A budget impact analysis of iron polymaltose and ferric carboxymaltose infusions

Background In Australia, iron deficiency anaemia can be managed by ferric carboxymaltose, and iron polymaltose given via either a traditional slow or new rapid infusion protocol. These differ in their manufacturing, administration, and monitoring requirements, with unknown associated costs. Aim To compare the direct costs of iron infusions used in Australia; and explore potential savings associated with increased uptake of the least-expensive option at a local hospital. Method A time-motion method was used to determine the labour and consumables associated with each infusion protocol. Secondly, a frequency analysis identified the most common iron infusion doses prescribed at the study site. The total direct costs per protocol were compared at these doses and then the potential savings from switching to the lowest-costing of these protocols where possible were explored. Results The most common doses were 0.5 g, 1 g, 1.5 g and 2 g. At these dose points, ferric carboxymaltose infusions are the least expensive, but only if national health subsidies are applied. In cases where they do not apply, iron polymaltose prepared from ampoules and infused using the rapid protocol (‘Iron Polymaltose Ampoules Rapid’) is the least expensive. Switching all applicable ferric carboxymaltose infusions and iron polymaltose infusions administered using the slow infusion protocol to Iron Polymaltose Ampoules Rapid is projected to yield up to $12,000 worth of savings annually. Conclusions Increased use of the Iron Polymaltose Ampoules Rapid protocol when government-subsidised options are not available is projected to have cost-saving outcomes. Investigation of implementation strategies to increase the use of this protocol are warranted.


Impacts on Practice
• In a time where healthcare budgets are increasingly strained, this study identifies the potential for significant cost savings with simple and easy to implement switches in iron infusion protocols. These savings would allow a greater number of patients to be treated for iron deficiency, and/or for resources to be redirected to more urgent priority areas. • Future studies exploring the budget impact of similar swaps in other countries would be warranted, as well as the impact of these changes from the patient's perspective.

Introduction
Iron deficiency anaemia (IDA) is a common health problem. Between 1990 and 2010, approximately 32.9% of the global population had anaemia, most commonly caused by iron deficiency [1]. Given the proportion of people affected by IDA and the considerable financial implications of largescale intravenous iron use, cost-efficient interventions are necessary to both minimise the associated treatment burden, as well as to allow greater access to treatments. In Australia, there are three main types of intravenous iron products available: iron sucrose, iron polymaltose (IPM) and ferric carboxymaltose (FCM). This article will only focus on IPM and FCM as iron sucrose use is limited to specialist settings in Australia [2]. Both IPM and FCM have good efficacy and safety profiles and there is no conclusive evidence indicating one is superior over the other [3][4][5][6]. They differ however, significantly in procurement costs as well as in their requirements for manufacturing, administration and monitoring; together this may have budget and patient flow implications.
Current protocols recommend that up to 2.5 g of IPM be given slowly over 2 to 3 h, and for up to a total duration of 5 h [7,8]. In contrast, FCM can be given in a single 15-min infusion but it is restricted to doses of 1 g, of which only one can be administered weekly [9,10]. FCM typically has a less demanding monitoring regimen of observations: at baseline, immediately after the infusion and 30 min later. Comparatively, locally, patients' vital signs are recorded by nurses every 10 min for the first 30 min, then half-hourly until a standard IPM infusion is completed [8]. Local safety protocols dictate that any infusion requiring more than five ampoules of medication be manufactured in the pharmacy department, where they are made in the sterile suite (infusions made up by nurses on the ward are not). Due to the currently available proprietary formulations, this only commonly applies to IPM infusions, and not FCM, because the usual dose range (1 + gram) requires at least ten IPM ampoules, but only two FCM ampoules at a time [7,9].
In recent years, an alternative 'rapid infusion' protocol of IPM has been implemented at several sites in Australia [11]. Studies report that this protocol is not inferior in efficacy or safety compared to the standard IPM protocol [12][13][14][15]. This protocol approves administering doses ≤ 1.5 g over 75 min, and doses up to 2 g over 105 min, with fewer monitoring requirements compared to the slow IPM protocol [7].
Numerous economic-related analyses carried out have concluded that FCM is cost-effective in the treatment of IDA [16][17][18][19][20]. However, there has been recent studies which show the potential of alternative intravenous iron therapy to yield cost savings. For example, an observational study by Delpeuch, et al. presented iron sucrose as a cost-saving alternative [21]. And Pollock et al. conducted cost analyses comparing FCM and iron isomaltoside in multiple countries showing that fewer infusions were required per patient to manage iron deficiency using iron isomaltoside than FCM, resulting in reduced costs [22,23]. While these iron infusion therapies are not recommended in Australia, the studies prompted an investigation comparing the costs associated with different iron infusion protocols utilised locally, as well as the potential opportunity for costs savings by implementing switches to within a local hospital [2].

Aim of the study
Given the range of options now available, the aim of the present study was to calculate and compare the direct costs of each infusion type at different doses, and to explore the potential budget savings associated with a tertiary hospital switching to the lowest costing of these protocols where possible.

Ethics approval
The study was approved by the Tasmanian Human Research Ethics Committee (H0017749, 28 November 2018).

Method
The methodology was two-fold. Firstly, a time-motion study was conducted at a local hospital to ascertain the labour and consumable requirements needed for both the established FCM and slow IPM infusion protocols. This methodology involved an external observer (CKL) attending the manufacturing and administration of these infusions at the study hospital. Initial observations were undertaken to record the necessary manufacturing and administration steps taken to produce and deliver these infusions, as well as the associated consumable requirements and the type of staff involved in each step. Worksheets (see Electronic Supplementary Material A) were subsequently developed in discussion with hospital staff. These worksheets were then used to record the time needed for each step using a stopwatch, with the observer watching unobtrusively from a distance. It was planned that twenty timed observations would be made of both the FCM and IPM slow protocols each between November 2018 and March 2019; this sample size was based on studies of similar designs [24][25][26].
Secondly, an estimation of the cost of each infusion type was calculated at the doses most used at the study site to allow comparisons across different doses (see Table 1), as well as the annual budget contribution of the FCM and slow IPM protocols to the study hospital. Analyses then were carried out to determine the budget implications associated with several different substitution scenarios (detailed under Analyses).

Population
The time-motion aspect of the study was conducted at Royal Hobart Hospital (RHH), Tasmania's major public (tertiary) hospital, in late 2018 to early 2019. For the budget impact analysis, relevant dispensing data from 2018 was also used from this site.

Drug procurement costs
FCM infusions were prepared from ampoules containing 500 mg/10 mL of FCM. The slow IPM infusions were administered using either bags made on site from ampoules containing 100 mg/2 mL of IPM or by using premix bags containing 1000 mg/500 mL of IPM. The wholesale price of each drug formulation in Australian Dollars (AUD) in 2019 was $142.80, $2.50, and $101.30 for a single FCM ampoule, IPM ampoule and IPM premix bag, respectively.

Consumables
For FCM manufacturing, a 20 mL syringe, one 19-Gauge sharp needle, one 250 mL sodium chloride (NaCl) 0.9% bag, one drawing up needle, two pairs of gloves and an alcohol swab were used. By contrast, consumables used in IPM manufacturing constituted of one 500 mL NaCl 0.9% bag, two 21-Gauge needles, one 0.8-micron filter, one additive cap, five iso wipes, one drawing up needle, two pairs of gloves and a light-protective bag. Sterile suite garments were not costed into the IPM calculations as many different infusions are made in one 'sitting', and the suite costs are considered part of the normal pharmacy department costs for daily work.
The consumables used during administration of FCM and IPM infusions were the same: a cannula, a tourniquet, an extension set, a giving set, two alcohol swabs, one cotton wool ball, one pressure pad and a 500 mL NaCl 0.9% bag. These costs were obtained from the site Supply and Purchasing Officer. The cost of the volumetric pump and tourniquets were not included as they are standard reusable hospital equipment.

Labour
The type of staff who carried out each step of the observed infusions was noted. Only time spent preparing and administering infusions (including monitoring patients receiving the infusions) was recorded; interruptions were ignored.

Analyses
All calculations were performed using Microsoft Excel version 2010 (Microsoft Corporation, Redmond, WA, USA). The time recorded from observations were summarized as means with standard deviation (SD). Pharmacy and nursing labour costs were calculated by multiplying the mean times taken in the preparation, administration, and monitoring of infusions, by the respective average hospital wage award for the staff type attending to these tasks. A frequency analysis was conducted using the dispensing data to ascertain the iron doses and protocols most commonly prescribed at the study site in 2018.
At the time of study, there were three different intravenous iron therapy at the study site: FCM, IPM prepared from ampoules and infused using the slow protocol (IPM Ampoules Slow), and IPM prepared from premix bags, also infused using the slow protocol (IPM Premix Slow). The approval of an IPM rapid protocol would increase this to five, with the addition of IPM prepared from ampoules and infused using the rapid protocol (IPM Ampoules Rapid) and IPM prepared from premix bags, also infused using the rapid protocol (IPM Premix Rapid).
The cost of each protocol at the most commonly prescribed doses was calculated by combining relevant drug acquisition, labour and consumables costs. For FCM infusions, labour and consumables costs were doubled when doses were above 1 g. Since IPM rapid infusions were not implemented at the RHH at the time the study was being conducted, its direct costs were estimated by extrapolating data from observed infusions (see Electronic Supplementary Material A).
The total direct costs per infusion for each infusion protocol were compared to each other at the common dose points to determine the lowest costing infusion protocol. In Table 1 Costing methodology used to calculate the direct cost of each protocol type a For FCM doses > 1 g, these were doubled to account for patients needing to return for their second dose at least a week later b This refers to the labour associated with the manufacturing, administration and patient monitoring of associated iron infusions  [27,28]. Furthermore, PBS reimbursements only apply to hospital dispensed medication as long as the medication is prescribed and dispensed on the day of discharge or if the patient is an outpatient. Thus, the difference in costs with (FCM-PBS and IPM-PBS) and without PBS reimbursement (FCM-NPBS and IPM-NPBS) were also accounted for in these analyses. Finally, a budget impact analysis, which is defined as an economic assessment that estimates the financial consequences of adopting a new intervention on an institution's budget, was undertaken. In this study, the lowest costing infusion protocol was projected to be used in place of other existing infusion protocols thereby estimating the potential cost savings associated. The initial estimation was made assuming that all infusions were changed to the lowest costing infusion type. A second analysis was then conducted exploring the budget impact of infusions only being changed to their lowest-costing contemporary i.e. where those involving premade bags were only switched to the lowest-costing of those that involved a premade bag, and those involving ampoules only being switched to the lowest-costing infusion type requiring ampoules. This second analysis was conducted to determine the cost implication of making protocol changes without adding to the workload of hospital staff who manufacture these infusion bags in-house. All analyses were carried out based on the inference that these protocols are not inferior to each other in terms of safety and efficacy.

Frequency analysis
Excluding iron sucrose prescriptions, there were a total of 1274 iron infusions prescribed at the study site in 2018, the majority comprising of FCM infusions (1136 or 89.2%), and specifically PBS-reimbursed FCM infusions (1100, 86.3%, N = 1274). The next most common form of iron infusion was the premix IPM bags (105, 8.2%), followed by IPM infusions made on site from ampoules (33, 2.6%).
Forty patients received FCM repeat doses within a month (31 days) of their first dose. Taking into account FCM's 1 g limit per week, [10] these were defined as part of their initially prescribed dose. No patients who received an IPM infusion received a repeat dose within the month of their first dose. The most common doses prescribed for any iron infusion (i.e. IPM and FCM combined) were 1 g (1042 infusions (84.4%)), 500 mg (115 infusions (9.3%)), 2 g (45 infusions (3.7%)) and finally, 1.5 g (17 infusions, (1.4%), N = 1234); these four dose points accounted for approximately 98.8% of doses.

Direct costs comparison
The average labour times associated with each infusion type are listed in Table 2. Table 3 outlines the total direct cost breakdown for each infusion protocol at the four aforementioned common dose points. PBS reimbursements were greater than the drug acquisition costs for FCM at all doses, making it the lowest costing infusion at all dose points. Conversely, when not attracting PBS subsidies it was the most expensive protocol; IPM Ampoules Rapid had the lowest total direct cost at all common dose points for non-subsidised infusions.

Budget-impact analysis
The annual cost of FCM-NPBS was approximately $11,804.94 (Table 4). If these infusions were changed to IPM Ampoules Rapid infusions (assuming no clinical contraindications to IPM existed), the projected annual cost would be reduced to $4,312.65, amounting to a total of up to $7,492.29 in saving. The annual costs of IPM Ampoules Slow and IPM Premix Slow were $4,717.62 and $17,926.30 respectively (Table 5). Converting these to IPM Ampoules Rapid infusions would reduce the projected annual cost to $18,158.78 i.e. a saving of $4,485.14 annually.
Together the budget impact of swapping all infusions that did not attract PBS subsidies to IPM Ampoules Rapid

Discussion
Our findings reveal that FCM infusions amass the lowest cost from the hospital perspective, but only if PBS subsidies apply. When prescriptions cannot be subsidised, IPM Ampoules Rapid is the least expensive protocol, irrespective of the dose prescribed. The potential savings associated with increasing use of the IPM Ampoules Rapid protocol could allow up to $12,000 to be re-directed for other patient care purposes. Other similarly sized hospitals without protocols in place requiring infusions with more than 5 ampoules be made in the pharmacy sterile suite would stand to save more. Currently however, these local protocols do exist, and an increase in the use of the IPM Ampoules Rapid protocol would require more infusions to be manufactured in the pharmacy's sterile suite each year-an extra 174 infusions annually, based on 2018 data. If there was resistance to this, for example due to hospital and staff capacity constraints, then changing all FCM-NPBS infusions and IPM infusions of doses up to 2 g to their respective IPM rapid counterpart would still result in a saving of up to $7,053 annually for the hospital (see Electronic Supplementary Material B). The term 'up to' is used in the above paragraph (as well as in the results section) in acknowledgement that switching from FCM to IPM is not always appropriate. FCM is used locally for in-patients in: (1) the management of IDA for women in the second or third trimester of pregnancy where the benefits of treatment are deemed to outweigh the risks to the fetus or (2) patients with known and documented intolerance or allergy to iron polymaltose [10]. FCM may be switched to IPM Ampoules Rapid only in cases of the former indication, where neither FCM nor IPM were found to be superior to one another when used in pregnant women [6].
Physicians are recommended to give greater consideration to the use of intravenous iron due to the introduction of these new infusion therapies which not only have shorter administration duration times, but also favourable safety profiles; however the cost implications have not been previously fully investigated [29]. A study by Khalafallah et al. [30] compared the cost of a 15-min FCM infusion with a 2-h IPM infusion in the treatment of IDA during pregnancy. No significant difference between the average overall costs of the two infusion options was found but only a single standardized 1-g dose was investigated, and the costing incorporated other aspects of patient care over and above the infusion itself. By contrast, our study measured and accounted for the actual labour costs associated with each infusion. Furthermore, it accounted for variations in manufacturing and medication costs at different doses. To the best of our knowledge, our study is the only study which has investigated costs at a range of doses for FCM and various types of IPM protocols. Our findings support the increasing uptake of IPM rapid protocols in Australia [11,15].
Between 2013 and 2017, there has been an increase in the use of intravenous iron which is may be due to the listing of ferric carboxymaltose on the PBS in 2014 [31]. One of the drivers for the increasing FCM use in recent years can be related to its short infusion duration and thus increased capacity for infusions per day compared to IPM infusion protocols, particularly when only 1 g of iron is required. For example, a higher number of patients can be administered 1 g infusions in a standard workday (7.5 h) if FCM is administered instead of IPM rapid (10 vs 6 a day, respectively). When doses over 1 g are required however, the numbers are more comparable, accounting for FCM patients needing to return for a second infusion at least a week later. Interestingly, Gilmartin et al. [32] found that only 14% of patients who require more than 1 g of FCM return for their second dose. From a clinical perspective, IPM rapid infusions may be preferred for patients requiring more than 1 g of iron, despite costing more.
The main limitation of this study was the difficulty in IPM data collection due to the ready availability of the premix IPM bags and the local preference for using FCM. Only six IPM infusions were manufactured during the data collection period. Instead of going through the time-motion process, these infusions were time-stamped by pharmacy manufacturing staff; although not ideal, this was sufficient to generate a mean labour time for the manufacturing of IPM infusions. Similarly, no observations were made of the administration and associated patient monitoring for any IPM infusion. We maneuvered around this by extrapolating IPM costs from FCM infusion administration costs. This was possible as the process to set-up the infusions once manufactured is the same, and the parameters involved in monitoring the patient for each infusion type are similar.

Conclusion
The IPM Ampoules Rapid protocol was cheaper than both FCM without government subsidies and the traditional slow IPM infusion protocol. Switching any FCM infusions that do not meet subsidy requirements and IPM slow infusion protocols to IPM Ampoules Rapid at our local study site is projected to yield up to almost twelve thousand dollars' worth of savings annually. Further studies considering all costs beyond the hospital perspective (i.e. patient costs associated with repeat FCM infusions) would be useful to further explore and compare the cost implications of these protocols.

Conflicts of interest
The authors have declared no potential conflicts of interest.