Impact of Intravenous Iron Replacement Therapy on Healthcare Costs for Patients with Iron Deficiency Anemia in the USA: A Retrospective Analysis

Abstract

Background

Intravenous iron replacement therapy for the treatment of iron deficiency anemia is often required in patients with chronic diseases including cancer, heart failure, or chronic kidney disease.

Objective

We aimed to compare healthcare resource utilization and costs for patients treated with intravenous ferric carboxymaltose (FCM) or low-dose iron for iron deficiency anemia.

Methods

This analysis of Optum Research Database administrative claims data included patients with iron deficiency anemia who received intravenous iron from 2017 to 2019 and had diagnoses of cancer, heart failure, or chronic kidney disease. Patients were continuously enrolled for 6-month baseline and 12-month follow-up periods. Follow-up all-cause total costs for FCM and low-dose iron cohorts were compared using generalized linear models; inpatient costs were estimated with two-part models to account for patients without hospitalizations. Models were adjusted for age, sex, geographic region, insurance type, index year, baseline comorbidity scores, and healthcare costs. Sensitivity analyses compared FCM with an iron sucrose subgroup.

Results

For patients with cancer (n = 10,763), mean adjusted all-cause total costs were numerically lower for FCM than low-dose iron by $2369 (cost ratio [CR] 0.97, P = 0.182) and significantly lower for FCM than iron sucrose by $6712 (CR 0.93, P < 0.001). For heart failure (n = 8337), the mean all-cause total cost was numerically lower for FCM than low-dose iron by $2022 (CR 0.97, P = 0.198) and significantly lower for FCM than iron sucrose by $3892 (CR 0.95, P = 0.024). For chronic kidney disease (n = 10,617), the mean all-cause total cost was statistically significantly lower for FCM than low-dose iron by $3623 (CR 0.94, P = 0.006) and iron sucrose by $4161 (CR 0.93, P = 0.004). For all groups, the FCM and low-dose iron cohorts differed in both the odds of having any inpatient costs and the level of inpatient cost (cancer: odds ratio 0.79, P < 0.001; CR 0.88, P < 0.001; heart failure: odds ratio 0.76, P < 0.001; CR 0.89, P < 0.001; chronic kidney disease: odds ratio 0.75, P < 0.001; CR 0.84, P < 0.001). Inpatient cost results were consistent for iron sucrose.

Conclusions

Despite the typically higher drug acquisition cost of FCM versus low-dose intravenous iron, the price differential was offset by the lower inpatient cost incurred in the FCM cohort in each patient population. These findings suggest the potential economic benefit of FCM to reduce inpatient utilization and associated costs to patients and health plans compared with low-dose intravenous iron.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40801-025-00496-9.

Plain Language Summary

This study compared the use of healthcare services and healthcare costs for patients in the USA treated with different types of intravenous iron replacement medicines. Patients in the study had cancer, heart failure, or chronic kidney disease, and they also had iron deficiency anemia. Patients treated with a high-dose intravenous iron replacement, ferric carboxymaltose, had healthcare costs that were lower than costs for patients treated with low-dose intravenous iron replacement because of lower hospitalization costs. Patients treated with ferric carboxymaltose were both less likely to be admitted to the hospital and had lower costs when they were hospitalized.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40801-025-00496-9.

Key Points

For patients with chronic kidney disease, cancer, and heart failure, patients treated for iron deficiency anemia with the intravenous iron replacement treatment ferric carboxymaltose had lower inpatient utilization and costs, suggesting better clinical outcomes compared with patients treated with the low-dose intravenous iron replacement.

The study found that ferric carboxymaltose can reduce the total cost of care relative to iron sucrose, the most used low-dose intravenous iron product, because of fewer inpatient stays, fewer inpatient days, and a lower blood transfusion rate.

Introduction

Iron deficiency anemia (IDA) affects approximately 1% of men and 5% of women in the USA [1] and is most commonly caused by gastrointestinal and menstrual blood loss [2]. However, IDA also occurs in cancers and many chronic inflammatory conditions, affecting an estimated 44% of patients with hematologic malignancies [3], 50% of patients with solid tumors [3], 50% of patients with heart failure (HF) [4], and 58–73% of patients with non-dialysis-dependent chronic kidney disease (CKD) [5]. Inflammatory mechanisms that contribute to anemia of chronic disease are exacerbated by poor nutrition, decreased gastrointestinal absorption of iron, and blood loss (e.g., bleeding tumors, frequent blood sampling, dialysis) [2, 6]. In cancer, myelosuppressive treatment may result in anemia, and hematologic cancers may also serve as an underlying cause of anemia [7]; use of erythropoiesis-stimulating agents to treat anemia may result in depletion of iron stores and corresponding IDA [8] requiring supplementation with intravenous (IV) iron [9].

Untreated IDA poses a substantial cost to the US healthcare system [2]. Iron deficiency anemia may be overlooked and untreated because IDA symptoms (e.g., fatigue and weakness) are also common symptoms of cancer, HF, and CKD [2]. Untreated IDA can lead to emergency hospitalization and multiple blood transfusions in patients with chronic diseases [2]. In addition, IDA may impair responses to cancer treatment and reduce overall survival [10] and is known to worsen exercise capacity for patients with HF [11].

Intravenous iron therapy is the preferred route of administration for iron replacement in patients with cancer [12], HF [4], and CKD [13] because it is associated with rapid correction of hemoglobin levels, fewer adverse events, and better adherence than oral administration [2]. Several IV iron preparations are available in the USA, including high- and low-dose formulations. Ferric carboxymaltose (FCM; Injectafer^®) is a high-dose formulation administered as a single administration of a weight-based dose of up to 1000 mg, or as two infusions at least 7 days apart with a cumulative dose of up to 1500 mg depending on patient weight and hemoglobin levels [2, 14, 15]. Low-dose formulations require multiple weekly or daily administrations to replenish iron stores with total cumulative dosages typically 1000–1200 mg, including iron sucrose (Venofer^®; maximal single dose 200 mg, five administrations), iron dextran (INFeD^®; maximal single dose 100 mg, 12 administrations), and sodium ferric gluconate complex in sucrose (Ferrlecit^®; maximal single dose 125 mg, eight administrations) [15–17]. Product selection for IV iron therapy ultimately depends on patient/physician preferences and cost considerations [2].

The number of total treatment courses for IDA varies depending on the severity of IDA and the individual treatment response; however, for cancer, HF, and CKD, the overall cumulative course of IDA treatment appears to align with the chronic nature of these conditions. In patients with cancer, the median time to resolution of iron deficiency has been found to be 1.9 years, with 58% of patients not achieving resolution of IDA by 3 years [19]. Patients with HF may require repeated treatment courses over several years, depending on their iron repletion sustainability and ongoing blood loss or inflammation, though precise “lifetime” numbers vary widely [20]. Among patients with CKD, the exact frequency is tailored to each patient’s needs, blood losses, and responses to therapy, but guidelines recommend regular monitoring of iron measures and redosing of IV iron to achieve target levels [21].

Clinical benefits of IV iron replacement therapy with FCM have been demonstrated, such as improved exercise capacity, functional status, and quality of life in patients with chronic HF [22–26]. Existing real-word studies by Kwong et al. [28] utilized a database that consisted of a mostly commercially insured population, and found higher costs among patients with low-dose IV iron. However, this previous study limited the FCM sample to patients with at least two administrations, so was not fully representative of real-world treatment patterns. There are otherwise limited real-world data on the economic value of IV iron treatment in these populations in the USA, and a need to replicate findings and understand how results may vary in a population representing both commercial and Medicare (older) patients [28, 29]. The primary objective of this study was to compare healthcare resource utilization, including blood transfusions and healthcare visits, as well as healthcare costs for patients with IDA treated with IV FCM or low-dose iron therapy.

Methods

Study Design and Patients

This retrospective analysis used the Optum Research Database (ORD), a large, national, US medical and pharmacy claims database. The ORD includes data from approximately 8% of US commercial health plan enrollees and 18% of Medicare Advantage enrollees. Medical claims in the ORD include diagnosis and procedure codes, site of service codes, and paid amounts among other data. Pharmacy claims include information such as drug name, National Drug Code, dosage form, drug strength, fill date, and financial information. Patients with an IV iron treatment between 1 January, 2017 and 31 December, 2019, and continuous enrollment for an 18-month observation period were included. Demographics and clinical characteristics were assessed during a 6-month baseline period before the index date (defined as the first medical claim for IV iron treatment). Healthcare resource utilization, healthcare costs, and IDA-related treatment patterns were assessed during the baseline period and a 12-month follow-up period (including the index date). Iron deficiency anemia-related costs were assessed only during the 12-month follow-up period. The analysis included patients if death limited the follow-up period to < 12 months. Variables including the diagnosis code of IDA, cancer, HF, and CKD were assessed at any point during the 6-month baseline or 12-month follow-up period because patients may have delayed or infrequent appointments for these conditions.

Cancer, HF, and CKD were selected as diseases for this study as part of the aim was to perform a sensitivity analysis of the previous study [28], and because they are labeled indications for FCM, with patients with cancer having a high likelihood of IDA. Eligible patients had a diagnosis of cancer, HF, or CKD as well as IDA based on one of more medical claims with a diagnosis of IDA in the baseline or follow-up period. A diagnosis of cancer, HF, or CKD was defined as two or more medical claims with the same diagnosis ≥ 30 days apart within the 18-month observation period. The database is de-identified and patient confidentiality was maintained for this analysis.

The study excluded patients with one or more medical claims in the baseline period for CKD stage 5, end-stage renal disease, or dialysis for consistency with the FCM label indications. Patients were also excluded if they had unknown sex, or evidence of any IV iron therapy, including high-dose (ferric carboxymaltose) iron therapy, low-dose IV iron therapy (i.e., iron dextran, ferric gluconate, iron sucrose), ferric derisomaltose, ferric pyrophosphate citrate, or ferumoxytolin in the baseline period. Applying this criterion meant that patients were treatment naïve to IV iron therapy as of the index date (based on the evidence in the baseline period).

Treatment Groups, Healthcare Resource Utilization, and Costs

Patients were classified into two primary IV iron treatment cohorts based on the index treatment date: FCM or low-dose iron (e.g., iron sucrose, iron dextran, or sodium ferric gluconate complex in sucrose). In addition, an iron sucrose subgroup was created because iron sucrose was the most used low-dose IV iron treatment, accounting for 62.4% of the cancer subpopulation, 69.8% of the HF subpopulation, and 73.5% of the CKD subpopulation. Patients were retained in their initial treatment cohort (intent-to-treat analysis), regardless of any switching (not assessed) or additional treatment (such as oral iron treatment) over the 12 month follow-up. Outcomes included all-cause total costs, comprising ambulatory care costs (including both office and hospital outpatient visits), emergency department (ED) costs, inpatient costs, other medical costs, and pharmacy costs. The study included measurement of non-IV iron-related ambulatory costs as an assessment of potentially disease-related costs (for cancer, HF, and CKD) rather than costs that might be associated with the IV iron treatment itself. Costs were calculated as the combined total costs paid by both the plan and patient, and Consumer Price Index adjusted to year 2021 [30]. Additional utilization was measured over the baseline and follow-up for the use of blood transfusions and the number of days with blood transfusions.

Statistical Analysis

Baseline demographics, clinical characteristics, and resource utilization outcome measures were analyzed descriptively within each patient disease population (cancer, HF, or CKD), including counts and percentages for categorical variables and means, medians, and standard deviations (SDs) for continuous variables. Within each disease population, bivariate comparisons for IV iron treatment cohorts (FCM vs low-dose iron and FCM vs iron sucrose) were provided using appropriate tests (e.g., t test, Chi-square depending on the distribution of each measure).

Multivariable analyses were used to control for confounding in the association between IV iron treatment cohorts and key outcomes. The three outcomes were: (1) all-cause total costs; (2) all-cause inpatient costs; and (3) non-IV iron-related ambulatory costs. Separate multivariable models compared the three outcome variables of interest between two treatment cohorts (FCM vs low-dose iron and FCM vs iron sucrose as a sensitivity analysis) in each of the three patient populations (i.e., cancer, HF, CKD). Total cost and non-IDA-related ambulatory cost comparisons were assessed using a generalized linear model (GLM) with a Gamma distribution and log link to account for the typical skewed distribution of healthcare cost variables [31, 32]. The exponentiated coefficients from the GLM represent estimated cost ratios (e.g., the ratio of costs for the FCM vs the low-dose iron cohort). Cost ratios with 95% confidence intervals (CIs) are presented for the treatment cohort variable in the final model (adjusted for other covariates). For ease of interpretation and comparison with the bivariate results, the adjusted average cost was predicted for each cohort.

For inpatient cost comparisons, two-part regression models were estimated to handle the greater percentage of patients with no inpatient cost and to account for the skewed distribution of cost variables. The first part of the two-part model used a logistic regression model to estimate the probability that a patient would have inpatient costs (i.e., hospital admission), while the second part consisted of a GLM model with a log link and Gamma distribution to estimate inpatient costs for those with non-zero inpatient costs [34]. Odds ratios (ORs) with 95% CIs are presented for the cohort variable in the final adjusted logistic model that estimated the probability of non-zero costs. Cost ratios with 95% CIs are presented for the cohort variable in the GLM model estimating the level of costs. For ease of interpretation and comparison with the bivariate results, we calculated the expected inpatient cost for each cohort by multiplying the probability of accruing any inpatient costs based on the logistic regression model by the conditional expected cost (given that the patient has cost > $0) produced by the GLM model.

For each model, predictors were selected based upon clinical rationale and significance. Regression diagnostics were performed for each model to assess goodness of fit and violations of model assumptions (e.g., multicollinearity). Predictor variables were age (continuous); sex (male vs female); geographic region (Midwest, Northeast, South, West); insurance type (commercial vs Medicare Advantage); year of treatment index date (2017, 2018, 2019); National Cancer Institute comorbidity score (for the cancer population) [35] or Quan-Charlson comorbidity score (for HF and CKD populations) (0, 1–2, 3–4, ≥ 5) [36]; and baseline inpatient stay (no vs yes). Analyses were conducted using SAS Release 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Analysis Sample

Between 1 January, 2017 and 31 December, 2019, 104,667 patients with one or more claims for IV iron therapy were identified. The final analysis sample included 21,374 patients with IDA who met eligibility criteria (Fig. 1 of the Electronic Supplementary Material [ESM]). Based on the requirement for two or more claims with the relevant diagnostic code ≥ 30 days apart within the observation period, the three patient groups for analysis included patients with cancer (n = 10,763), patients with HF (n = 8,337), and patients with CKD (n = 10,617).

Baseline Demographics and Clinical Characteristics

Overall, patients were predominantly white, non-Hispanic female individuals with a mean age in the FCM cohort of 70.6 years (cancer), 74.4 years (HF), and 74.1 years (CKD) who resided in the South and were insured by Medicare Advantage plans (in FCM: cancer, 79.4%; HF, 91.8%; CKD, 89.7%) (Table 1). In each patient population, demographics and clinical characteristics were generally similar across FCM, low-dose iron, and iron sucrose cohorts. However, for patients with cancer, the mean National Cancer Institute comorbidity score was lower for the FCM cohort, and for patients with HF or CKD, the mean Quan-Charlson comorbidity scores were higher for the FCM cohorts.

Table 1.

Baseline demographics and clinical characteristics^a

Characteristic	Patients with cancer (n = 10,763)			Patients with heart failure (n = 8337)			Patients with chronic kidney disease (n = 10,617)
	FCM (n = 5892)	Low-dose ironb (n = 4871)	Iron sucroseb (n = 3040)	FCM(n =3364)	Low-dose ironb (n = 4973)	Iron sucroseb (n = 3471)	FCM (n = 4377)	Low-dose ironb (n = 6240)	Iron sucroseb (n = 4584)
Age, years, mean	70.6	70.9 P = 0.323	71.0 P = 0.201	74.4	75.0 P = 0.012	75.2 P < 0.001	74.1	74.4 P = 0.106	74.5 P = 0.030
Age group, years, n (%)
18–39	114 (1.9)	98 (2.0) P = 0.775	69 (2.3) P = 0.290	18 (0.5)	19 (0.4) P = 0.302	14 (0.4) P = 0.425	34 (0.8)	44 (0.7) P = 0.670	29 (0.6) P = 0.414
40–64	1256 (21.3)	1061 (21.8) P = 0.559	656 (21.6) P = 0.775	461 (13.7)	677 (13.6) P = 0.906	469 (13.5) P = 0.817	565 (12.9)	824 (13.2) P = 0.655	600 (13.1) P = 0.799
65–79	3209 (54.5)	2493 (51.2) P < 0.001	1499 (49.3) P < 0.001	1757 (52.2)	2444 (49.1) P = 0.006	1658 (47.8) P < 0.001	2378 (54.3)	3243 (52.0) P = 0.017	2356 (51.4) P = 0.005
≥ 80	1313 (22.3)	1219 (25.0) P < 0.001	816 (26.8) P < 0.001	1128 (33.5)	1833 (36.9) P = 0.002	1330 (38.3) P < 0.001	1400 (32.0)	2129 (34.1) P = 0.022	1599 (34.9) P = 0.004
Sexc, n (%)
Female	3501 (59.4)	2792 (57.3) P = 0.028	1724 (56.7) P = 0.014	1988 (59.1)	2925 (58.8) P = 0.800	2034 (58.6) P = 0.677	2624 (60.0)	3766 (60.4) P = 0.676	2738 (59.7) P = 0.832
Male	2391 (40.6)	2079 (42.7) P = 0.028	1316 (43.3) P = 0.014	1376 (40.9)	2048 (41.2) P = 0.800	1437 (41.4) P = 0.677	1753 (40.1)	2474 (39.6) P = 0.676	1846 (40.3) P = 0.832
Race/ethnicityc, n (%)
Hispanic	343 (5.8)	578 (11.9) P < 0.001	380 (12.5) P < 0.001	164 (4.9)	461 (9.3) P < 0.001	342 (9.9) P < 0.001	244 (5.6)	696 (11.2) P < 0.001	526 (11.5) P < 0.001
Non-Hispanic Asian	100 (1.7)	128 (2.6) P < 0.001	88 (2.9) P < 0.001	42 (1.2)	79 (1.6) P = 0.203	53 (1.5) P = 0.326	67 (1.5)	127 (2.0) P = 0.056	100 (2.2) 0.023
Non-Hispanic Black	1118 (19.0)	771 (15.8) P < 0.001	451 (14.8) P < 0.001	632 (18.8)	970 (19.5) P = 0.414	650 (18.7) P = 0.949	894 (20.4)	1,334 (21.4) P = 0.235	971 (21.2) P = 0.377
Non-Hispanic White	4076 (69.2)	3216 (66.0) P < 0.001	2007 (66.0) P = 0.002	2394 (71.2)	3249 (65.3) P < 0.001	2286 (65.9) P < 0.001	3018 (69.0)	3858 (61.8) P < 0.001	2825 (61.6) P < 0.001
Region, n (%)
Midwest	1507 (25.6)	1300 (26.7) P = 0.191	991 (32.6) P < 0.001	955 (28.4)	1731 (34.8) P < 0.001	1421 (40.9) P < 0.001	1183 (27.0)	1899 (30.4) P < 0.001	1577 (34.4) P < 0.001
Northeast	628 (10.7)	775 (15.9) P < 0.001	589 (19.4) P < 0.001	365 (10.9)	639 (12.9) P = 0.006	476 (13.7) P < 0.001	458 (10.5)	693 (11.1) 0.295	535 (11.7) 0.069
South	3352 (56.9)	2350 (48.2) P < 0.001	1171 (38.5) P < 0.001	1876 (55.8)	2227 (44.8) P < 0.001	1298 (37.4) P < 0.001	2477 (56.6)	3123 (50.0) P < 0.001	2059 (44.9) P < 0.001
West	405 (6.9)	446 (9.2) P < 0.001	289 (9.5) P < 0.001	168 (5.0)	376 (7.6) P < 0.001	276 (8.0) P < 0.001	259 (5.9)	525 (8.4) P < 0.001	413 (9.0) P < 0.001
Insurance coverage, n (%)
Commercial	1214 (20.6)	1108 (22.7) P = 0.007	708 (23.3) P = 0.003	275 (8.2)	424 (8.5) 0.570	294 (8.5) 0.659	450 (10.3)	622 (10.0) P = 0.598	449 (9.8) P = 0.444
Medicare advantage	4678 (79.4)	3763 (77.3) P = 0.007	2332 (76.7) P = 0.003	3089 (91.8)	4549 (91.5) P = 0.570	3177 (91.5) P = 0.659	3927 (89.7)	5618 (90.0) P = 0.598	4135 (90.2) P = 0.444
NCI or Quan-Charlson comorbidity score, mean (SD)	2.4 (2.1)	2.6 (2.3) P < 0.001	2.7 (2.3) P < 0.001	4.5 (2.4)	4.3 (2.3) P < 0.001	4.3 (2.2) P < 0.001	3.9 (2.4)	3.8 (2.3) P = 0.050	3.7 (2.3) P = 0.004
Year of index date, n (%)
2017	1556 (26.4)	1839 (37.8) P < 0.001	983 (32.3) P < 0.001	802 (23.8)	1665 (33.5) P < 0.001	1078 (31.1) P < 0.001	1098 (25.1)	2173 (34.8) P < 0.001	1503 (32.8) P < 0.001
2018	2065 (35.0)	1467 (30.1) P < 0.001	1006 (33.1) P = 0.065	1171 (34.8)	1564 (31.5) P = 0.001	1161 (33.5) P = 0.235	1488 (34.0)	1962 (31.4) P = 0.006	1525 (33.3) P = 0.466
2019	2271 (38.5)	1565 (32.1) P < 0.001	1051 (34.6) P < 0.001	1391 (41.3)	1744 (35.1) P < 0.001	1232 (35.5) P < 0.001	1791 (40.9)	2105 (33.7) P < 0.001	1556 (33.9) P < 0.001
6-Month baseline all-cause costs,d USD, mean (SD) [median]
Total	39,775 (50,531) [22,868]	37,466 (54,741) [19,550]	39,104 (58,981) [19,953]	34,080 (48,076) [20,702]	34,260 (56,197) [19,602]	34,363 (58,222) [19,397]	27,043 (38,392) [14,424]	29,190 (53,990) [14,435]	29,334 (57,178) [14,066]
Inpatient stays	11,292 (25,396) [0]	12,667 (32,930) [0]	13,370 (36,547) [0]	14,944 (31,983) [1569]	17,248 (43,864) [3034]	17,581 (46,177) [3557]	10,052 (24,569) [0]	13,670 (42,394) [0]	13,883 (44,702) [0]
ED visits	866 (2048) [0]	1272 (4033) [209]	1209 (2375) [206]	1432 (2694) [497]	2176 (4577) [917]	2029 (3072) [883]	1010 (2096) [182]	1634 (3161) [451]	1545 (2977) [370]
Ambulatory visits	19,419 (33,913) [7877]	16,139 (32,649) [5956]	17,133 (36,760) [6037]	9680 (16,247) [5252]	8171 (18,172) [4040]	8065 (18,827) [3966]	8874 (16,178) [4135]	7518 (16,352) [3571]	7506 (17,286) [3447]
Pharmacy	6353 (17,001) [1359]	5628 (15,383) [1223]	5677 (15,767) [1193]	6260 (14,810) [2605]	4892 (12,389) [2261]	4876 (11,779) [2252]	5855 (15,516) [2195]	4866 (13,177) [2028]	4838 (13,197) [2065]
Other medical	1845 (7347) [436]	1760 (10,530) [405]	1714 (6391) [382]	1763 (13,232) [485]	1773 (14,009) [477]	1812 (12,810) [478]	1254 (6500) [363]	1501 (11,293) [364]	1561 (11,695) [357]

ED emergency department, FCM ferric carboxymaltose, NCI National Cancer Institute, SD standard deviation, USD US dollars

^aData are n (%) unless otherwise noted

^bP-values are for comparison with the FCM cohort in the corresponding patient population and based on two-sample t tests for continuous measures and Pearson chi-square tests for binary measures

^cSex (male or female) and race/ethnicity were based on information linked to and included in the Optum Research Database; patients with undefined sex were removed from the study sample

^dCosts were adjusted using the annual medical care component of the Consumer Price Index to reflect inflation between the date of the claim and 2021 [30]¹⁵

Patients with Cancer

For patients with cancer (n = 10,763), 5892 (54.7%) were in the FCM cohort, 4871 (45.3%) in the low-dose IV iron cohort, and 3040 (28.2%) in the iron sucrose cohort. During the follow-up period, < 1% of patients with cancer received oral iron therapy. In the FCM cohort, 77.8% received one or more additional FCM doses within 21 days of the index dose. The most common type of IV iron therapy initiated (including patients who switched from index to other IV iron treatment) was FCM (57.2%), followed by iron sucrose (29.7%), iron dextran (11.7%), and ferric gluconate (6.4%).

During the 6-month baseline period, the percentages of patients with all-cause ED and inpatient visits and the mean number of all-cause ED visits, inpatient visits, and inpatient days were each significantly lower for the FCM cohort than for either low-dose iron or iron sucrose cohorts, whereas the mean number of ambulatory visits was higher for the FCM cohort compared with other cohorts (P < 0.001 for all comparisons) [Table 1 of the ESM]. A similar pattern was observed during the 12-month follow-up period for all-cause and IDA-related visits (Table 1 of the ESM).

During the baseline period, mean (SD) unadjusted all-cause total healthcare costs in patients with cancer were higher for the FCM cohort compared with the low-dose iron cohort ($39,775 [$50,531] vs $37,466 [$54,741]) (Table 1). Mean all-cause ambulatory care visit, pharmacy, and other medical costs were higher and ED and inpatient costs were lower for the FCM cohort than the low-dose iron cohort. A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Over the follow-up, 22.2% of FCM, 32.4% of low-dose, and 34.2% of iron sucrose patients received blood transfusions (P < 0.001) [Table 4 of the ESM]. The FCM cohort had a significantly lower mean (SD) number of blood transfusion days (15.5 [23.2]) compared with low-dose iron (18.5 [26.2]; P < 0.001) and iron sucrose cohorts (19.0 [27.8]; P < 0.001).

During the follow-up period, unadjusted mean (SD) all-cause total costs were $82,525 ($104,616) for the FCM cohort and $83,723 ($115,304) for the low-dose iron cohort (Fig. 1). Of note, unadjusted mean (SD) inpatient costs were lower for the FCM cohort than the low-dose iron cohort ($19,929 [$45,633] vs $27,119 [$61,134]), whereas ambulatory costs were higher for the FCM cohort than the low-dose iron cohort ($44,005 [$74,499] vs $38,857 [$77,986]). A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Fig. 1.

Unadjusted follow-up all-cause healthcare costs^a. CKD chronic kidney disease, ED emergency department, FCM ferric carboxymaltose, HF heart failure, USD US dollars. ^aTotals were calculated prior to rounding for subcategories

Multivariable adjusted healthcare costs in the follow-up period for patients with cancer are presented in Table 2 and Fig. 2A. After controlling for covariates, the mean adjusted all-cause total cost was numerically but not statistically significantly lower for the FCM cohort than the low-dose iron cohort (cost ratio, 0.97 [95% CI 0.93–1.01]; P = 0.182; $82,425 vs $84,794), and statistically significantly lower for the FCM than the iron sucrose cohort (cost ratio, 0.93 [95% CI 0.88–0.97]; P < 0.001; $83,028 vs $89,740).

Table 2.

Adjusted cost ratios comparing intravenous ferric carboxymaltose with low-dose iron and iron sucrose in patients with cancer, HF, and CKD during the follow-up period

	Cancer population (n = 10,763)		HF population (n = 8337)		CKD population (n = 10,617)
	Cost ratio (95% CI)	Odds ratio (95% CI)	Cost ratio (95% CI)	Odds ratio (95% CI)	Cost ratio (95% CI)	Odds ratio (95% CI)
FCM vs low-dose irona
Total cost	0.97 (0.93, 1.01)		0.97 (0.93, 1.02)		0.94 (0.90, 0.98)
Inpatient costb	0.88 (0.82, 0.94)	0.79 (0.73, 0.86)	0.89 (0.83, 0.95)	0.76 (0.70, 0.84)	0.84 (0.79, 0.90)	0.75 (0.70, 0.82)
Non-IV iron ambulatory visit cost	1.01 (0.95, 1.08)		1.04 (0.98, 1.11)		1.04 (0.98, 1.11)
FCM vs iron sucrosea
Total cost	0.93 (0.88, 0.97)		0.95 (0.90, 0.99)		0.93 (0.89, 0.98)
Inpatient costb	0.89 (0.82, 0.95)	0.72 (0.65, 0.79)	0.88 (0.82, 0.94)	0.72 (0.64, 0.80)	0.85 (0.79, 0.91)	0.75 (0.69, 0.82)
Non-IV iron ambulatory visit cost	0.99 (0.92, 1.06)		1.01 (0.94, 1.09)		1.03 (0.96, 1.10)

CI confidence interval, CKD chronic kidney disease, FCM ferric carboxymaltose, HF heart failure, IV intravenous, USD US dollars

^aWholesale acquisition costs per mg (USD): FCM, $1.85; iron dextran, $0.36; iron sucrose, $0.36; sodium ferric gluconate complex in sucrose, $0.51

^bFor inpatient costs, the odds ratio is provided for whether patients had costs > $0 during the follow-up period in addition to the generalized linear model cost ratio for the level of costs among patients with cost > $0 during the follow-up period

Fig. 2.

Adjusted annual healthcare costs during the follow-up period: patients with cancer (A), heart failure [HF] (B), and chronic kidney disease [CKD] (C). FCM ferric carboxymaltose, IV intravenous, USD United States dollars

In the adjusted analysis, the FCM and low-dose iron cohorts differed in both the odds of having any inpatient costs (OR 0.79 [95% CI 0.73–0.86]; P < 0.001) and the level of cost for patients with an inpatient admission (cost ratio, 0.88 [95% CI 0.82–0.94]; P < 0.001); mean adjusted inpatient costs were $5392 (20.7%) lower for the FCM cohort than the low-dose iron cohort ($20,642 vs $26,034); inpatient cost results were similar for the iron sucrose cohort. The adjusted non-IV iron-related ambulatory visit costs were not statistically significantly different for the FCM than the low-dose iron cohort (cost ratio 1.01 [95% CI 0.95–1.08] P = 0.655; $22,263 vs $21,949); results were similar for the comparison of FCM and iron sucrose cohorts.

Patients with HF

For patients with HF (n = 8337), 3364 (40.4%) were in the FCM cohort, 4973 (59.6%) in the low-dose IV iron cohort, and 3471 (41.6%) in the iron sucrose cohort. During the follow-up period, < 0.5% of patients with HF received oral iron therapy. In addition, in the FCM cohort, 79.0% received one or more additional FCM doses within 21 days of the index dose. The most common type of IV iron therapy initiated (included switched treatments) was iron sucrose (44.2%), followed by FCM (43.7%), ferric gluconate (10.8%), and iron dextran (8.6%).

During the 6-month baseline period, the percentages of patients with all-cause ED and inpatient visits and the mean number of all-cause ED visits, inpatient visits, and inpatient days were each significantly lower for the FCM cohort than for either low-dose iron or iron sucrose cohorts, while ambulatory visits were higher for FCM compared with other cohorts (P < 0.001 for all comparisons) [Table 2 of the ESM]. A similar pattern was observed during the 12-month follow-up period for all-cause and IDA-related visits (Table 2 of the ESM).

During the baseline period, mean (SD) unadjusted all-cause total healthcare costs in patients with HF were lower for the FCM cohort compared with the low-dose iron cohort ($34,080 [$48,076] vs $34,260 [$56,197]) (Table 1). Mean all-cause ambulatory care visit and pharmacy costs were greater, whereas ED visit, inpatient, and other medical costs were less for the FCM cohort than the low-dose iron cohort. A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Over the follow-up, 27.0% of FCM patients, 36.6% of low-dose iron patients, and 36.3% of iron sucrose patients received blood transfusions (P < 0.001) [Table 4 of the ESM]. Of note, the FCM cohort had a numerically but not statistically significantly lower mean (SD) number of blood transfusion days (20.1 [29.2]) compared with the low-dose iron cohort (21.9 [32.3]; P = 0.149) and the iron sucrose cohort (22.8 [35.3]; P = 0.054).

During the follow-up period, unadjusted mean (SD) all-cause total costs were $68,594 ($90,495) for FCM and $69,117 ($95,927) for low-dose iron cohorts (Fig. 1). Of note, unadjusted mean (SD) inpatient costs were lower for the FCM cohort than the low-dose iron cohort ($27,838 [$58,328] vs $34,234 [$64,043]), whereas ambulatory costs were higher for the FCM cohort than the low-dose iron cohort ($21,977 [$38,710] vs $18,698 [$51,201]). A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Multivariable adjusted healthcare costs in the follow-up period for patients with HF are presented in Table 2 and Fig. 2B. After controlling for covariates, the mean adjusted all-cause total cost was numerically but not statistically significantly lower for the FCM cohort than the low-dose iron cohort (cost ratio, 0.97 [95% CI 0.93–1.02]; P = 0.198; $67,752 vs $69,774), and statistically significantly lower for the FCM cohort than the iron sucrose cohort (cost ratio, 0.95 [95% CI 0.90–0.99]; P = 0.024; $68,071 vs $71,962).

In the adjusted analysis, the FCM and low-dose iron cohorts differed in both the odds of having any inpatient costs (OR, 0.76 [95% CI 0.70–0.84]; P < 0.001) and the level of cost for patients with an inpatient admission (cost ratio, 0.89 [95% CI 0.83–0.95]; P < 0.001). Mean adjusted inpatient costs were $6458 (18.9%) lower for the FCM than the low-dose iron cohort ($27,786 vs $34,244), and inpatient cost results were similar for FCM and iron sucrose cohorts. The adjusted non-IV iron-related ambulatory visit costs were not statistically significantly different for the FCM than the low-dose iron cohort (cost ratio, 1.04 [95% CI 0.98–1.11] P = 0.235; $11,562 vs $11,115), and results were similar for comparison of the FCM cohort to the iron sucrose cohort.

Patients with CKD

For patients with CKD (n = 10,617), 4377 (41.2%) were in the FCM cohort, 6240 (58.8%) in the low-dose IV iron cohort, and 4584 (43.2%) in the iron sucrose cohort. During the follow-up period, < 0.5% of patients with CKD received oral iron therapy. In the FCM cohort, 80.2% received one or more additional FCM doses within 21 days of the index dose. The most common type of IV iron therapy initiated (including switched treatments) was iron sucrose (45.5%), followed by FCM (44.0%), ferric gluconate (9.2%), and iron dextran (7.7%).

During the 6-month baseline period, the percentages of patients with all-cause ED and inpatient visits and the mean number of all-cause ED visits, inpatient visits, and inpatient days were each significantly lower for the FCM cohort than for either the low-dose iron or iron sucrose cohorts, whereas ambulatory care visits were higher for the FCM cohort than for the other cohorts (P < 0.001 for all comparisons) [Table 3 of the ESM]. A similar pattern was observed during the 12-month follow-up period for all-cause and IDA-related visits (Table 3 of the ESM).

During the baseline period, mean (SD) unadjusted all-cause total healthcare costs for patients with CKD were lower for the FCM cohort than the low-dose iron cohort ($27,043 [$38,392] vs $29,190 [$53,990]) (Table 1). Mean all-cause ambulatory care visit and pharmacy costs were greater, whereas ED visit, inpatient, and other medical costs were lower for the FCM cohort than the low-dose iron cohort. A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Over the follow-up, 22.9% of FCM and 30.7% of low-dose iron patients and 29.6% of iron sucrose patients received blood transfusions (P < 0.001) [Table 4 of the ESM]. The FCM cohort had a significantly lower mean (SD) number of blood transfusion days (18.9 [29.4]) compared with the low-dose iron cohort (21.5 [32.0]; P = 0.026) and the iron sucrose cohort (22.3 [34.9]; P = 0.009).

During the follow-up period, unadjusted mean (SD) all-cause total costs were $56,737 ($78,257) for the FCM cohort and $60,016 ($92,758) for the low-dose iron cohort (Fig. 1). Of note, unadjusted mean (SD) inpatient costs were lower for the FCM cohort than the low-dose iron cohort ($19,504 [$47,558] vs $26,697 [$59,729]), whereas ambulatory costs were higher for the FCM cohort than the low-dose iron cohort ($21,402 [$38,513] vs $18,112 [$47,421]). A similar pattern was observed when comparing FCM and iron sucrose cohorts.

Multivariable adjusted healthcare costs in the follow-up period are shown in Table 2 and Fig. 2C for the CKD population. After controlling for covariates, the mean adjusted all-cause total cost was statistically significantly lower for the FCM cohort than the low-dose iron cohort (cost ratio, 0.94 [95% CI 0.90–0.98]; P = 0.006; $56,334 vs $59,957) and the iron sucrose cohort (cost ratio, 0.93 [95% CI 0.89–0.98]; P = 0.004; $56,321 vs $60,483)s.

In the adjusted analysis, the FCM and low-dose iron cohorts differed in both the odds of having any inpatient costs (OR, 0.75 [95% CI 0.70–0.82]; P < 0.001) and the level of cost for patients with an inpatient admission (cost ratio, 0.84 [95% CI 0.79–0.90]; P < 0.001). Mean adjusted inpatient costs were $6646 (25.2%) lower for the FCM cohort than the low-dose iron cohort ($19,778 vs $26,424), and inpatient cost results were similar for the comparison of the FCM cohort to the iron sucrose cohort. The adjusted non-IV iron-related ambulatory visit costs were not statistically significantly different for FCM and low-dose iron cohorts (cost ratio, 1.04 [95% CI 0.98–1.11] P = 0.227; $10,795 vs $10,381), and results were similar for the FCM versus iron sucrose cohort comparison.

IDA-Related Total Costs of Care

The FCM cohorts in all three patient populations had lower IDA-related total costs of care in the follow-up period compared with low-dose and iron sucrose cohorts (Fig. 3). Whereas the FCM cohorts had higher IDA-related ambulatory visit and pharmacy costs, they had lower IDA-related inpatient, ED visit, and other medical costs compared with the low-dose and iron sucrose cohorts.

Fig. 3.

Iron deficiency anemia-related costs in the 12-month follow-up period^a. CKD chronic kidney disease, ED emergency department, FCM ferric carboxymaltose, HF heart failure, USD United States dollars. ^aCosts were defined as iron deficiency anemia related if the claim had a diagnosis for iron deficiency anemia in any position on the claim or was a medical claim for intravenous iron therapy

Discussion

Findings

In this retrospective claims analysis, a statistically significantly lower inpatient cost incurred in the FCM cohort in all patient populations. Mean all-cause total healthcare costs adjusted for baseline demographics and clinical characteristics were numerically lower for the FCM cohort compared with the low-dose iron cohort for patients with cancer and HF but not statistically significantly different; however, for patients with CKD, adjusted total costs were statistically significantly lower in the FCM cohort versus the low-dose iron cohort. Sensitivity analyses in each patient population consistently demonstrated significantly reduced all-cause total costs and inpatient costs in the FCM cohort compared with the subgroup of the iron sucrose cohort. At wholesale acquisition costs (2022/2023), minimum medication costs for a course of treatment are an estimated $2695 for 1500 mg of FCM, $360 for 1000 mg of iron sucrose, $406 for 1200 mg of iron dextran, and $509 for ferric gluconate [37–39]. Despite the higher drug acquisition cost of FCM relative to low-dose IV iron, the FCM cohort had generally lower inpatient and total costs that offset this difference.

In addition, for all three patient populations, there was a greater difference in the number of transfusion days between FCM and iron sucrose cohorts than between FCM and low-dose iron cohorts. The difference in transfusion days may provide a clinical explanation for the cost savings observed during the follow-up period. Overall, in addition to better clinical outcomes implied by the lower inpatient utilization and costs, these findings show that FCM can reduce the total cost of care relative to iron sucrose because of fewer inpatient stays, fewer inpatient days, and a lower blood transfusion rate.

The populations overall did not appear to supplement or switch treatment to oral iron therapy. The numbers of patients initiating each type of treatment over follow-up (including switched treatment) closely echoed the sizes of the treatment cohorts, suggesting few patients switched to an IV iron therapy other than their index treatment; thus, the intent-to-treat approach for the study cohort treatments provided the most realistic assessment of the impact of starting the selected treatment including the potential impact of discontinuation or low adherence. This study included patients who had only one FCM treatment, so patients may not have had sufficient exposure to impact clinical outcomes and subsequent utilization, as we could not determine whether this was a deliberate one-dose schedule or an incomplete treatment series; the inclusion of the patients with only one treatment was designed as a sensitivity analysis to address a limitation of the Kwong et al. study that required at least two treatments. In addition, iron sucrose requires more visits because of the dosing schedule, which would contribute to costs, although patients may be less adherent owing to the increased time required.

Our findings differ slightly from previous analyses using the IQVIA PharMetrics^® Plus database. Using a more restrictive definition for FCM claims (two or more doses of FCM in the identification period, which excluded patients with a single FCM dose) revealed that after controlling for baseline covariates, FCM treatment was associated with significantly lower inpatient, outpatient, and total costs in the follow-up period than low-dose iron for patients with cancer, HF, and CKD [28]. One of the main differences between the studies is the patient ages. In Kwong et al.’s study, commercially insured younger patients were overrepresented. Hence, we hypothesize perhaps this is the reason for fewer significant differences in the current study; in a more elderly patient population, cost-saving effects of IV iron in patients with cancer and HF cohorts were not high enough to produce statistical significance, although we do find a numerical cost reduction. Similar results were observed for comparisons between FCM and iron sucrose in each population. In comparison, patients in the current analysis were older, had higher baseline comorbidity scores, and a greater percentage were enrolled in Medicare Advantage plans.

Our findings are consistent with several non-US analyses that compared FCM with iron sucrose for the treatment of IDA in similar patient populations [32, 33]. Cost savings with FCM versus iron sucrose were because of a reduced inpatient stay length [41], personnel cost reductions [42], and reduced bed and nursing costs [43].

Limitations

The strengths of this analysis include a large sample size (N = 21,374), representing three chronic conditions, and both 6-month baseline and 12-month follow-up periods. Limitations include those inherent to claims data (e.g., coding errors, missing data, and reporting variations across clinical practices). Coding of healthcare resource utilization specific to IDA can be underutilized or misused. In addition, identification of chronic conditions from medical claims may not capture all conditions for a patient (particularly those with multiple comorbidities); however, the 6-month baseline period likely mitigated this source of potential misclassification bias. Additionally, potential bias resulting from unobservable differences between study cohorts cannot be eliminated, such as clinical variables that influence treatment choice (e.g., disease severity, cancer stage, medication side effects, clinical rationale, personal preference). Patients may have had multiple conditions, and the IDA could have been attributable to any of them, and this is difficult to discern using claims data. Rather than restricting the study to patients with only a single condition, patients were allowed to contribute to the populations for each condition they had. This allowed for the incorporation of patients with complex medication conditions in lieu of a more limited and less severely ill population. Baseline hemoglobin measures were not available, although it is possible that FCM use is targeted towards patients with worse baseline hemoglobin levels. There is variation in the cancer subgroups, but the effect represents the average effect across the groups. Our study found some differences in initial comorbidity levels between the cohorts of patients treated with FCM versus low-dose iron, which were addressed by the performed multivariable analyses to adjust for these and other observed differences between cohorts. Finally, results may not be generalizable to other US populations, such as patients who are uninsured or insured through Medicaid plans.

Conclusions

Despite the higher drug acquisition cost of FCM relative to low-dose IV iron, the drug price differential was offset by the lower inpatient cost incurred in the FCM cohort. All-cause total healthcare costs were lower for FCM versus low-dose IV iron products — numerically but not statistically significantly lower for FCM versus low-dose IV iron for patients with cancer and HF, and significantly lower for patients with CKD. When compared with iron sucrose, all-cause total healthcare costs were consistently significantly lower for FCM for patients with cancer, HF, and CKD, showing evidence of cost savings for FCM versus iron sucrose, the most used low-dose IV iron product. These findings further suggest the economic benefit of FCM to reduce inpatient utilization and associated costs to patients and health plans compared with low-dose IV iron products.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Funding

Optum Life Sciences was contracted by Daiichi Sankyo, Inc. to conduct this analysis. The analysis was funded by Daichi Sankyo, Inc.

Conflicts of Interest/Competing Interests

Nicole M. Engel-Nitz is an employee of Optum Life Sciences and a shareholder in UnitedHealth Group. Amy Anderson and Summer Tran were employees of Optum Life Sciences at the time the analysis was conducted, and Amy Anderson is a shareholder in UnitedHealth Group. Kevin Wang is an employee of Daiichi Sankyo and Winghan Jacqueline Kwong was an employee of Daiichi Sankyo at the time this analysis was conducted.

Ethics Approval

Institutional review board approval or a waiver of approval was not required for this study because the study data were secondary and de-identified in accordance with the United States Department of Health and Human Services Privacy Rule’s requirements for de-identification codified at 45 C.F.R. § 164.514(b).

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Availability of Data and Material

The data contained in the database used for the study contain proprietary elements owned by Optum and, therefore, cannot be broadly disclosed or made publicly available at this time. The disclosure of these data to third parties assumes certain data security and privacy protocols are in place and that the third party has executed a standard license agreement, which includes restrictive covenants governing the use of the data.

Code Availability

Not applicable.

Authors’ Contributions

All named authors meet the International Committee of Medical Journal Editors criteria for authorship for this article, take responsibility for the integrity of the work as a whole, contributed to the writing and reviewing of the manuscript, and have given final approval of the version to be published. NMEN, WJK, ST, and KW were involved in the conception and design of the study and data interpretation. NMEN and AA were involved in the acquisition of data. NMEN, ST, and AA were involved in the data analysis. All authors read and approved the final version.