Home versus hospital immunoglobulin treatment for autoimmune neuropathies: A cost minimization analysis

Abstract Background Prior clinical trials have suggested that home‐based Ig treatment in multifocal motor neuropathy (MMN) and chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) and its variant Lewis–Sumner syndrome (LSS) is safe and effective and is less costly than hospital‐administered intravenous immunoglobulin (IVIg). Methods A French prospective, dual‐center, cost minimization analysis was carried out to evaluate IVIg administration (5% concentrated) at home versus in hospital with regard to costs, patients’ autonomy, and patients’ quality of life. The primary endpoint was the overall cost of treatment, and we adopted the perspective of the payer (French Social Health Insurance). Results Twenty‐four patients aged 52.3 (12.2) years were analyzed: nine patients with MMN, eight with CIDP, and seven with LSS. IVIg (g/kg) dosage was 1.51 ± 0.43 in hospital and 1.52 ± 0.4 at home. Nine‐month total costs per patient extrapolated to 1 year of treatment were €48,189 ± 26,105 versus €91,798 ± 51,125 in the home and hospital groups, respectively (p < .0001). The most frequently reported factors for choosing home treatment were the good tolerance and absence of side effects of IVIg administration, as well as a good understanding of the advantages and drawbacks of home treatment (75% of respondents). The mRankin scores before and after switch to home treatment were 1.61 ± 0.72 and 1.36 ± 0.76, respectively (p = .027). Discussion The switch from hospital‐based to home‐based IVIg treatment for patients with immune neuropathy represents potentially significant savings in the management of the disease.


Immunoglobulins rank among the top drug expenses for hospital in
France. In addition to the high drug costs, IVIg treatments require the use of hospital resources for recurrent infusions. Home infusion has been used since the 1990s (Ochs et al., 1987;Ryan, Thomson, & Webster, 1988) and is now considered to be a safe alternative to hospital care. The possibility of providing safe home treatment with IVIg makes it possible to reduce treatment costs, make better use of hospital resources, and improve the patient's quality of life (QoL).
The aim of this analysis was to estimate and compare the costs of home-based vs. hospital-based recurrent infusions.

| PATIENTS AND METHODS
Patients currently treated for autoimmune neuropathy with IVIg in a hospital outpatient settings were recruited for a before-after study from two tertiary referral care centers in France.

| Design
The analysis was designed as a before-after analysis, with each patient being his or her own control. The design was chosen for practical reasons, as the patients enrolled in this analysis were stable and their IVIg treatment costs did not vary over time prior to switching to home treatment ( Figure 1). We compared the two IVIg treatment procedures in a cost minimization analysis, considering all direct costs to the healthcare system and community.

| Patient selection
Patients from the two French referral centers were monitored before and after switch to home infusion. For patients to be eligible for home treatment, they had to: 1. Be 18 years or older.

3.
Have received previous treatment with IVIg for a minimum of five cycles.

4.
Be responders to IVIg and clinically stable.

5.
Have received one home treatment for the same disease and agree to continue home treatment for at least three more cycles.
Patients were excluded if they were currently participating in another therapeutic trial, had combined neuropathy and monoclonal anti myelin-associated glycoprotein (anti MAG) gammopathy, or were already alternating hospital and home treatments. After individual consent was obtained, patients were asked to rank their reasons for switching to home treatment out of the following: comfort, no previous history of adverse events, good tolerance of IVIg, autonomy, costs, and family organization.

| Intervention
All patients had a caregiver (e.g., trusted family member) with them during the infusion, who had been trained on home management while the patient was in hospital. Before beginning home treatment, nurses in the referral facility were trained for IVIg infusion (Tegeline ® , 5% freeze-dried, sucrose-stabilized) or had previous experience in administering and monitoring these treatments. Hospital pharmacy staff provided additional training in handling and transporting blood products. Vascular access was reviewed and clinical competency for the nursing staff achieved. An administration and surveillance protocol was established according to the recommendations of the manufacturer (LFB Biomedicaments). A routine blood workup (urea, creatinine, blood cell count, hemoglobin, ALT/AST, coagulation assessment, ESR, C-reactive protein) was performed 24 hr prior to infusion and 24 hr after the end of the cycle as per current European guidelines (Elovaara et al., 2008). These tests are performed identically in all patients regardless of the place of treatment, and we did not report their costs in the comparison. Results were then systematically transmitted to the referral specialist for formal approval to proceed with the infusion. If F I G U R E 1 Study design premedication was used in the previous hospital infusion, the same protocol was used. The IVIg was administered using an infusion pump.
Prior to the home visit, the hospital pharmacy dispensed the IVIg directly to the nurse after the doctor orders and prescriptions were cross-checked. The IVIg was packaged in a validated, temperaturecontrolled cooler (monitored via a thermometer). Once at home, the patient was premedicated if this was part of the determined protocol. Venous access was achieved and baseline vital signs taken. The IVIg was administered following a predefined protocol (doses, infusion time, premedication). The patient status was monitored throughout the infusion. Side effects were reported to the principal investigator and to the pharmacy and recorded in the patient's medical chart. Once the infusion was complete, the IV line was discontinued and the patient was monitored for another 60 minutes to ensure that he or she was stable. Unused IVIg was returned to the pharmacy.

| Data collection
Healthcare resource utilization both in the hospital and in the community was recorded. Data for the "before" phase (hospital-based infusion) were extracted from the patient's medical charts, supplemented by the hospital's claims database. The claims database has linked records of all inpatient and outpatient admissions with a unique patient identifier. For all patients in the study, the following patient characteristics were recorded: age, sex, diagnosis, comorbidities, modified Rankin scale (mRankin) score, living conditions, profession, and eligibility for welfare benefits (Rankin, 1957).  The secondary endpoints assessed the quality of life of patients treated at home and the impact of home treatment on patients' autonomy. The impact of home treatment on autonomy was assessed by comparing the mRankin score before and after the switch to home treatment.
No ethics approval was required for this analysis as only routine care was given. Patients gave consent to the data collection and did not oppose the use of data already recorded in administrative database. All patient information was anonymized in the database using coded identification numbers, and no information in the database could be backtraced to reveal the patient's identity.
The analysis was undertaken between 2012 and 2014. All resources were valued at 2016 prices, and costs are reported in €2016.

| Sample size calculation
We calculated that 22 patients would provide 80% power, with a two-sided alpha level of 0.05, to detect a 50% relative decrease in the yearly total cost of IVIg treatment as compared to the in-hospital treatment (estimated total hospital cost of €100,000). This calculation method was conservative: Cost distributions tend to be skewed and follow a gamma distribution. Sample size calculations based on differences in means were found to be very conservative, giving numbers which substantially exceed the required power (Cundill & Alexander, 2015).
Given the need for continued monitoring of patients' disease, we did not expect any missing value on the use of healthcare resources.
The unit of analysis was the patient, using an intention-to-treat analysis based on period. All costs of the "before" period were attributed to hospital-based treatment, and all costs of the "after" period, including hospital admissions, were attributed to home-based infusion. The total cost of each period was divided by the number of months and multiplied by 12 to obtain a yearly patient cost. This calculation assumed that all patients were receiving a stable IVIg maintenance regimen. Outliers were not removed.
Continuous data were reported as mean ± SD, and the paired Student's t test was employed when comparisons were made for parametric data. Nonparametric data were analyzed with the paired Wilcoxon test, and we used 1,000 bootstrap replications to estimate the 95% confidence interval of the costs and cost difference. All tests were two-tailed, and a p value of <.05 was predetermined to represent statistical significance. Analyses were carried out using the SAS version 9.1 (SAS Institute, Cary, NC).

| RESULTS
The two centers identified 24 patients who were monitored for 9 months. Table 2 shows the baseline characteristics of the study patients. The mean age was 52.3 years (±12.2) with a male to female sex ratio of 2:1. Of the 24 patients included, 14 were working full-or part-time, one patient was unemployed, and the others were retired.
Six of 24 patients used implantable venous access devices, and the remaining 18 used IV lines. Patients were monitored for an average of 8.53 (±2.85) months and 4.96 (±2.86) months during the "before" and "after" periods, respectively.
During the "before" period, in hospital, the doses ranged from 37.2 to 203.7 grams per treatment cycle (1-2 g/kg), with a mean dose of 114.81 (±32.87) grams dosed every three to four weeks as per the protocols in the centers. Before the switch, 12 patients were treated as inpatients and 12 as outpatients. Patients traveled on average 20 km to go to the hospital.
After the switch to home treatment, Ig was delivered directly to the patients in six cases, and to a nurse in the remaining 18. Infusion devices were IV line and electric pump, and infusion times at home averaged 3 hr. Doses ranged from 58.5 to 222.3 g per treatment cycle, with a mean dose of 119.38 (±38.14) g.
During the "after" period (at home), two patients were admitted to hospital and one patient discontinued home treatment for personal reasons.
Overall costs per patient per cycle before the switch amounted to

| DISCUSSION
Home-based IVIg for treatment of primary (PID) or secondary immunodeficiency (SID) at a low dose (0.4 g/kg/cycle) has been used since the 1990s (Ochs et al., 1987;Ryan et al., 1988) in Europe and North America. As early as 1994, use of high doses of IVIg at home was authorized in the Netherlands for the treatment of immune neuropathies (Cats, Van der Pol, Bertens, & van den Berg, 2011), subject to certain criteria being me prior to treatment. These included administration of at least one cycle of IVIg (corresponding to a cumulative dose of 2 g per kg per cycle) in hospital, the presence of a nurse specialized in home management during the last cycle administered in hospital, prescription of an anti-allergic reaction kit (epinephrine, prednisone, and an antihistamine), blood pressure monitoring, and verification of the possibility of a venous access port. During the same period, home infusion was not yet a widespread practice in patients with autoimmune diseases in France and many other European countries, probably due to a lack of experience among hospital practitioners and/or concern about adverse events (AEs). Even though IVIg is generally safe, serious adverse events such as thromboembolic events (Marie, Maurey, Hervé, Hellot, & Levesque, 2006;Rajabally & Kearney, 2011) or renal failure (Caress, Kennedy, & Eickman, 2010) can still occur, especially in patients treated by a high dose of IVIg or with concomitant diseases. In this analysis, the mean number of AEs was not statistically different between hospital and home treatment, and most AEs were ranked as mild or moderate. Home administration of IVIg began in France in the 1990s as a cost-saving measure, as well as for the benefit of patients' comfort and quality of life (Hachulla et al., 2002), and it is now considered to be a safe alternative to hospital-based treatments.
Home-based IVIg as maintenance therapy clearly reduced hospital costs in our group of patients with autoimmune neuropathies.
Total yearly treatment costs were divided roughly in half, from an average of €91,000 down to €48,000 per patient, with consistent savings across all three indications. Cost savings were achieved through fewer admissions and, to a much lesser extent, fewer commutes.
Lower cost of MMN patients was explained by more outpatient hospital care.
Patients were satisfied by the switch to home treatment and experienced a small but significant reduction in their mRankin score.
However, only patients that had already tolerated home-based treatment and wanted to continue were included, which might bias the results toward a positive opinion on home treatment. We did not attempt to establish a causal relationship between satisfaction and home switch as with a randomized trial, but rather to propose a costreducing alternative for selected patients. It was designed as a before-after study with no randomization and would be useful to policymakers only insofar as "switchable" population could be identified and be large enough to justify investing in home-based treatment.
T A B L E 3 Resource utilization and treatment costs in € before and after the home treatment switch.