Treatment Preferences for Subcutaneous Immunoglobulin Among Patients with Chronic Inflammatory Demyelinating Polyradiculoneuropathy, Caregivers, and Physicians: A Discrete Choice Experiment

Abstract

Introduction

Subcutaneous immunoglobulin (SCIG) is a recommended maintenance treatment option for chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) and could offer greater patient convenience than intravenous immunoglobulin (IVIG). We evaluated the burden of CIDP and assessed the value that patients with CIDP, caregivers, and physicians place on various attributes of SCIG treatment options.

Methods

This was an online discrete choice experiment survey designed through targeted literature review and patient/caregiver/physician interviews, conducted in Germany, the UK, and the USA. The survey was administered to adults with CIDP who were stable on IVIG and/or receiving SCIG, and caregivers/physicians. Participants chose preferred treatment profiles comprising six SCIG treatment attributes with varying levels: chance of CIDP symptom worsening, treatment frequency, number of sites per infusion, time spent on infusion per month, chance of infusion site reaction, and infusion location. Subgroup analyses were also performed. A conditional logistic regression model estimated preference weight for each attribute.

Results

Overall, 146 patients, 52 caregivers, and 81 physicians completed the survey; mean (standard deviation) patient age was 58.1 (12.8) years; time since CIDP diagnosis was 8.8 (9.3) years; 98.6% had received IVIG, and 30.8% had received SCIG. Patients and caregivers experienced a substantial burden of disease that negatively impacted quality of life and work productivity. The most important factors in treatment decision-making were a lower chance of CIDP symptom worsening in 6 months, less frequent treatment administration, and home infusion (all p < 0.001 vs reference categories). Patients considered a lower chance (10% vs 20%) of symptom worsening and lower treatment frequency (every 3–4 weeks vs weekly) as the most important SCIG attributes, accounting for 64.6% of the decision weight.

Conclusion

There remains a substantial burden on patients and caregivers due to CIDP. Patients with CIDP highly valued efficacy and convenience when considering preferred SCIG options. Preference awareness may facilitate shared patient–physician decision-making.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40120-026-00915-1.

Key Summary Points

Why carry out this study?

To understand if subcutaneous immunoglobulin (SCIG) can offer a more convenient alternative to intravenous immunoglobulin as maintenance treatment for chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) owing to the option for home-based self-administration.

The study evaluated the burden of CIDP on physical and social activities, and work productivity, and assessed which treatment attributes contribute to preferences for SCIG treatment options among patients, caregivers, and physicians.

What was learned from the study?

CIDP imposes a substantial burden on patients and caregivers, affecting their quality of life and work productivity.

Among the attributes included in the survey, treatment preferences among patients, caregivers, and physicians were primarily driven by efficacy and convenience, and treatment frequency.

Awareness of preferences for different treatment attributes may facilitate shared patient–physician decision-making when optimizing treatment strategies.

Introduction

Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) encompasses a spectrum of immune-mediated neuropathies that cause weakness and sensory symptoms in a chronic progressive or relapsing–remitting disease course [1]. CIDP places significant burden on patients, including impaired physical function, psychosocial distress, and increased prevalence of depression [2].

Intravenous immunoglobulin (IVIG) and corticosteroids are guideline-recommended first-line treatments for CIDP [1]. Although these treatments are effective in controlling CIDP symptoms, logistical, convenience, and tolerability aspects of IVIG, and safety concerns regarding long-term corticosteroid use could be burdensome for patients [3]. For instance, IVIG treatment is associated with rare but serious risks such as thromboembolism and requires venous access and administration by a healthcare professional (HCP), which increases the burden on both patients (e.g., travel requirements) and healthcare providers [4, 5]. Since 2018, subcutaneous immunoglobulin (SCIG) has been available as an alternative to IVIG for the maintenance treatment of CIDP [5]. SCIG is associated with a lower risk of systemic adverse events than IVIG, does not require venous access, and allows for convenient home-based self-administration [5]. However, the maximum volume that can be delivered in the subcutaneous space is limited and may therefore necessitate weekly infusions, which may be restrictive for some patients and limit their activities [5].

Given that various options are now available for administration of immunoglobulin G (IgG) therapy, it is timely to consider the relative importance that physicians and patients with CIDP place on various attributes of these different treatments. It is also pertinent to review the preferences of caregivers, given they may be actively involved in treatment administration to a broad range of patients, including older patients with dexterity issues and pediatric patients with CIDP, an important consideration given that SCIG is indicated in the EU for use in adolescents and children of any age [6]. Discrete choice experiments (DCEs) are a well-established approach for determining stated preferences in healthcare settings, allowing estimation of the relative importance of different treatment attributes and the trade-offs between them [7]. This study evaluated the burden of CIDP on health-related quality of life (HRQoL) and work productivity among patients with CIDP and caregivers, and used the DCE method to assess the extent to which treatment attributes contribute to the relative preferences for different SCIG formulations from the perspectives of patients, caregivers, and physicians.

Methods

Overview of Study Design

The study was conducted in two phases: a qualitative phase and a quantitative phase, following the International Society of Pharmacoeconomics and Outcomes Research (ISPOR) good-practice guidance on conjoint analysis design [8]. In the qualitative phase, a targeted literature search was conducted to compile a list of the key attributes of SCIG treatment options. A series of one-on-one interviews with eligible patients, caregivers, and physicians was carried out to inform and validate the identified treatment attributes of interest, which were then used to develop DCE surveys. In the quantitative phase of the study, participant characteristics (e.g., patient and caregiver demographics, disease burden, and treatment characteristics) and the DCE surveys were used to collect data from patients, caregivers, and physicians to establish their stated preferences for different SCIG treatment options. The study was conducted in accordance with the Helsinki Declaration of 1964 and its later amendments, and was granted exemption from a full review by the Pearl Institutional Research Board according to FDA 32 Code of Federal Regulations (CFR) 56.104 and 45CFR46.104(b)(2): (2) Tests, Surveys, Interviews on 01/11/2022 (amendment approved on 03/31/2022). Informed consent was obtained from participants prior to interviews and online survey completion. Participants received an honorarium payment for their time spent and effort participating in this research.

Participants

Recruitment of patients and caregivers was conducted in collaboration with a patient advocacy group (Guillain-Barré Syndrome/CIDP Foundation International [https://www.gbs-cidp.org/]) and an established recruitment agency through a range of outreach activities. For example, working with the GBS/CIDP Foundation, mail blasts were used to reach their members to inform them of the opportunity to participate in the study. Outside of the Foundation, a study flyer was circulated at a CIDP symposium. Where an individual was interested in participating in the study, they could reach out to the recruitment agency using the contact information on the flyer to confirm their eligibility for participation. The recruitment agency also worked with their established network of providers to secure additional referrals. Physicians were recruited by the same agency.

For inclusion in the study, patients and caregivers were required to be at least 18 years of age and residing in the USA, the UK, or Germany. To best reflect the population that would make decisions between different hypothetical attributes among SCIG treatment options that were or may become available in clinical practice, eligible patients had a self-reported diagnosis of CIDP, were stable on IVIG for CIDP for at least 3 months, or had prior experience with SCIG. Caregivers must have provided care for at least 3 months to patients who were stable on IVIG for CIDP. Physicians were neurologists or neuromuscular specialists who had treated patients with CIDP for at least 3 years and had treated at least 3 patients in the previous 3 years.

Qualitative Phase: Defining Treatment Attributes

A targeted literature search was performed to identify relevant attributes and levels pertaining to the experiences and preferences of patients, caregivers, and physicians with CIDP treatments (with a particular focus on SCIG). Online publication databases (PubMed, Google Scholar) and product labels of approved CIDP treatments, as well as recent publications of clinical trials of SCIG treatment for CIDP were searched. This resulted in an initial list of CIDP treatment attributes and corresponding levels that was explored further in one-on-one interviews, including 4 patients with CIDP, 4 caregivers, and 4 neurologists meeting the same respective eligibility criteria, in the USA and the UK. These interviews were used to inform and refine participants’ perceptions of the most important SCIG attributes and corresponding levels that may affect treatment choices.

Quantitative Phase: DCE Survey Design

In total, six attributes, with two to four levels for each attribute, were identified and confirmed during the qualitative phase of the study (Table S1); these were consistent across physicians, patients, and caregivers. No changes were made to the treatment attributes and levels following the one-on-one interviews. Based on the predefined attributes and corresponding levels, 36 DCE choice cards were generated using the SAS macro DCE package (SAS Institute, Cary, NC, USA) to achieve an orthogonal, balanced, and efficient design, assigning two hypothetical SCIG treatment profiles to each choice card. The choice cards were randomly divided into three blocks (i.e., 12 choice cards in each block). Each participant was randomly assigned to respond to one of the blocks. The order of choice cards was also randomized to avoid potential ordering effects. An example of a DCE choice card is shown in Fig. 1.

Fig. 1.

Example of a treatment choice card

Each block included a pair of specifically constructed choice cards to evaluate internal validity and the consistency of participants’ responses. The two choice cards were designed to have the same treatment profiles but in a reversed order to assess test–retest reliability (i.e., choice card 1 has Profile A vs Profile B whereas choice card 2 has Profile B vs Profile A). In addition, one treatment profile (dominant profile) was designed to represent a commonly perceived unambiguously better treatment option.

A 5-min tutorial was designed using plain English and visual aids to provide a basic description for each of the treatment attribute levels, and to familiarize participants with the concept of choosing a preferred treatment profile between two options on a choice card.

Quantitative Phase: DCE Survey Data Collection

Following completion of the qualitative phase of the study, surveys were developed for patients, caregivers, and physicians. Surveys were conducted via a web-based secure portal, and were anticipated to take approximately 20 min for participants to complete.

The surveys incorporated screening items to confirm participants’ eligibility and then collected data relating to participants’ characteristics, including the experiences of patients with CIDP treatments and the impact of the disease on their daily lives (Table S2). Established outcome measures captured various aspects of the impact of CIDP experienced by patients and/or caregivers: the Inflammatory Neuropathy Cause and Treatment Disability Scale (INCAT) [9], the Rasch-Built Overall Disability Scale (R-ODS) [10], the EQ-5D-3L, the EQ-VAS [11–13], and Work Productivity and Activity Impairment (WPAI) pertaining to CIDP [14].

Subgroup Analyses

Exploratory analyses of patient preferences were conducted in the following categories: age (< 60 vs ≥ 60 years, above vs below the median), sex (male vs female), CIDP type (typical CIDP vs non-typical CIDP), use of SCIG (with vs without prior SCIG experience), and INCAT score (INCAT = 0–1 vs ≥ 2).

Statistical Analysis

Sociodemographic characteristics of physicians, patients, and caregivers, and clinical characteristics of patients were summarized descriptively. Scores for patient- and caregiver-reported outcomes were calculated according to the scoring algorithms of the individual instruments. Participants’ preferences for the CIDP treatments were assessed separately for patients, caregivers, and physicians using multivariable conditional logistic regressions. Regression coefficients and their 95% confidence intervals were reported as preference weights (i.e., part-worth utilities). Coefficients from the regression analysis were also used to assess the relative importance of treatment attributes from the perspective of the respective populations.

Internal validity was assessed by evaluating the consistency of responses by an individual on the constructed dominant profile and test–retest choice card pairs. Percentages of participants who satisfied the internal validity check were summarized for the overall sample and specific cohorts.

For the exploratory subgroup analyses, patients’ preferences for CIDP treatments for each subgroup were assessed using stratified multivariable conditional logistic regression models. For each subgroup, p values were obtained according to the interaction terms of the respective categories with each of the attributes.

All analyses were performed in SAS 9.4 (SAS Institute, Inc., Cary, NC).

Results

Participant Characteristics

In total, 146 patients, 52 caregivers, and 81 physicians met the study criteria and completed the survey. Each caregiver provided data for 1 patient under their care. Most patients and caregivers were from the USA (87.0% and 65.4%, respectively; Table 1), whereas the distribution of physicians was more even across the USA, the UK, and Germany (Table 2).

Table 1.

Patient and caregiver sociodemographic characteristics

	Patients (N = 146)	Caregivers (N = 52)
		Self	Patients under their care
Age, years, mean (SD)	58.1 (12.8)	52.3 (14.4)	56.4 (15.0)
Sex, n (%)
Male	64 (43.8)	19 (36.5)	24 (46.2)
Female	81 (55.5)	33 (63.5)	28 (53.8)
Prefer not to answer	1 (0.7)	–	–
Country of residence, n (%)
USA	127 (87.0)	34 (65.4)	–
Germany	10 (6.8)	10 (19.2)	–
UK	9 (6.2)	8 (15.4)	–
Employment status,a n (%)
Retired	48 (32.9)	12 (23.1)	19 (36.5)
Unable to work due to disability	46 (31.5)	1 (1.9)	19 (36.5)
Full-time	40 (27.4)	15 (28.8)	7 (13.5)
Self-employed	12 (8.2)	4 (7.7)	0 (0.0)
Part-time	11 (7.5)	7 (13.5)	5 (9.6)
Unemployed	9 (6.2)	6 (11.5)	7 (13.5)
Homemaker	1 (0.7)	6 (11.5)	0 (0.0)
Student	–	1 (1.9)	2 (3.8)
Other	2 (1.4)	2 (3.8)	3 (5.8)

SD standard deviation

^aCategories were not mutually exclusive

Table 2.

Physician sociodemographic characteristics and experience of treating CIDP

	Physicians (N = 81)
Age, years, mean (SD)	51.4 (9.9)
Sex, n (%)
Male	23 (28.4)
Female	55 (67.9)
Prefer not to answer	3 (3.7)
Country of practice, n (%)
USA	30 (37.0)
Germany	26 (32.1)
UK	25 (30.9)
Years of practice, mean (SD)	19.6 (7.2)
Years of treating patients with CIDP, mean (SD)	17.9 (7.2)
Primary practice setting,a n (%)
Academic institution/medical centers	41 (50.6)
Private practice	30 (37.0)
Public or government institution	12 (14.8)
Non-academic hospital or health system	9 (11.1)
Number of patients with CIDP treated over the past 3 years,b mean (SD)	24.4 (27.4)
Types of therapies prescribed to patients with CIDP,a n (%)
Corticosteroids	81 (100.0)
IVIG	80 (98.8)
Immunosuppressants	75 (92.6)
Plasma exchange	59 (72.8)
Physiotherapy	58 (71.6)
SCIG	47 (58.0)
Other	8 (9.9)
Aware of SCIG as a treatment for patients with CIDP, n (%)	81 (100.0)
Prescribed SCIG as a treatment for CIDP, n (%)	46 (56.8)
Reasons for not prescribing SCIG,a,c n (%)
Access/availability issues	21 (60.0)
Lack of monitoring at home	17 (48.6)
Patients’ ability to administer the infusion	15 (42.9)
Lack of information on efficacy and/or safety	12 (34.3)
Concerns about patients’ compliance	8 (22.9)
Patient preference	5 (14.3)
Lack of personnel to provide training to patients	5 (14.3)
Cost/insurance coverage	5 (14.3)
Efficacy	3 (8.6)
Safety concerns	3 (8.6)
Drug–drug interaction	1 (2.9)
Other	6 (17.1)

CIDP chronic inflammatory demyelinating polyradiculoneuropathy, IVIG intravenous immunoglobulin, SCIG subcutaneous immunoglobulin, SD standard deviation

^aCategories were not mutually exclusive

^bOne physician’s response was imputed as missing due to an improbable value

^cAmong the physicians (N = 35) who had not prescribed SCIG as a treatment for CIDP

Patients

The mean (standard deviation [SD]) age of the patient group was 58.1 (12.8) years and 55.5% were female. Fewer than half (43.2%) were in the work force (full-time, part-time, or self-employed) (Table 1). The mean (SD) time since CIDP diagnosis was 8.8 (9.3) years and the most common CIDP subtype was typical CIDP (67.8%). In addition, most patients (78.8%) had experienced a CIDP relapse in the past and more than half (58.3%) had a relapse in the previous year. Almost all patients (95.2%) had experienced at least one CIDP-related symptom within the previous 7 days, with one or more of the following symptoms being the most prevalent: tingling in the arms and legs (69.2%), fatigue (69.9%), and gradual weakening of the arms and legs (55.5%). The mean (SD) INCAT disability score reported by patients was 3.4 (2.5), representing moderate activity limitation of the arms and legs. The mean (SD) R-ODS centile score was 59.2 (19.9), reflecting moderate limitation of activity and social participation. The most common comorbidities were hypertension (35.6%), depression (22.6%), obesity (19.9%), and anxiety (19.2%) (Table 3).

Table 3.

Patients’ disease and treatment characteristics

	Patients (N = 146)	Patients under care of caregivers (N = 52)
Time since CIDP diagnosis,a years, mean (SD)	8.8 (9.3)	6.8 (9.9)
CIDP subtype, n (%)
Typical CIDP	99 (67.8)	35 (67.3)
Distal CIDP	23 (15.8)	4 (7.7)
Multifocal CIDP	9 (6.2)	6 (11.5)
Motor CIDP	3 (2.1)	0 (0.0)
Sensory CIDP	3 (2.1)	1 (1.9)
Focal CIDP	–	3 (5.8)
Unknown/not sure	9 (6.2)	3 (5.8)
Comorbidities,b,c n (%)
Hypertension	52 (35.6)	10 (19.2)
Depression	33 (22.6)	13 (25.0)
Obesity	29 (19.9)	7 (13.5)
Anxiety	28 (19.2)	11 (21.2)
Asthma	25 (17.1)	5 (9.6)
Other autoimmune disease	23 (15.8)	6 (11.5)
Gastrointestinal disease	22 (15.1)	5 (9.6)
Thyroid disease	20 (13.7)	1 (1.9)
Diabetes (type 1 or 2)	18 (12.3)	6 (11.5)
Other	26 (17.8)	7 (13.5)
CIDP-related symptoms experienced in the past 7 days,c n (%)
Tingling in arms and legs	101 (69.2)	38 (73.1)
Fatigue	102 (69.9)	39 (75.0)
Gradual weakening of arms and legs	81 (55.5)	33 (63.5)
Burning or stabbing pain	71 (48.6)	19 (36.5)
Loss of reflexes	70 (47.9)	19 (36.5)
Loss of balance and ability to walk	67 (45.9)	23 (44.2)
Loss of feeling in arms and legs	59 (40.4)	18 (34.6)
Double vision	18 (12.3)	6 (11.5)
Difficulty swallowing	13 (8.9)	7 (13.5)
Other	17 (11.6)	4 (7.7)
Did not have any symptoms	7 (4.8)	3 (5.8)
INCAT disability scored
Mean (SD)	3.4 (2.5)	4.4 (2.9)
Median (IQR)	3 (2–4)	4 (2–6)
Arm disability score
Mean (SD)	1.7 (1.4)	2.3 (1.6)
Median (IQR)	2 (1–2)	2 (1–2)
Leg disability score
Mean (SD)	1.8 (1.4)	2.2 (1.6)
Median (IQR)	1 (1–3)	2 (1–4)
R-ODS score,e mean (SD)	59.2 (19.9)	48.8 (20.9)
Received IVIG, n (%)	144 (98.6)	43 (82.7)
Received SCIG, n (%)	45 (30.8)	16 (30.8)
Location of infusion,c n (%)
At infusion clinic or doctor’s office	48 (32.9)	22 (42.3)
At home by a HCP	45 (30.8)	16 (30.8)
At home by patient and/or a caregiver	42 (28.8)	13 (25.0)
In the hospital	26 (17.8)	14 (26.9)
Mean time spent travelling to infusion clinic or hospital, n (%)	N = 66f	N = 31g
≤ 30 min	32 (48.5)	17 (54.8)
> 30 min to ≤ 1 h	25 (37.9)	11 (35.5)
> 1 h to ≤ 2 h	5 (7.6)	3 (9.7)
> 2 h	4 (6.1)	0 (0.0)

CIDP chronic inflammatory demyelinating polyradiculoneuropathy, HCP healthcare professional, INCAT Inflammatory Neuropathy Cause and Treatment Disability Scale, IVIG intravenous immunoglobulin, IQR interquartile range, R-ODS Rasch-Built Overall Disability Scale, SCIG subcutaneous immunoglobulin, SD standard deviation

^aIf CIDP diagnosis year was available but month was missing, CIDP diagnosis month was imputed as July

^bExperienced by > 10% patients in either group

^cCategories were not mutually exclusive

^dINCAT is a measure of activity limitation of the arms and legs, ranging from 0 (no signs of disability) to 12 (most severe disability score)

^eR-ODS is a linearly weighted scale that captures activity and social participation limitations, ranging from 0 (most severe activity and social participation restrictions) to 100 (no activity and social participation limitations)

^fOnly collected among patients who had received IVIG but not started SCIG

^gCalculated for caregivers who provided average time spent travelling to infusion clinic or hospital

Nearly all patients (98.6%) had received IVIG treatment previously (mean [SD] duration 6.3 [5.8] years) and 30.8% had received SCIG (mean [SD] duration 2.9 [2.3] years). Approximately 60% of patients received IVIG or SCIG at home (with 30.8% receiving treatment from an HCP and 28.8% by self-administration or from a caregiver), or at an infusion clinic or doctor’s office (32.9%). Among those who travelled to an infusion clinic or hospital to receive infusions, 86.4% of patients spent less than an hour in travelling time on average (Table 3).

Caregivers and Patients Under Their Care

The mean (SD) ages of caregivers and the patients under their care were 52.3 (14.4) and 56.4 (15.0) years, respectively. More than half of the caregivers (63.5%) and the patients (53.8%) were female. Half of the caregivers were working (full-time, part-time, or self-employed), although most of the patients (76.9%) being cared for by these caregivers were not in the work force (Table 1). All caregivers considered themselves to be the primary caregiver of the patients, with 46.2% being the spouse, 15.4% parents/step-parents, and 11.5% the children/stepchildren of the patients. Most of the caregivers (84.6%) had provided care for the patients for longer than a year.

Disease and treatment characteristics of the patients cared for by the caregivers were generally similar to those of the patient cohort (Table 3).

Physicians

Half (50.6%) of the participating physicians were practising in an academic institute or medical center and had been treating patients with CIDP for a mean (SD) of 17.9 (7.2) years with a mean (SD) of 24 (27) patients treated over the past 3 years. All physicians had prescribed corticosteroids for their patients with CIDP, 98.8% had prescribed IVIG, and 58.0% had prescribed SCIG. The most common reasons for not prescribing SCIG included access/availability issues (60.0%), lack of monitoring at home (48.6%), patients’ ability to self-administer their infusions (42.9%), and lack of information on efficacy and/or safety of SCIG (34.3%) (Table 2).

HRQoL Burden of Patients and Caregivers

Mean (SD) EQ-5D-3L utility index and EQ-5D VAS scores in the patient cohort were 0.68 (0.23) and 62.3 (19.4), respectively. Patients reported that CIDP prevented them (quite a bit or very much) from performing activities (50.7%), affected their social activities with family and friends (39.7%), and imposed financial burden (37.7%). The mean (SD) WPAI total activity impairment pertaining to CIDP was 51.1% (26.8). Among patients who were currently employed, mean (SD) WPAI total productivity impairment pertaining to CIDP was 37.9% (26.6) (Table 4).

Table 4.

Burden of CIDP for patients and caregivers

	Patients (N = 146)	Caregivers (N = 52)
EQ-5D-3L, mean (SD)
EQ-5D-3L utility indexa	0.68 (0.23)	0.79 (0.28)
EQ-5D VAS scoreb	62.3 (19.4)	76.8 (15.6)
WPAI pertaining to CIDPc
Absenteeism, % (SD)	11.9 (21.5)	11.3 (21.7)
Presenteeism, % (SD)	32.0 (22.3)	26.1 (17.5)
Total productivity impairment, % (SD)	37.9 (26.6)	30.8 (21.0)
Total activity impairment, % (SD)	51.1 (26.8)	45.8 (23.3)
Prevented from performing activities by CIDP, n (%)
Not at all	8 (5.5)	11 (21.2)
A little	35 (24.0)	11 (21.2)
Somewhat	29 (19.9)	14 (26.9)
Quite a bit	44 (30.1)	11 (21.2)
Very much	30 (20.5)	5 (9.6)
Social activities with family and/or friends affected by CIDP, n (%)
Not at all	20 (13.7)	9 (17.3)
A little	38 (26.0)	15 (28.8)
Somewhat	30 (20.5)	13 (25.0)
Quite a bit	37 (25.3)	10 (19.2)
Very much	21 (14.4)	5 (9.6)
Financial burden imposed on self and family by CIDP, n (%)
Not at all	23 (15.8)	13 (25.0)
A little	32 (21.9)	13 (25.0)
Somewhat	36 (24.7)	7 (13.5)
Quite a bit	32 (21.9)	10 (19.2)
Very much	23 (15.8)	9 (17.3)

CIDP chronic inflammatory demyelinating polyradiculoneuropathy, SD standard deviation, WPAI Work Productivity and Activity Impairment questionnaire

^aEQ-5D-3L summary index reflects how good or bad a health state is according to the preferences of the general population of the specific country/region. The maximum index score of 1 indicates the best health state

^bEQ-5D VAS records the patient’s self-rated health on a vertical visual analogue scale ranging from 0 to 100, in which 100 is labelled “The best health you can imagine” and 0 is labelled “The worst health you can imagine”

^cAbsenteeism, presenteeism, and total productivity impairment were calculated for participants who were currently employed (N = 58/146 patients; N = 24/52 caregivers), and total activity impairment was calculated for all participants. WPAI outcomes were expressed as impairment percentages, with higher numbers indicating greater impairment and less productivity

Among caregivers, mean (SD) EQ-5D-3L utility index and EQ-5D VAS scores were 0.79 (0.28) and 76.8 (15.6), respectively. More than a quarter (28.8%) of the caregivers reported that CIDP affected (quite a bit or very much) their social activities with family and/or friends, while 36.5% reported that CIDP imposed financial burden on themselves and their family. The mean (SD) WPAI total activity impairment pertaining to CIDP of the caregivers was 45.8% (23.3). Among caregivers who were employed, mean (SD) WPAI total productivity impairment pertaining to CIDP was 30.8% (21.0) (Table 4).

Stated Preferences for SCIG Treatment

All identified attributes had a significant role in the stated preferences for SCIG treatment decisions of patients, caregivers, and physicians. Across all participants, internal validity checks indicated that the quality of data collected was high (success rate 95% for the dominant profile test and 78.9% for the test–retest assessment) (Table S3).

Patients

From the perspective of the patients included in the study, the attribute with the highest preference weight was efficacy (lower chance of CIDP symptom worsening in 6 months), followed by treatment frequency (less frequent infusions), and location of infusion (by HCP or by self/caregiver at home) (Fig. 2).

Fig. 2.

Participants’ preferences for SCIG treatment attributes according to DCE. *p < 0.05, **p < 0.01, ***p < 0.001. Note: A positive part-worth utility indicates that patients prefer a profile with this level over the reference level of a particular attribute, while a negative part-worth utility indicates that patients prefer the reference level over this level. Values are reported with 95% confidence intervals. CI confidence interval CIDP chronic inflammatory demyelinating polyradiculoneuropathy, DCE discrete choice experiment

Treatment efficacy (chance of CIDP symptom worsening in 6 months of 10% vs 20%) accounted for the highest choice decision weight (38.9%), followed by treatment frequency (receiving treatment every 3–4 weeks vs once a week; 25.7%), and number of infusion sites (single vs multiple infusion sites; 15.2%) (Fig. 3). Location of infusion was valued similarly by patients when an infusion was administered by the patient or caregiver at home versus a HCP in their home, accounting for 3.3% of the decision weight. However, if “by healthcare professional in a clinic” was selected as the anchor profile, the location of infusion (self/caregiver at home vs HCP in a clinic) accounted for the largest choice decision weight (33.8%), followed by treatment efficacy (26.6%), and treatment frequency (17.6%) (Fig. S1).

Fig. 3.

Relative importance of attributes of SCIG treatment for CIDP from the perspective of A patients, B caregivers, or C physicians. CIDP chronic inflammatory demyelinating polyradiculoneuropathy, SCIG subcutaneous immunoglobulin

Caregivers

From the perspective of caregivers included in the study, the attribute with the highest preference weight was efficacy (lower chance of CIDP symptom worsening in 6 months), followed by location of infusion (by HCP at home—stronger preference—or by the patient or caregiver at home) (Fig. 2).

Treatment efficacy (chance of CIDP symptom worsening in 6 months of 10% vs 20%) accounted for the highest choice decision weight (42.0%), followed by number of infusion sites used during each infusion (single vs multiple infusion sites; 25.4%), and safety (chance of an infusion site reaction of 20% vs 30%; 12.7%). In contrast to patients’ perspectives, treatment frequency (once every 3–4 weeks vs once a week) only accounted for 8.5% of the decision weight for the caregivers. Caregivers’ preferences for location of infusion were largely similar regarding infusion by patient/caregiver or by a HCP at home, accounting for 8.4% of the decision weight (Fig. 3). However, if “by healthcare professional in a clinic” was selected as the anchor profile, location of infusion by self/caregiver at home was strongly preferred over infusion by HCP in a clinic, accounting for the second highest choice decision weight (29.1%) (Fig. S1).

Physicians

From the physicians’ perspective, the attribute with the highest preference weight was efficacy (lower chance of CIDP symptom worsening in 6 months), followed by treatment frequency (once every 3–4 weeks vs once a week), time spent on infusion in a month (for fewer hours), and location of infusion (by patient or caregiver at home). A non-significant trend towards a preference for administration by a healthcare provider in a clinic compared with by a healthcare provider at home was also noted (Fig. 2).

Treatment efficacy (chance of CIDP symptom worsening of 10% vs 20%) accounted for the highest choice decision weight (37.8%), followed by location of infusion (infusion by patient or caregiver at home vs by healthcare provider at home; 19.1%), and time spent on infusion in a month (2 vs 8 h; 16.0%) (Fig. 3). Anchoring the location of infusion on “infusion by healthcare professional in a clinic” had a negligible impact on the relative importance of attributes among physicians (Fig. S1).

Treatment Preferences in Patient Subgroups

Results from the exploratory subgroup analyses are included in Table 5. When stratified by age (< 60 vs ≥ 60 years), older patients had a stronger preference for less frequent treatment administration (once every 2 weeks or once every 3–4 weeks vs 2–3 times a week) compared with younger patients. In contrast, younger patients had a stronger preference for receiving the infusion by a healthcare provider at home (vs in a clinic) compared with older patients.

Table 5.

Stratification of patients’ preferences for SCIG treatment attributes according to DCE

	Patients (N = 146)
	Age, years			Sex			CIDP type			Prior SCIG treatment experience			INCAT score
Attributes	Subgroup	Part-worth utilitya	p value	Subgroup	Part-worth utilitya	p value	Subgroup	Part-worth utilitya	p value	Subgroup	Part-worth utilitya	p value	Subgroup	Part-worth utilitya	p value
Treatment frequency
2–3 times a week	Reference			Reference			Reference			Reference			Reference
Once a week	< 60	0.337 (0.06, 0.61)	0.091	Male	0.639 (0.37, 0.91)	0.288	Typical	0.624 (0.39, 0.86)	0.053	Experienced	0.476 (0.11, 0.84)	0.696	0–1	0.776 (0.36, 1.19)	0.148
	≥ 60	0.660 (0.40, 0.92)		Female	0.486 (0.22, 0.75)		Non-typical	0.242 (− 0.07, 0.56)		No experience	0.555 (0.33, 0.78)		≥ 2	0.438 (0.23, 0.65)
Once every 2 weeks	< 60	0.725 (0.44, 1.01)	0.045*	Male	0.873 (0.59, 1.16)	0.527	Typical	1.063 (0.83, 1.30)	0.005*	Experienced	0.782 (0.41, 1.16)	0.392	0–1	1.284 (0.85, 1.72)	0.056
	≥ 60	1.120 (0.85, 1.39)		Female	1.033 (0.76, 1.31)		Non-typical	0.441 (0.08, 0.80)		No experience	0.973 (0.74, 1.20)		≥ 2	0.813 (0.60, 1.03)
Once every 3–4 weeks	< 60	0.648 (0.32, 0.97)	0.020*	Male	0.884 (0.57, 1.20)	0.622	Typical	1.191 (0.91, 1.47)	< 0.001*	Experienced	0.829 (0.39, 1.26)	0.623	0–1	1.361 (0.87, 1.85)	0.033*
	≥ 60	1.174 (0.87, 1.48)		Female	1.086 (0.76, 1.41)		Non-typical	0.177 (− 0.21, 0.57)		No experience	0.969 (0.71, 1.23)		≥ 2	0.785 (0.54, 1.03)
Number of infusion sites during each infusion
Requires multiple infusion sites	Reference			Reference			Reference			Reference			Reference
Option to have one infusion site	< 60	0.214 (0.06, 0.37)	0.750	Male	0.120 (− 0.04, 0.28)	0.060	Typical	0.246 (0.11, 0.38)	0.894	Experienced	0.085 (− 0.12, 0.29)	0.093	0–1	0.113 (− 0.12, 0.35)	0.255
	≥ 60	0.251 (0.10, 0.40)		Female	0.316 (0.17, 0.47)		Non-typical	0.228 (0.04, 0.42)		No experience	0.296 (0.17, 0.42)		≥ 2	0.267 (0.15, 0.39)
Chance of CIDP symptom worsening in 6 months
35%	Reference			Reference			Reference			Reference			Reference
20%	< 60	0.542 (0.31, 0.78)	0.484	Male	0.288 (0.06, 0.51)	0.028*	Typical	0.634 (0.44, 0.83)	< 0.001*	Experienced	0.448 (0.14, 0.75)	0.711	0–1	0.611 (0.27, 0.96)	0.366
	≥ 60	0.427 (0.22, 0.64)		Female	0.673 (0.45, 0.90)		Non-typical	0.077 (− 0.20, 0.35)		No experience	0.519 (0.33, 0.70)		≥ 2	0.438 (0.26, 0.61)
10%	< 60	1.120 (0.88, 1.36)	0.804	Male	1.047 (0.81, 1.28)	0.567	Typical	1.238 (1.03, 1.44)	0.004*	Experienced	1.145 (0.83, 1.46)	0.868	0–1	1.003 (0.64, 1.36)	0.681
	≥ 60	1.076 (0.85, 1.30)		Female	1.181 (0.95, 1.41)		Non-typical	0.722 (0.44, 1.01)		No experience	1.123 (0.93, 1.32)		≥ 2	1.099 (0.92, 1.28)
Time spent on infusion per month lePara>
15 h	Reference			Reference			Reference			Reference			Reference
8 h	< 60	0.367 (0.16, 0.57)	0.120	Male	0.112 (− 0.11, 0.33)	0.053	Typical	0.243 (0.07, 0.42)	0.176	Experienced	0.403 (0.13, 0.68)	0.294	0–1	0.249 (− 0.07, 0.57)	0.910
	≥ 60	0.137 (− 0.07, 0.34)		Female	0.420 (0.22, 0.62)		Non-typical	0.460 (0.18, 0.74)		No experience	0.236 (0.06, 0.41)		≥ 2	0.272 (0.11, 0.43)
2 h	< 60	0.515 (0.27, 0.76)	0.123	Male	0.150 (− 0.10, 0.40)	0.017*	Typical	0.392 (0.19, 0.59)	0.807	Experienced	0.419 (0.09, 0.75)	0.915	0–1	0.315 (− 0.04, 0.67)	0.603
	≥ 60	0.248 (0.02, 0.48)		Female	0.561 (0.33, 0.79)		Non-typical	0.434 (0.12, 0.75)		No experience	0.407 (0.21, 0.61)		≥ 2	0.405 (0.22, 0.59)
Chance of infusion site reaction
40%	Reference			Reference			Reference			Reference			Reference
30%	< 60	0.195 (− 0.01, 0.40)	0.552	Male	0.078 (− 0.13, 0.29)	0.290	Typical	0.226 (0.05, 0.40)	0.192	Experienced	0.031 (− 0.24, 0.31)	0.221	0–1	0.124 (− 0.19, 0.43)	0.772
	≥ 60	0.112 (− 0.09, 0.31)		Female	0.252 (0.05, 0.45)		Non-typical	0.025 (− 0.24, 0.29)		No experience	0.235 (0.06, 0.41)		≥ 2	0.173 (0.01, 0.33)
20%	< 60	0.360 (0.13, 0.59)	0.500	Male	0.180 (− 0.05, 0.41)	0.105	Typical	0.363 (0.17, 0.56)	0.496	Experienced	0.201 (− 0.11, 0.51)	0.368	0–1	0.379 (0.04, 0.72)	0.682
	≥ 60	0.254 (0.05, 0.46)		Female	0.468 (0.25, 0.68)		Non-typical lePara>	0.254 (− 0.02, 0.53)		No experience	0.358 (0.18, 0.54)		≥ 2	0.296 (0.12, 0.47)
Location of infusion
By a healthcare provider in a clinic	Reference			Reference			Reference			Reference			Reference
By a healthcare provider at home	< 60	1.058 (0.83, 1.29)	0.005*	Male	0.979 (0.75, 1.21)	0.083	Typical	0.794 (0.60, 0.98)	0.476	Experienced	1.072 (0.78, 1.37)	0.044*	0–1	0.705 (0.37, 1.04)	0.444
	≥ 60	0.610 (0.40, 0.82)		Female	0.698 (0.48, 0.91)		Non-typical	0.899 (0.62, 1.18)		No experience	0.726 (0.54, 0.91)		≥ 2	0.853 (0.68, 1.03)
By yourself and/or a caregiver at home	< 60	0.761 (0.54, 0.98)	0.771	Male	0.949 (0.72, 1.18)	0.048*	Typical	0.843 (0.66, 1.03)	0.051	Experienced	1.685 (1.37, 2.00)	< 0.001*	0–1	0.847 (0.51, 1.18)	0.608
	≥ 60	0.805 (0.59, 1.02)		Female	0.691 (0.48, 0.90)		Non-typical	0.496 (0.22, 0.77)		No experience	0.422 (0.24, 0.60)		≥ 2	0.750 (0.58, 0.92)

CIDP chronic inflammatory demyelinating polyradiculoneuropathy, DCE discrete choice experiment, INCAT Inflammatory Neuropathy Cause and Treatment Disability Scale, SCIG subcutaneous immunoglobulin

*Statistically significant

^aA positive part-worth utility indicates that patients prefer a profile with this level over the reference level of a particular attribute, while a negative part-worth utility indicates that patients prefer the reference level over this level. Values are reported with 95% confidence intervals

Women had a stronger preference for higher treatment efficacy (significant for a 20% lower chance of CIDP symptom worsening in 6 months vs a 35% chance), and for less time spent on infusion (strongly prefer spending 2 h in a month vs 15 h) compared with men. In contrast, men had a stronger preference for administering the infusion by themselves or by a caregiver at home (vs a healthcare provider in a clinic) compared with women.

When stratified by CIDP type, patients with typical CIDP had stronger preferences for less frequent treatment administration (once every 2 weeks or once every 3–4 weeks vs 2–3 times a week) and higher treatment efficacy (a 10% or 20% chance vs a 35% chance of CIDP symptom worsening) compared with patients with non-typical CIDP.

In terms of treatment experience, treatment preferences between SCIG-experienced patients and those without prior SCIG experience were generally comparable. However, patients with SCIG experience had a stronger preference for receiving an infusion at home by a healthcare provider or by themselves and/or a caregiver (vs receiving it in a clinic) compared with patients with no prior SCIG experience.

Patients with less severe disability (i.e., an INCAT score of 0–1) had a stronger preference for less frequent treatment administration (particularly once every 3–4 weeks vs 2–3 times a week) compared with patients with more severe disability (i.e., an INCAT score ≥ 2).

Discussion

This study reports insights into the burden of disease experienced by patients with CIDP and caregivers, and their preferences for key attributes of SCIG therapy using a survey that incorporated DCE questions.

Our study highlighted the substantial burden associated with CIDP among patients, with fewer than half being employed despite the population, on average, being of working age (mean 58.1 years). On average, patient participants had a disease duration of 9 years, more than half reported a relapse in the past year, and over 90% had experienced some CIDP-related symptoms in the past 7 days. The burden experienced by patients had a negative impact on their ability to participate in physical and social activities, and limited their ability to work, which, together, imposed an additional financial burden. High disease, treatment, and financial burdens associated with CIDP were reported in a global survey (N = 595) that also highlighted the need for treatments that improve HRQoL [15]. The patient population for the current study was broadly similar to those in other studies that reported a high burden of CIDP and reasons for preferring SCIG treatment over IVIG for those that require long-term treatment [2], and after switching from IVIG to SCIG [16]. In the current study, all caregivers reported that they considered themselves to be the primary care provider. Caregivers experienced an impact on social and family activities, and ability to perform work-related tasks. The burden of disease experienced by patients and by caregivers likely shapes their perception of available treatment options that drive treatment choice.

The main drivers of treatment preferences were attributes related to efficacy and convenience. For all participant groups, efficacy (chance of CIDP symptom worsening in 6 months) was found to be the most important factor in SCIG treatment decisions according to the DCE. The importance of other treatment attributes varied across the participant groups. For example, lower treatment frequency (once every 3–4 weeks vs once a week) was of a higher priority for patients than for caregivers or physicians. Additionally, the location of infusion was more important to patients and caregivers than physicians. In contrast, a shorter time spent on infusion was preferred by physicians compared with patients and caregivers.

It should be emphasized that the treatment attributes in the DCE represent hypothetical scenarios and do not provide evidence of clinical benefit of any particular treatment. Several options are available for Ig therapy, including facilitated SCIG (fSCIG) that incorporates a high-purity recombinant form of hyaluronidase, which transiently increases the permeability of subcutaneous tissue to IgG that allows high-volume IgG administration (equivalent to volumes administered intravenously) into the subcutaneous tissue over a short time [17]. fSCIG 10% (HYQVIA, Takeda Pharmaceuticals USA Inc., MA, USA; HyQvia, Baxalta Innovations GmbH, Vienna, Austria) was approved in 2024 in the USA as a maintenance therapy for adults, and in the EU as a maintenance therapy for CIDP after stabilization with IVIG for patients of all ages [6, 18]. fSCIG can be administered once every 3–4 weeks (at the same dose and frequency as the patient’s prior IVIG therapy if applicable) [6, 18]. The dosing profile of fSCIG coupled with the opportunity for self-administration at home may be preferred by some patients over SCIG, which warrants further investigation.

The results of this study showed that treatment preferences of patients, caregivers, and physicians were in broad overall alignment across several parameters. However, there were some notable differences across participating groups, which reveal that effective dialogue between patients, caregivers, and physicians can support a shared decision-making approach. A shared decision-making process is a concept that considers both the perspectives of HCPs and patients, which can offer the potential to optimize management strategies for each patient on an individualized basis [19]. In the setting of IgG therapy for CIDP, the recent availability of fSCIG may provide greater flexibility in treatment choice, because it may enable less frequent infusions and use of fewer infusion sites than conventional SCIG at an equivalent monthly dose [17]. Real-world evidence in other approved indications for therapeutic IgG suggests that, although patients have a tendency to prefer their current treatment administration route over routes with which they have no experience, they may look to switch from IVIG or conventional SCIG to fSCIG if they encounter clinical problems with their current treatment, or when seeking greater convenience and flexibility in dosing [20, 21]. A systematic review reported that 16 of 18 studies that compared IVIG with SCIG found that patients preferred the subcutaneous administration route, and this was also the case in all six studies that were conducted in pediatric populations; the primary driver of patient preference for SCIG was related to the option of home administration due to convenience and comfort, as well as the avoidance of the need for travel to attend hospital appointments [21]. However, caution should always be applied when extrapolating findings of patient experience between different indications for SCIG/fSCIG therapy. Some of the discordance in treatment preferences between physicians and their patients observed in this study underscores the value of appreciating the views of all relevant stakeholders and encouraging active communication between all parties when considering the effects of treatment preferences on outcomes in any chronic disease.

In an effort to better understand preferences in patient cohort, we performed an exploratory subgroup analysis of preferences that revealed that older patients preferred lower treatment frequencies, while younger patients preferred receiving infusions at home by a healthcare provider. Women prioritized higher treatment efficacy and less time spent on infusion, whereas men preferred administration of infusions by themselves or a caregiver at home. In addition, patients with typical CIDP and those with SCIG favored lower treatment frequencies and higher treatment efficacy compared with their counterparts with non-typical CIDP or with no experience of SCIG therapy. Patients’ choices reflect the frequency and setting of treatment administration, which may be determined by factors that impact their ability to self-infuse SCIG (e.g., age, level of disability), differing personal attitudes towards SCIG, and lifestyle factors. In this respect, the importance of patient views and choices independent of families, caregivers, and physicians should be considered. Heterogeneity in preferences for use of IgG replacement therapy in primary immune deficiency diseases, even within the patient population, underlies calls for a formal shared decision-making process [22], with preliminary tools to support this process now being developed as a decision-making aid for primary immunodeficiency [23].

Regulatory agencies and health technology assessment bodies are increasingly recognizing the importance of understanding patient preferences to inform their decision-making [24, 25]. DCE has, in recent years, become the most commonly used method for quantification of preferences in the healthcare setting from the perspective of a range of different stakeholders [26]. DCE is a valuable tool for this purpose as it simulates real-world scenarios rather than focusing on treatment attributes in isolation, as in ranking exercises, for example [26]. As such, DCE represents an appropriate and robust approach to explore preferences for treatment attributes of SCIG in our study. The success rates of the dominant profile test and the test–retest assessment, performed as an internal validity check, were in accordance with those reported from other DCE studies (mean reported success rates 80% and 76%, respectively, vs 95% and 79% in our study) [27]. Similar methodology was also adopted in a recent study that evaluated the preferences of patients with primary antibody immune deficiency disorders for attributes of IVIG and SCIG across different patient characteristics [22]. The authors suggested that the DCE tool developed to evaluate treatment attributes in their study could be adjusted to help document patients’ views in a clinical setting to facilitate shared decision-making [22].

Possible limitations of this study include the generalizability of the data. For example, the selected options for SCIG treatments for CIDP included in the DCE questions do not necessarily reflect experience in the real world. In addition, patients and caregivers were recruited from a patient advocacy group and a recruiting agency and may not have been fully representative of respective real-world populations. Patients provided a self-reported diagnosis of CIDP rather than that established by a physician, which could potentially lead to misdiagnosis. Assessment of patient, caregiver, and physician preferences for SCIG was limited to the included treatment attributes; additional treatment attributes that were not specified in the survey may have affected the results of the preference analysis. For example, the efficacy of SCIG treatment is generally similar and in this study is defined narrowly in terms of “chance of CIDP symptoms worsening”; hence, the preference weight of efficacy may have been underestimated. In addition, it should be noted that the chance for CIDP symptom worsening of 10% vs 20% is hypothetical and does not reflect clinically established differences between SCIG treatment options. The assessment of additional efficacy variables, such as improvement among patients who are non-relapsing, could have provided a more comprehensive perception of treatment efficacy. Participants’ eligibility and all data in the survey were self-reported and may have been subject to recall bias; clear explanations and definitions for terms that may have subjective interpretation were provided to minimize this bias, and no time restrictions were applied for completion of the survey questions. Caregivers completed the INCAT assessment and R-ODS questionnaire to provide an impression of the degree of disability among patients under their care; however, it should be noted that INCAT and R-ODS are not validated for proxy administration. Finally, caution should be exercised when interpreting treatment preferences because the limited size of study sample limited the exploratory subgroup analyses to investigate potential impacts of participants characteristics on preference decision-making for CIDP treatments.

Conclusion

This study underscores the importance of shared decision-making in optimizing treatment strategies because preferences for treatment attributes varied between patients and caregivers compared with physicians. Treatment preferences among patients, caregivers, and physicians are primarily driven by efficacy and convenience, and treatment frequency. Findings from this study also highlight the significant burden patients with CIDP and their caregivers experience, including impacts on their physical and social activities, work productivity, and financial stability.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This study was funded by Takeda Development Center Americas, Inc. The journals’ Rapid Service fees were also funded by Takeda Development Center Americas, Inc.

Data Availability

The data sets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of Interest

Min Yang, Danni Yang, and Mo Zhou are employees of Analysis Group, Inc., which received funding from Takeda Development Center Americas, Inc. to perform this study. Danni Yang and Mo Zhou are former employees of Analysis Group, Inc. Miriam L Zichlin and Hakan Ay are employees of Takeda Development Center Americas, Inc. and are Takeda shareholders. Chelsey Fix and Lisa Butler are employees of GBS/CIDP Foundation International. Colin Anderson-Smits is a former employee of Takeda Development Center Americas, Inc. Current affiliation, Gilead Sciences, Foster City, CA, USA.

Ethical Approval

The study was conducted in accordance with the Helsinki Declaration of 1964 and its later amendments, and was granted exemption from a full review by the Pearl Institutional Research Board according to FDA 32 Code of Federal Regulations (CFR) 56.104 and 45CFR46.104(b)(2): (2) Tests, Surveys, Interviews on 01/11/2022 (amendment approved on 03/31/2022). Informed consent was obtained from participants prior to interviews and online survey completion. Participants received an honorarium payment for their time spent and effort participating in this research.