75.0 [5.8]	75.8 [5.7]	0.021 a
Time from initial MIBC diagnosis to recurrence or matched index (years)	1.9 [1.8]	2.0 [1.7]	0.153
Sex			0.307
Female	121 (24.1%)	161 (26.7%)
Male	382 (75.9%)	441 (73.3%)
White	434 (86.3%)	517 (85.9%)	0.848
Year of recurrence or matched index			<0.001 a
2007-2011	112 (22.3%)	102 (16.9%)
2012-2016	236 (46.9%)	215 (35.7%)
2017-2020	155 (30.8%)	285 (47.3%)
Clinical characteristics
CCI during baseline period	1.9 [1.7]	2.0 [1.9]	0.371
Any proxy for cisplatin contraindications b	272 (54.1%)	286 (47.5%)	0.030 a
Renal insufficiency	131 (26.0%)	114 (18.9%)	0.005 a
Peripheral neuropathy	56 (11.1%)	58 (9.6%)	0.415
Sensorineural hearing loss	41 (8.2%)	41 (6.8%)	0.397
Cardiac disease	147 (29.2%)	167 (27.7%)	0.586
Recurrence type
Locoregional only	29 (5.8%)	—	—
Distant metastasis only	458 (91.1%)	—	—
Locoregional followed by distant metastasis	16 (3.2%)	—	—
HCRU during 12 months before index, PPPY
All-cause HCRU
Inpatient admissions	1.9 [1.9]	1.5 [1.7]	<0.001 a
Emergency department visits	1.0 [1.6]	0.8 [1.3]	0.004 a
Outpatient visits	32.9 [18.0]	26.4 [17.2]	<0.001 a
Skilled nursing facility stays	0.3 [0.9]	0.6 [2.7]	0.085
BC-related HCRU
Inpatient admissions	1.4 [1.5]	1.1 [1.3]	<0.001 a
Emergency department visits	0.2 [0.6]	0.2 [0.5]	0.772
Outpatient visits	10.6 [10.5]	8.1 [8.9]	<0.001 a
Skilled nursing facility stays	0.2 [0.5]	0.2 [0.9]	0.284
Health care costs during 12 months before index, PPPM (2022 US Dollars)
All-cause health care costs
Total costs	5,636 [6,004]	4,538 [4,868]	0.001 a
Medical costs	5,373 [5,936]	4,274 [4,694]	0.001 a
BC-related medical costs	2,903 [2,974]	2,455 [2,866]	0.011 a
Medicare Part D pharmacy costs	264 [647]	264 [1,201]	0.999

Data are presented as mean [SD] or n (%).

^a Statistically significant.

^b Specific conditions listed under cisplatin contraindications are further described in Nyame YA, Holt SK, Diamontopoulos LN, et al. Social and clinical correlates of neoadjuvant chemotherapy in Medicare beneficiaries with muscle invasive bladder cancer from 2004-2015. Urology. 2021;149:154-60. doi:10.1016/j.urology.2020.12.020.

BC = bladder cancer; CCI = Charlson Comorbidity Index; HCRU = health care resource utilization; MIBC = muscle-invasive bladder cancer; PPPM = per patient per month; PPPY = per patient per year.

During the 12-month baseline period, patients in the recurrence cohort had higher rates of HCRU than patients in the nonrecurrence cohort (Table 1). Specifically, patients with recurrence had significantly more inpatient admissions (all-cause: 1.9 vs 1.5 admissions PPPY; BC-related: 1.4 vs 1.1 admissions PPPY; both P < 0.001) and outpatient visits (all-cause: 32.9 vs 26.4 visits PPPY; BC-related: 10.6 vs 8.1 visits PPPY; both P < 0.001) than patients without recurrence. In terms of health care costs during the baseline period, patients with recurrence had significantly higher all-cause total medical costs ($5,373 vs $4,274 PPPM; P = 0.001), as well as significantly higher BC-related total medical costs ($2,903 vs $2,455 PPPM; P = 0.011) than patients without recurrence.

HCRU AND HEALTH CARE COSTS ASSOCIATED WITH MIBC AND MIBC RECURRENCE

In the overall population, patients had on average 3.5 inpatient admissions, 1.0 ED visits, and 25.8 outpatient visits per year for all-cause HCRU and 2.8 inpatient admissions, 0.2 ED visits, and 7.9 outpatient visits per year for BC-related HCRU (Supplementary Figure 2 ^{(250.8KB, pdf)} ). The all-cause total medical costs PPPM were $11,250, with $9,158 for inpatient admissions (Supplementary Table 3 ^{(250.8KB, pdf)} ). BC-related total medical costs and costs for inpatient admissions were $7,618 and $6,752, respectively.

Compared with patients without recurrence, those in the recurrence cohort had significantly higher rates of HCRU during the study period (Figure 2). Specifically, patients with recurrence had significantly more all-cause inpatient admissions (3.8 vs 1.5 admissions PPPY; adjusted IRR [aIRR], 2.4; 95% CI, 2.2-2.6), ED visits (1.9 vs 0.7 visits PPPY; aIRR, 2.7; 95% CI, 2.4-3.0), outpatient visits (39.4 vs 19.9 visits PPPY; aIRR, 2.0; 95% CI, 2.0-2.1), and more days hospitalized (30.3 vs 14.1 days PPPY; adjusted mean difference, 18.9) compared with patients without recurrence (all P < 0.001). Differences were more pronounced between the 2 cohorts in terms of BC-related HCRU. Patients in the recurrence cohort had significantly more BC-related inpatient admissions (2.7 vs 0.6 admissions PPPY; aIRR, 4.1; 95% CI, 3.6-4.6), ED visits (0.5 vs 0.1 visits PPPY; aIRR, 3.5; 95% CI, 2.6-4.6), and outpatient visits (14.4 vs 4.0 visits PPPY; aIRR, 3.2; 95% CI, 3.1-3.4) than patients without recurrence (all P < 0.001). Unadjusted comparisons of all-cause and BC-related HCRU between patients with and without recurrence are summarized in Supplementary Table 4 ^{(250.8KB, pdf)} .

FIGURE 2.

All-Cause (A) and BC-Related (B) HCRU for Patients by Recurrence Status

^aStatistically significant.

aIRR = adjusted incidence rate ratio; BC = bladder cancer; HCRU = health care resource utilization; MIBC = muscle-invasive bladder cancer; PPPY = per patient per year.

Compared with patients without recurrence, those in the recurrence cohort had significantly higher mean monthly all-cause and BC-related health care costs (Table 2). The adjusted costs for patients in the recurrence cohort compared with those in the nonrecurrence cohort were $7,191 higher for all-cause total medical expenses, including costs that were $4,542 higher for inpatient admissions, $84 higher for ED visits, and $1,610 higher for outpatient visits on average (all P < 0.001). For BC-related health care costs, the adjusted mean differences in monthly cost per patient were $3,073, $2,171, $46, and $546, for total medical costs, inpatient admissions, ED visits, and outpatient visits, respectively (all P < 0.001).

TABLE 2.

Comparison of All-Cause and BC-Related Health Care Costs (2022 US Dollars) Between Patients With MIBC With vs Without Recurrence

	Mean monthly cost (PPPM)		Cost difference [recurrence cohort - nonrecurrence cohort]
	Patients with recurrence (n = 503)	Patients without recurrence (n = 602)	Unadjusted	P value	Adjusted a	P value
All-cause health care costs
Total costs	10,324 [16,420]	3,662 [7,427]	6,661	<0.001 b	7,026	<0.001 b
Medical costs	10,030 [16,377]	3,343 [7,243]	6,686	<0.001 b	7,191	<0.001 b
Inpatient costs	6,654 [15,889]	2,102 [6,419]	4,553	<0.001 b	4,542	<0.001 b
Emergency department costs	133 [386]	48 [186]	85	<0.001 b	84	<0.001 b
Outpatient costs	2,095 [3,101]	412 [721]	1,683	<0.001 b	1,610	<0.001 b
Skilled nursing facility costs	351 [1,152]	343 [1,672]	8	0.917	11	0.889
Other costs c	796 [1,126]	439 [742]	358	<0.001 b	359	<0.001 b
Medicare Part D pharmacy costs	294 [644]	319 [1,504]	−25	0.372	18	0.543
BC-related health care costs
Total costs	3,649 [6,344]	766 [2,802]	2,883	<0.001 b	3,094	<0.001 b
Medical costs	3,633 [6,331]	765 [2,802]	2,868	<0.001 b	3,073	<0.001 b
Inpatient costs	2,258 [5,759]	377 [2,119]	1,881	<0.001 b	2,171	<0.001 b
Emergency department costs	42 [295]	8 [144]	34	<0.001 b	46	<0.001 b
Outpatient costs	706 [1,397]	88 [375]	618	<0.001 b	546	<0.001 b
Skilled nursing facility costs	135 [691]	137 [1,334]	−2	0.962	46	0.527
Other costs c	493 [982]	156 [520]	337	<0.001 b	306	<0.001 b
Medicare Part D pharmacy costs	16 [352]	0 [4]	16	0.009 b	7	0.072

Data are presented as mean [SD].

^a Adjusted analyses controlled for age at diagnosis, time from initial MIBC diagnosis to radical cystectomy, sex, race, ethnicity, region, TNM stage at diagnosis, hypertension during baseline, year of index, Charlson Comorbidity Index score during baseline, any proxy for cisplatin contraindications, any inpatient admissions during baseline, and any emergency department visits during baseline.

^b Statistically significant.

^c The “Other” cost category includes hospice care, home health agency, and durable medical equipment costs.

BC = bladder cancer; MIBC = muscle-invasive bladder cancer; PPPM = per patient per month.

Discussion

This study provides real-world evidence that patients with surgically resected MIBC experience substantial HCRU and costs, with an average of 3.5 all-cause inpatient admissions, 1.0 ED all-cause visits, and 25.8 all-cause outpatient visits PPPY and all-cause total medical costs of $11,250 PPPM. In addition, it demonstrates that patients with recurrence have a significantly higher economic burden than those without recurrence. Patients with recurrence had higher rates of all-cause HCRU, including a 2.4 times higher rate of inpatient admissions, 2.7 times higher rate of ED visits, and 2.0 times higher rate of outpatient visits. Total medical costs were $7,191 PPPM higher for the recurrence cohort than the nonrecurrence cohort, primarily driven by inpatient admission costs. Although the general economic burden of recurrence in MIBC is recognized, there are limited available real-world data on economic burden of recurrence compared with no recurrence among US patients with surgically resected MIBC; therefore, these results help benchmark outcomes for this population. ¹⁴

The findings from this study on the overall burden of MIBC after RC are largely aligned with previous studies assessing real-world postcystectomy economic burden among patients with MIBC. Shi et al described real-world HCRU and health care costs for patients with MIBC, overall and among those who had a cystectomy. ¹⁰ In the MIBC population with cystectomy (mean age 63 years), mean all-cause total health care costs PPPY were $136,104, with 76% of total health care costs attributable to BC ($103,152). This is largely consistent with the all-cause total medical costs of $11,250 PPPM (extrapolated to $135,000 PPPY) and BC-related total medical costs of $7,618 PPPM (extrapolated to $91,416 PPPY) in our study in the overall RC-treated MIBC population. Similarly, a study by Williams et al examined costs among patients with BC that underwent RC (mean age 70 years) and found an overall median cost of $103,579 at year 1. ⁹ Whereas previous studies primarily focused on the general adult MIBC population, this SEER-Medicare study provides additional insights specifically for patients aged 65 and older with Medicare insurance coverage.

Although the overall economic burden of MIBC after RC has been examined, few prior studies have compared HCRU and health care costs between patients with MIBC with and without recurrence. In Clark et al, an economic modeling study, treatment and resource costs increased as the disease recurred or progressed. ¹⁵ For patients expected to experience a recurrence, the estimated average cost per patient with MIBC was $56,247, which is higher than the approximately $35,000 to $37,000 average cost per patient with newly diagnosed MIBC. Avritscher et al estimated costs for 61 patients with MIBC in Texas and found that the mean cost for treatment of recurrences was $36,101. ¹⁶ In our study, the additional costs for those in the recurrence cohort were $7,191 PPPM for all-cause total medical expenses. Although these studies provide different estimates, the different methodologies (modeling estimates vs SEER-Medicare data) as well as different cost components described (treatment costs vs all-cause medical costs) likely drive this difference. Further research comparing patients with recurrence with patients without recurrence would be beneficial.

Although not directly studied in this study, the higher burden of health care visits and health care costs among patients with MIBC with recurrence compared with patients without recurrence may be due to various factors. For example, the index date was set to be 30 days before the recurrence diagnosis, so procedures, tests, and visits associated with the recurrence diagnosis were captured in the described HCRU and costs. Further, patients who experience a recurrence may initiate additional treatments that result in additional visits, medication costs, and monitoring compared with patients who do not experience a recurrence. ^{14 , 17} For example, in this study we see approximately double the rate of inpatient visits, ED visits, and outpatient visits among patients with recurrence compared with patients without recurrence. In addition, historically, there has been poor prognosis for patients with recurrent disease and costs are known to increase toward the end of life, which may have contributed to patterns observed in this study. ^{17 – 19} Lastly, a large portion of the total costs were not BC related for patients with recurrence and patients without recurrence, with only 35% of total costs for patients with recurrence and 21% of total costs for patients without recurrence being BC related. Although some of this may be related to how diagnosis codes are recorded, this may also indicate that patients have a large burden of other health care concerns in this population. Further studies examining the specific cost drivers for patients with recurrent disease would be insightful.

This study had several important strengths. To the best of our knowledge, this retrospective observational study is one of the first to quantify HCRU and health care costs associated with recurrence compared with no recurrence among patients with surgically resected MIBC in a population-level real-world dataset. Second, the linked SEER-Medicare data used in this study provide detailed clinical and demographic information for all the enrolled patients, which allowed adjustments of potential baseline confounders when comparing recurrence vs nonrecurrence cohorts. Third, the algorithm used to identify disease recurrence in the claims data yielded a rate of recurrence similar to the reported recurrence rate in this population, suggesting its validity. Fourth, the patient population included in this study represented diverse disease stages and treatment histories.

LIMITATIONS

It is also important to acknowledge the limitations of this study. First, owing to the nature of administrative claims data, no codes were available to identify MIBC recurrence directly and coding inaccuracies may have led to misclassification bias or misidentification of patients with MIBC recurrence. However, we implemented a comprehensive algorithm that incorporated procedure codes, diagnosis codes, and drug codes. This approach was further strengthened through clinical consultation to ensure accurate assessment of recurrence. In addition, the majority of patients with recurrence were identified based on diagnosis-based indicators. Second, the patient population in the linked SEER-Medicare database only included Medicare patients who were aged 65 years or older ²⁰ ; therefore, the results from this study may not be generalizable to a younger population of patients with MIBC or those with different insurance types or status. On the other hand, as more than 70% of patients with BC in the United States were diagnosed at age 65 years or older (with a median age at diagnosis of 73 years), ²¹ patients in the SEER-Medicare databases are likely to be representative of the general population with MIBC in the United States. Third, although adjusted analyses based on key patient characteristics were considered in this analysis whenever feasible, the results may be confounded by unmeasured characteristics; caution must be used when considering causal inference from this study. Fourth, the median follow-up time from the time of recurrence or matched index date was 1.0 years for patients with recurrence and 1.6 years for those without recurrence. Although these follow-up times are not extensive, they are sufficient to capture a holistic picture of economic burden in the immediate period following disease recurrence.

Although the linked SEER-Medicare database provides detailed clinical and demographic information together with enriched HCRU data for all the enrolled patients, it lacks data on patients who do not have Medicare coverage. As the real-world data on MIBC adjuvant therapy continue to accumulate, future studies with larger sample sizes, longer follow-up periods, and the inclusion of new treatments could extend the learning from the current analysis.

This study provides insights for payers, providers, and health systems into the substantial economic burden faced by patients with MIBC after RC. This includes high rates of inpatient, ED, and outpatient care as well as substantial costs in this population. Further, patients with MIBC after RC who experience a recurrence have a higher economic burden than patients without recurrence. These findings suggest that finding ways to prevent or delay recurrence, such as through early detection and the development of novel treatments, may help reduce the economic burden associated with MIBC. Future studies examining the impact of interventions to reduce recurrence on economic outcomes would be valuable to payers and providers.

Conclusions

This retrospective, US population–based study found that patients with surgically resected MIBC have considerable HCRU and costs. Additionally, those who experienced recurrence had substantial all-cause and BC-related HCRU and health care costs following surgery. These findings highlight the importance of early detection before progression into MIBC and developing more effective treatments in preventing or delaying recurrence following surgical resection of MIBC to alleviate the economic burden on patients and the health care system.

Disclosures

Patrick Squires, Aljosja Rogiers, Haojie Li, and Ching-Yu Wang report employment with and ownership interest in Merck & Co., Inc. Erin E. Cook, Yan Song, Anya (Xinyi) Jiang, Adina Zhang, and Shravanthi M. Seshasayee are employees of Analysis Group, Inc., a consulting company that provided paid consulting services to Merck & Co. Ronac Mamtani reports research grants or contracts to his institution from Astellas and MSD and consulting fees from Astellas, Bristol Myers Squibb, Merck & Co, Inc, and Seagen outside the submitted work. This study was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.