Evaluating the burden of pruritus due to atopic dermatitis in Japan by patient-reported outcomes

Abstract

Aims

Although pruritus is a hallmark feature of atopic dermatitis, no study has investigated the associated impact of pruritus, due to atopic dermatitis in Japan. The study aimed to evaluate the real-life burden of atopic dermatitis by pruritus severity in adult Japanese patients. The primary objective was to assess the correlation between pruritus severity and work productivity and activity impairment. A secondary objective was to characterize the impact of pruritus on quality of life and to evaluate the burden of symptoms severity and frequency.

Materials and methods

A cross-sectional internet-based survey was conducted. Eligible patients were currently employed and working adults with atopic dermatitis for at least 6 months. Stratification on pruritus severity assessed by the Worst Pruritus Numerical Rating Scale at the screening was performed to ensure that different severity groups are represented. Correlations were assessed using Spearman’s rank-order correlation coefficient. Multivariate regression analyses were performed to assess the impact of pruritus severity on work productivity and quality of life.

Results

The study analyzed 370 patients. Pruritus severity significantly correlated with work impairment (Rho = 0.622, P value (H0: Rho > 0.5) <.001). A greater pruritus severity was associated with greater work productivity and activity impairment and a greater impact on quality of life, sleep, emotional state, and everyday activities. Patients with a greater pruritus severity carried a higher economic burden of treatment and were more often not satisfied with the received therapy.

Limitations

All data were self-reported by patients via an online survey, which is associated with the risk of misclassification for diagnosis, recall bias, and limited participation of patients.

Conclusions

The study provides evidence that pruritus is associated with the overall disease burden and impacts many important life aspects of patients with atopic dermatitis in Japan.

Keywords:

• Atopic dermatitis
• pruritus
• disease burden
• patient-reported outcomes
• COVID-19

JEL CODES:

• I10
• I1
• I
• I18

Introduction

Atopic dermatitis (AD) is a chronic inflammatory skin disease that can have repetitive exacerbations and remissions¹. AD is one of the most common chronic conditions. Global prevalence is approximately 230 million with variance across countries and geographic regions^2,3. In Japan, the total number of patients with AD is estimated as 450,000 and it has demonstrated the increasing trend⁴. The prevalence of adult AD patients ranged between 2.1 and 4.9% in an international, cross-sectional, web-based survey performed in Japan and other countries⁵. Another survey on 4,826 Japanese adults reported a prevalence of AD to be 6.8%⁶. AD usually develops by 5 years of age^1,7 and it may resolve during childhood, however, in 20–50% of patients it may become chronic and persist through adulthood⁸. Recent studies suggested that the prevalence of AD in adults may be higher than previously recognized^9–11.

The main treatment goal, according to the most recent Japanese guidelines for AD, is to either eliminate skin problems or reach a stable state with mild symptoms without exacerbations to enable the daily functioning of patients⁸. Unfortunately, symptoms remain uncontrolled in a significant proportion of patients¹² and persistent clearance of skin lesions is not achieved by most patients with moderate to severe AD at baseline¹³.

Patients with AD suffer from frequent and intense pruritus^14,15 – the core symptom of AD, which can result in marked sleep deprivation, anxiety and depression, impaired quality of life, and reduced daily productivity^16–18. Almost two-thirds of patients with moderate to severe atopic dermatitis report that pruritus is present most of the time, 12 h a day or more¹⁹. In addition, pruritus in AD differs from pruritus in other skin diseases such as psoriasis, scabies, or urticaria in the frequency of other associated symptoms or influence of modulating factors^14,20. Alleviating pruritus in AD is an important treatment goal for the patients^16,21.

Several studies reported the disease burden of AD in Japan^22–25. Based on research investigating the burden of pruritic skin disease, negative effects of the disease on productivity and quality of life have been indicated^18,26. However, there have been no studies evaluating the impact of pruritus due to AD in Japan, and it has been difficult to estimate how much pruritus due to AD affects the patients and their daily life.

Thus, this study investigated the specific impact of pruritus due to AD on the lives of Japanese patients by evaluating the real-life disease burden by pruritus severity. The primary objective was to assess the correlation between pruritus severity and work productivity and activity impairment. A secondary objective was to characterize the impact of pruritus on quality of life (including sleep, daily activities, and emotional state), symptom severity, and symptom frequency. Additionally, the impact of COVID-19 on working conditions and healthcare resource use is also addressed in our study.

Methods

Study design

We conducted a cross-sectional study using an online survey for data collection.

Study population and data source

Potential participants were randomly selected from a patient web panel owned by Intage Inc., a medical market research firm based in Tokyo^27–29. The panel contains approximately 669,058 patients and 15,571 patients who are undergoing any treatment for AD throughout Japan (as of August 2019). Patients who agreed to participate in the survey had to be currently employed and currently working adults (18–60 years old), have a diagnosis of AD for at least 6 months, be of Japanese nationality, and provide informed consent. The diagnosis of AD was based on patients’ self-reports of being diagnosed with AD by a physician. To maintain the data quality, the time of questionnaire completion was assessed and patients with an inappropriately short time were excluded, as well as patients with contradictory answers on pruritus severity scales. We also excluded patients with concomitant dermatologic or systemic conditions associated with pruritus to isolate the effect of pruritus due to AD (Supplementary Table S1). Patients were blinded to the research question when invited to take part in the study and could withdraw from the survey at any time.

The data collection was conducted during weekdays on October 26–27, 2020. To ensure balanced representation, we performed stratification on pruritus severity at the screening. At the stage of entering the data, electronic quality control was performed to avoid missing data.

Outcomes and covariates

Pruritus severity and disease burden were evaluated with several patient-reported outcome instruments. Pruritus severity was assessed with the Worst Pruritus Numerical Rating Scale (WP-NRS), which assesses pruritus severity on a scale from 0 (no itch) to 10 (worst itch imaginable) ^30–32. To ensure that different severity groups are represented, WP-NRS was used to enroll patients with different levels of pruritus severity at screening adopting the Itch (Pruritus) Numerical Rating Scale severity levels developed in a study by Vakharia et al. A pruritus score of 0–3 corresponds to mild self-reported pruritus severity, 4–6 to moderate and 7–10 to severe symptoms, respectively³³.

The impact on work productivity was measured with the Work Productivity and Activity Impairment Questionnaire: Atopic Dermatitis V2.0 (WPAI: AD). This validated instrument assesses total work productivity impairment, calculated from the combined effects of “absenteeism” (i.e. percentage of work hours missed) and “presenteeism” (i.e. percentage of impairment while working), as well as total activity impairment other than work due to AD during the previous week³⁴. Quality of life was evaluated with the Dermatology Life Quality Index (DLQI) which has been used for over a decade to assess and compare the quality of life of patients with different dermatologic conditions³⁵ and recommended by the Harmonising Outcome Measures for Eczema VII consensus meeting³⁶. This instrument consists of 10 questions assessing the impact of skin disease on symptoms and feelings, daily activities, leisure, work and school, personal relationships, and the impact of treatment over the previous week^37,38. The impact due to AD was evaluated with the Atopic Dermatitis Impact Scale (ADerm-IS), which assesses the impact of AD on sleep, daily activities (household, physical, and social activities and difficulty concentrating), and emotional state (feelings of self-consciousness, embarrassment, and sadness) ^30,39.

The intensity and frequency of AD symptoms were measured with the Atopic Dermatitis Symptom Scale (ADerm-SS) and the Patient-Oriented Eczema Measure (POEM), respectively^30,39,40. ADerm-SS is a questionnaire that assesses the severity of signs and symptoms of moderate to severe adult AD, including itch, skin pain, skin cracking, flaking, thickening, bleeding, and oozing, as well as rash and skin dryness^30,39. The ADerm-SS total score was calculated as the sum of all 11 items of the questionnaire. POEM is a validated tool designed to measure and monitor patient-reported symptom severity of AD^40,41.

Additional data were collected with a series of stand-alone questions on sociodemographic characteristics (e.g. age, sex), employment status, smoking status, AD diagnosis, and exclusionary conditions, atopic dermatitis duration, flares frequency, and intensity over the last 6 months, current and recent treatment during the last 6 months, hospitalization, healthcare resource use in the last 6 months, and overall satisfaction with the treatment. We also included questions that addressed the impact of the COVID-19 pandemic on patients’ life, work, and healthcare resource use. We asked whether there have been any changes since the COVID-19 pandemic (e.g. associated with commuting to work, online meetings, change in revenue, duration of sleep, time spent with family, increased frequency of hands washing) and if this change had a beneficial or detrimental influence on controlling AD symptoms. Two clinical experts (H.M., Y.K.) were involved in the selection of clinically relevant outcomes and variables evaluated in this study.

Before conducting the main study, the online questionnaire was tested on a pilot sample of 10 patients of the target population to verify the questionnaire’s understandability. Patients with contradictory answers related to pruritus severity (≥4-point difference) between the WP-NRS and ADerm-SS scales were excluded. A 4-point threshold was selected since in a validation study of Peak Pruritus NRS the most appropriate threshold for defining a clinically relevant, within-person response was a ≥ 2-4-point change^16,30. Also, in other studies, the change of 4 points in WP-NRS was considered the minimum change to show a clinically meaningful improvement^30,42.

The study was conducted according to the Declaration of Helsinki. Ethical approval was not an obligation for this observational study because the study used fully anonymized data of voluntarily preregistered respondents. The Japanese Ethical Guidelines for Medical and Health Research Involving Human Subjects do not apply to studies exclusively using anonymized data⁴³.

Statistical analysis

Sample size and justification

No data are available in the literature to define the expected correlation coefficient between pruritus severity (assessed with WP-NRS) and work productivity and activity impairment (assessed with WPAI percent overall work impairment due to AD). Hence, a correlation coefficient above 0.5 with a 10-point variability of the 95% confidence interval [0.4; 0.6] was proposed as a basis to calculate the sample size. Assuming a statistical power of 80%, a minimum sample size of 219 patients is necessary. Due to there being the uncertainty of the final number of eligible patients in a cross-sectional study and on the confidence interval, the sample size has been inflated to account for those limitations and to ensure the appropriate power in the final analysis.

Analytic approach

Descriptive statistics were used to describe the sociodemographic characteristics, clinical characteristics, healthcare resource use, quality of life, work productivity impairment, and COVID-19 pandemic influence on AD symptoms and working conditions. Bivariate statistical analysis was conducted to determine the statistical association between pruritus severity as measured with WP-NRS and the productivity measured with WPAI. Spearman’s rank-order correlation coefficient was calculated, and the two-sided student’s t test was performed at the significance level α = 0.05 to test whether the correlation is at least equal to 0.5. A multivariate regression analysis (using generalized linear models) was performed to assess the impact of pruritus severity on work productivity (WPAI-AD percent overall work impairment), quality of life (DLQI total score), and ADerm-IS domain scores, controlling for differences in patient characteristics.

In each model, the main covariates included sociodemographic characteristics, employment status, and clinical characteristics, such as the number of years with AD, history of hospitalization for AD, current and recent treatment, and flares frequency and intensity over the last 6 months. The final list of variables included in the model was determined using a stepwise selection process based on the lowest Akaike Information Criterion and based on clinical expert opinion. Data management and analyses were performed with the statistical software SAS version 9.4; in particular the PROC GLIMMIX for the generalized linear models.

Results

In this study, 7,059 patient panel members were invited by email to participate and 1,736 responded (survey response rate 24.6%). Of them, 22 were not considered due to discrepancy with the registered panel information, and 1,714 were included in the screening stage of the survey and assessed based on the inclusion and exclusion criteria. Among 1,714 potential participants, 73.7% (n = 1,264) were considered ineligible at the screening stage because they did not provide informed consent; could not confirm Japanese nationality, current employment, or being diagnosed with AD for at least 6 months; or they had concomitant dermatologic or systemic conditions associated with pruritus. The number of patients who met the selection criteria assessed at the stage of screening and entered the survey was 450. The median time of survey completion was approximately 8 min, and 99% of patients took approximately 3.5 min — none of the participants had an abnormally low completion time. Eighty participants were excluded due to contradictory answers on severity scales. In the analysis, 370 patients met all eligibility criteria and were included.

The overview of study population characteristics is presented in Table 1. Approximately half of the patients were male (55.9%). Most of the patients (68.1%) were older than 40 years; the mean age of study patients was 44.3 years. The mean time of AD symptoms duration was 30.8 years. The majority of patients in the total study population were employed full-time and worked at an office. In a group of patients with severe pruritus, 63.4% worked full-time compared with 45.4% of patients with mild pruritus. The mean severity of pruritus in the overall study population assessed with a WP-NRS score was 4.6. There were 141, 117, and 112 patients in the groups with mild, moderate, and severe pruritus, respectively.

Table 1. Characteristics of the study population (total population and stratified by disease severity).

Characteristic		All	Mild	Moderate	Severe
		(n = 370)	(n = 141)	(n = 117)	(n = 112)
Age, years	Mean (SD)	44.3 (9.4)	45.3 (9.1)	43.5 (9.4)	44.0 (9.6)
	18–30	37 (10.0)	9 (6.4)	13 (11.1)	15 (13.4)
	31–40	81 (21.9)	31 (22.0)	30 (25.6)	20 (17.9)
	41–50	135 (36.5)	53 (37.6)	37 (31.6)	45 (40.2)
	51–60	117 (31.6)	48 (34.0)	37 (31.6)	32 (28.6)
Sex	Male	207 (55.9)	71 (50.4)	70 (59.8)	66 (58.9)
	Female	163 (44.1)	70 (49.6)	47 (40.2)	46 (41.1)
Employment status	Employed full-time	205 (55.4)	64 (45.4)	70 (59.8)	71 (63.4)
	Employed part-time	118 (31.9)	52 (36.9)	36 (30.8)	30 (26.8)
	Self-employed/business owner	21 (5.7)	11 (7.8)	4 (3.4)	6 (5.4)
	Freelance and professional workers	15 (4.1)	7 (5.0)	4 (3.4)	4 (3.6)
	Directors of companies, organizations, public corporations, etc	11 (3.0)	7 (5.0)	3 (2.6)	1 (0.9)
Smoking status	Current smoker	61 (16.5)	21 (14.9)	18 (15.4)	22 (19.6)
	Former smoker	86 (23.2)	30 (21.3)	30 (25.6)	26 (23.2)
	Never smoker	223 (60.3)	90 (63.8)	69 (59.0)	64 (57.1)
Current work style	Working from home	53 (14.3)	21 (14.9)	14 (12.0)	18 (16.1)
	Working at office	282 (76.2)	103 (73.0)	95 (81.2)	84 (75.0)
	Working outside/apart from the office (e.g. visiting customers)	22 (5.9)	9 (6.4)	7 (6.0)	6 (5.4)
	Others	13 (3.5)	8 (5.7)	1 (0.9)	4 (3.6)
Time of atopic dermatitis symptoms duration, years	Mean (SD)	30.8 (13.8)	29.0 (13.9)	30.9 (12.9)	33.0 (14.4)
Time since AD diagnosis, years	Mean (SD)	29.0 (13.9)	27.1 (13.9)	29.6 (13.3)	30.9 (14.3)

Values are presented as n (%) unless noted otherwise.

Abbreviations. AD, atopic dermatitis; SD, standard deviation.

Work productivity and activity impairment in relation to pruritus severity

Pruritus severity correlated with overall work impairment (Spearman’s rank-order correlation coefficient Rho = 0.622). The correlation coefficient was significantly greater than 0.5 (p < .001). The mean overall work impairment was 5.2% in patients with mild pruritus, 21.7% and 38.1% in the groups of patients with moderate and severe pruritus, respectively. Patients with moderate or severe pruritus had significantly higher WPAI-AD scores compared with patients with mild pruritus (p < .001; Table 2).

Table 2. Overall work impairment was evaluated with the WPAI-AD†.

	All (n = 365‡)	Mild (n = 137)	Moderate (n = 116)	Severe (n = 112)
Mean (SD)	20.5 (24.1)	5.2 (12.4)	21.7 (20.3)	38.1 (26.2)
P value§	_	Ref.	<0.001	<0.001
1–20	242 (66.3)	128 (93.4)	74 (63.8)	40 (35.7)
21–40	47 (12.9)	3 (2.2)	24 (20.7)	20 (17.9)
41–60	50 (13.7)	6 (4.4)	14 (12.1)	30 (26.8)
61–80	24 (6.6)	0 (0.0)	3 (2.6)	21 (18.8)
81–100	2 (0.6)	0 (0.0)	1 (0.9)	1 (0.9)

Abbreviations. SD, standard deviation; WPAI-AD, Work Productivity and Activity Impairment Questionnaire: Atopic Dermatitis V2.0.

Values are provided as¬ n (%) unless noted otherwise.

†WPAI-AD score represents the percentage of overall work impairment.

‡Five patients reported that their working time in the preceding week was 0; therefore, the percentage of the overall work impairment score could not be calculated for those 5 patients (including 4 from the mild group and one from the moderate group).

§P value was generated using analysis of variance.

The multivariate model suggests that the impact of AD on work productivity is associated with pruritus (p < .001; Table 3). On average, a 4-point increase on WP-NRS was associated with a 19.7-point increase in overall work impairment after adjusting for other factors. Other statistically significant factors include monthly average copayment of AD treatment, part-time employment, and treatment with moisturizers during the last 6 months.

Table 3. Impact on work productivity: a generalized linear model of WPAI-AD percent overall work impairment.

Covariates	Category	Coefficient	Standard error	95% CI	P value
Intercept		−13.515	7.647	−28.58; 1.55	.078
WP-NRS		4.925	0.572	3.80; 6.05	< .001
Age, years	18–30	Ref.
	31–40	2.737	4.841	−6.80; 12.28	.572
	41–50	2.850	5.027	−7.06; 12.76	.571
	51–60	3.754	5.184	−6.46; 13.97	.470
Sex	Male	Ref.
	Female	5.787	3.094	−0.31; 11.88	.063
Employment status	Employed full-time	Ref.
	Employed part-time	−7.108	3.177	−13.37; −0.85	.026
	Other†	1.228	4.018	−6.69; 9.14	.760
Smoking status	Current smoker	−0.870	3.638	−8.04; 6.30	.811
	Former smoker	−0.825	3.219	−7.17; 5.52	.798
	Never smoker	Ref.
Number of years with AD	0–15	Ref.
	16–30	−6.155	4.054	−14.14; 1.83	.130
	31–45	−3.569	4.037	−11.52; 4.39	.378
	46–59	−1.459	5.240	−11.78; 8.87	.781
History of hospitalization for AD since the start of the disease	Yes	16.986	14.609	−11.80; 45.77	.246
	No	Ref.
Current and recent historical treatment received for AD during the last 6 months‡	Topical steroids	3.461	4.007	−4.43; 11.36	.389
	Moisturizers	6.132	2.747	0.72; 11.54	.027
	Tacrolimus (topical)	2.707	3.221	−3.64; 9.05	.401
	Antihistamines (oral)	2.197	2.742	−3.21; 7.60	.424
	Oral steroids	3.039	5.985	−8.75; 14.83	.612
Current work style	Working from home	Ref.
	Working at office	1.230	3.526	−5.72; 8.18	.728
	Working outside/apart from the office (e.g. visiting customers)	−2.521	5.795	−13.94; 8.90	.664
	Other	3.083	8.107	−12.89; 19.06	.704
Monthly average copayment of AD treatment (in ¥1,000)		0.631	0.203	0.23; 1.03	.002
How many times have the exacerbations of the problems with your skin occurred during the last 6 months?	0	Ref.
	1	−1.113	4.069	−9.13; 6.90	.785
	2	0.758	4.429	−7.97; 9.48	.864
	3+	4.279	3.539	−2.69; 11.25	.228

AD: atopic dermatitis; WPAI-AD, Work Productivity, and Activity Impairment Questionnaire: Atopic Dermatitis V2.0; WP-NRS, Worst Pruritus Numerical Rating Scale.

†Other: Includes self-employed/business owner, freelance and professional workers, and directors of companies, organizations, public corporations, etc.

‡For binary variables, the reference is no treatment with the specific treatment method (e.g. moisturizer users compared with moisturizer nonusers).

Clinical burden

AD clinical course and treatment pattern within the last 6 months

In patients with mild pruritus, 47.5% (n = 67) of patients during the last 6 months did not experience any flares and 21.3% (n = 30) only experienced one flare, whereas in 64.3% (n = 72) of patients with severe pruritus experienced three or more flares. In most patients, acute symptoms required adding treatment with TCSs or tacrolimus (Table 4). In the mild pruritus group, 44.7% (n = 63) of the patients did not receive any treatment for AD in the last 6 months and 18.0% (n = 14) were not receiving the treatment at the time of the study, while each portion in the severe pruritus groups was 22.3% (n = 25) and 6.9% (n = 6), respectively. Most patients receiving treatment used TCSs (n = 228, 88.4%) and skin moisturizers (n = 169, 65.5%). A considerable number of patients used oral antihistamines (44.8%, 20.4%, and 26.9% of patients in the groups with severe, moderate, and mild pruritus, respectively) and topical tacrolimus (21.8%, 22.6%, and 14.1% of patients, respectively). (Supplementary Table S2).

Table 4. Flares (frequency and intensity) in the last 6 months (total population and stratified by pruritus severity).

Question	Answer	All (n = 370)	Mild (n = 141)	Moderate (n = 117)	Severe (n = 112)
How many times have the exacerbations of the problems with your skin occurred during the last 6 months?	Mean (SD)	7.5 (22.9)	3.6 (11.7)	5.6 (12.5)	14.4 (36.4)
How long on average did the skin problems last (days)?	Mean (SD)	22.6 (38.0)	15.1 (31.7)	20.6 (36.3)	30.3 (43.0)
	No exacerbations	112 (30.3)	67 (47.5)	27 (23.1)	18 (16.1)
	1–5	84 (32.6)	31 (41.9)	26 (28.9)	27 (28.7)
	6–10	79 (30.6)	27 (36.5)	32 (35.6)	20 (21.3)
	>10	95 (36.8)	16 (21.6)	32 (35.6)	47 (50.0)
Did it require adding treatment with topical corticosteroids or tacrolimus?	Not applicable (no exacerbations)	112 (30.3)	67 (47.5)	27 (23.1)	18 (16.1)
	Yes	209 (81.0)	54 (73.0)	75 (83.3)	80 (85.1)
	No	49 (19.0)	20 (27.0)	15 (16.7)	14 (14.9)
Did it require adding systemic treatment (administered orally or with injection)?	Not applicable (no exacerbations)	112 (30.3)	67 (47.5)	27 (23.1)	18 (16.1)
	Yes	62 (24.0)	12 (16.2)	18 (20.0)	32 (34.0)
	No	196 (76.0)	62 (83.8)	72 (80.0)	62 (66.0)

Values are presented as n (%) unless noted otherwise.

Abbreviation. SD, standard deviation.

Satisfaction with the treatment

Among patients who received treatment currently, treatment satisfaction differed depending on pruritus severity. Only 4.7% of patients with mild pruritus were dissatisfied with their current treatment, whereas 29.6% of patients with severe pruritus reported dissatisfaction (p < .001). Lack of symptoms control was the most common reason in all groups for being unsatisfied with therapy (Table 5, Supplementary Table S3, and Supplementary Table S4).

Table 5. Satisfaction with the current treatment.

Question	Answer	All	Mild	Moderate	Severe
		(n = 370)	(n = 141)	(n = 117)	(n = 112)
Are you satisfied with the current medical care you receive?	No treatment received currently	140 (37.8)	77 (54.6)	32 (27.3)	31 (27.7)
	Treatment received currently	230	64	85	81
	Satisfied	93 (40.4)	39 (60.9)	28 (32.9)	26 (32.1)
	Neutral	101 (43.9)	22 (34.4)	48 (56.5)	31 (38.3)
	Not satisfied	36 (15.7)	3 (4.7)	9 (10.6)	24 (29.6)
	P value†	_	Ref.	0.003	< 0.001
Why are you unsatisfied with the medical care?‡	Not applicable (No treatment received currently or satisfied with treatment)	233 (63.0)	116 (82.3)	60 (51.3)	57 (50.9)
	Lack of symptoms control	73 (53.3)	9 (36.0)	33 (57.9)	31 (56.4)
	Safety concerns	15 (11.0)	4 (16.0)	3 (5.3)	8 (14.6)
	Lack of explanations by physicians/nurses	3 (2.2)	0 (0.0)	2 (3.5)	1 (1.8)
	Tolerability issues	3 (2.2)	0 (0.0)	2 (3.5)	1 (1.8)
	Insufficient convenience of use	8 (5.8)	1 (4.0)	4 (7.0)	3 (5.4)
	Treatment costs	17 (12.4)	5 (20.0)	6 (10.5)	6 (10.9)
	Other	18 (13.1)	6 (24.0)	7 (12.3)	5 (9.1)

Values are presented as n (%) unless noted otherwise.

†P value was generated using the Chi-square test.

‡All percentages are calculated from the total number of patients in each column.

Patients with severe pruritus who were not satisfied with their treatment (n = 24) presented a numerically higher average in all aspects of burdens measured in this study compared with patients with mild or moderate pruritus (Supplementary Table S5). This trend was observed particularly in the frequency and intensity of flares (Supplementary Table S6).

AD skin symptoms

In patients who reported severe pruritus, skin symptoms of AD were significantly more aggravated. The POEM score, assessing the frequency of skin symptoms during the last 7 days, was on average 5.1 in the group of patients with mild pruritus, 9.6 and 14.8 in the groups of patients with moderate and severe pruritus, respectively, and the difference was statistically significant (p < .001). The ADerm-SS score, evaluating the intensity of skin symptoms during the last 24 h, was on average 11.9 in the group of patients with mild pruritus, 32.1 and 57.5 in the groups of patients with moderate and severe pruritus, respectively, and the difference between groups was statistically significant (p < .001; Supplementary Figure S1).

Humanistic burden

Quality of life

The mean total DLQI score was significantly lower in the group with mild severity of pruritus (1.8) compared with the groups with moderate (4.2) and severe (8.1) pruritus, respectively (p < .001). Most patients with mild pruritus reported that the skin problems did not affect their life (n = 86; 61.0%), whereas, in the group of patients with severe pruritus, it was declared only by 9.8% of patients (n = 11). For most patients with severe pruritus, disease symptoms had a moderate (n = 39; 34.8%), large (n = 23; 20.5%), or very large (n = 5; 4.5%) influence on the quality of life (Supplementary Figure S2). The multivariate model suggests that the impact of AD on quality of life is associated with pruritus (p < .001; Supplementary Table S7).

Impact on sleep, daily activity, and emotional state

For the impact on sleep, daily activity, and emotional state, assessed with ADerm-IS, the mean scores were 1.6, 1.6, and 1.9, respectively, in the group of patients with mild pruritus, and 12.9, 11.0, and 13.5, respectively, in the group of patients with severe pruritus. Patients with moderate or severe pruritus had significantly higher ADerm-IS scores compared with patients with mild pruritus (p < .001; Supplementary Table S8). The multivariate model suggests that the impact of AD on sleep, daily activities, and emotional state is associated with pruritus (p < .001; Supplementary Table S9).

Economic burden

Monthly average copayment of AD treatment

The monthly average copayment of AD treatment was ¥1,443 (13.6 United States Dollars [USD]) in the group of patients with mild pruritus compared with ¥3,414 (32.2 USD) in the group with severe pruritus (Table 6)⁴⁴.

Table 6. Monthly average copayment of AD treatment (¥).

	All	Mild	Moderate	Severe
	(n = 370)	(n = 141)	(n = 117)	(n = 112)
Mean (SD)	2,468 (5,397)	1,443 (2,043)	2,797 (6,180)	3,414 (7,021)
Median (min; max)	1,500 (0.0; 60,000)	1,000 (0.0; 15,000)	1,800 (0.0; 60,000)	2,000 (0.0; 60,000)

Abbreviations. AD, atopic dermatitis; ¥, Japanese Yen; SD, standard deviation.

Impact of COVID-19 pandemic on working conditions and healthcare resource use

Most patients (78.1%) in the overall study population reported that there have been no changes in their work before and after the beginning of the COVID-19 pandemic. There was a significant difference in the proportion of patients who reported a change due to COVID-19 in the group with severe pruritus compared with patients with mild pruritus (31.2% vs. 14.2%; p = .001) (Supplementary Table S10). In terms of health resource use, most patients (87.3%) in the overall study population reported that there was no difference in the frequency of visits to the hospital/clinic or the amount or type of drugs prescribed before and after the beginning of the COVID-19 pandemic; 17.0% of patients with severe pruritus declared that it had changed since the COVID-19 pandemic, while only 8.5% declared such a change in the group with mild pruritus (Supplementary Table S11).

Discussion

The present study sought to evaluate the correlation between pruritus severity and work productivity and activity impairment and the impact of pruritus due to AD on the lives of Japanese patients by evaluating the real-life disease burden by pruritus severity.

As reported, pruritus is a core symptom of AD, and we used the WP-NRS to capture the severity of pruritus. Unlike other common measures to assess signs and symptoms of AD such as POEM, the WP-NRS is designed to evaluate the worst pruritus in the previous 24 h more reliably in terms of recall bias compared to other self-reported measures requiring longer recall (e.g. the past 7 days for POEM). Thus, this simpler scale was used in this study to allow a more rapid assessment of AD disease severity, which may benefit busy clinical practice.

The study showed a statistically significant correlation of pruritus severity with the percent overall work impairment measured with WPAI-AD. The main endpoint of our study, the Spearman's correlation coefficient Rho of 0.622, is interpreted as a moderate correlation in the literature^45,46 and it supports the interpretation that there is a close relation between pruritus and work productivity¹⁸ and more severe pruritus due to AD is associated with lower work productivity. Based on the average annual income of full-time employed workers in Japan, this overall work impairment was translated into the loss of ¥278,570 (2,628.0 USD) per patient with mild pruritus, ¥1,165,907 (10,999.1 USD) per patient with moderate pruritus, and ¥2,048,942 (19,329.6 USD) per patient with severe pruritus annually⁴⁷. The estimation supported the results of a previous study²⁴.

As the secondary objective, the study assessed the influence of pruritus severity on quality of life, as well as on skin symptoms and life aspects, including sleep, daily activities, and emotional state. The results showed that pruritus severity was independently associated with all the evaluated outcomes. Greater pruritus severity was consistently related to worsening of quality of life, skin symptoms, sleep, emotional state, and everyday activities.

The relationship between patient-reported outcomes (PROs) and various patients’ characteristics and relevant factors influencing work productivity were explored in this study. The monthly average copayment of AD treatment was independently associated with all outcomes under this study, suggesting that bearing monthly costs of treatment is a burden for patients and affects their quality of life in many different dimensions. Patients with severe pruritus reported higher monthly average copayments of AD treatment compared with patients with mild pruritus, implying a higher direct economic burden for patients with greater pruritus severity. Compared with part-time work, full-time employment has a negative influence on work productivity. It is understandable because full-time workers generally engage in their jobs for a longer time than part-time workers, and therefore they have more chances to be affected by pruritus. Treatment with moisturizers was significantly associated with overall work impairment compared with moisturizer nonusers, indicating the burden of topical treatment, which shore up the results in the previous studies^48,49.

The study also showed that patients with severe pruritus were more often dissatisfied with the received treatment; had other symptoms of AD more aggravated; and had greater work productivity and activity impairment, poorer quality of life, and worse quality of sleep, effects on daily activities, and emotional state compared with patients with mild pruritus, with all differences showing statistical significance. Patients with severe pruritus who were not satisfied with their treatment (n = 24) presented a higher average in all aspects of burdens compared with patients with mild or moderate pruritus. These results require careful interpretation because this study did not evaluate causal association and patients who are not satisfied with the treatment may have lower adherence, resulting in severer pruritus. Nonetheless, this finding may be helpful to uncover the existing unmet needs of patients with AD.

In this study, pruritus has significant socioeconomic consequences, and the disease burdens were significantly higher at greater levels of pruritus severity. Previous findings from a study conducted in Germany with 800 patients with active chronic skin diseases showed a similar level of correlation of pruritus severity with activity impairment due to pruritus (Rho = 0.506, p < .001) and overall work productivity loss due to pruritus (Rho work – presenteeism Rho = 0.508, p < .001), thereby increasing the certainty of our study. Patients with moderate to very strong pruritus experienced significant impairment in quality of life and sleep, as well as impaired sexual life, depressive mood, and anxiety⁵⁰.

A previous study that included Japanese patients showed a significant correlation between AD symptoms severity and total work productivity impairment score, as well as DLQI scores. The linear correlation in this study was weaker (r = 0.387) than the rank-based correlation (which can suggest a monotonic relationship) identified in our study (Rho = 0.617). A substantial difference is that the study population was patients visiting hospitals, and disease severity was assessed by physicians with the Severity Scoring of Atopic Dermatitis scale, which includes objective clinical signs. This assessment could be in discrepancy with symptom intensity assessed solely from patients’ perspectives. In our study, both symptom severity and work impairment were assessed based on patients’ perspectives using PRO measures, which might be the reason why our study showed a higher correlation between them²⁵.

Despite increasing physicians’ efforts to improve health outcomes, some patients were dissatisfied with the effectiveness of their current treatment, which may suggest the need for more efficacious and safe treatments for AD symptoms, especially pruritus^51,52. Recent advances in research to understand the phenotype and mechanisms of AD have provided new therapeutic options⁵³. Enhancing shared decision-making between patients and physicians to reflect each patient preference may lead to better treatment outcomes and higher satisfaction.

Strengths and limitations

Strengths of the study include sufficient sample size and that data were collected in a real-world setting, not in a controlled environment of a clinical trial. The survey was conducted only on weekdays to reflect as much as possible the work-related reality of patients with AD pruritus. Using PRO measures allowed us to capture patient perception, which is crucial in the assessment of pruritus due to the subjective nature of this symptom. The PRO instruments used in our study — WPAI-AD, ADerm-SS, ADerm-IS, and WP-NRS (conceptually equivalent to Peak Pruritus NRS) — are well-defined, validated, reliable, and sensitive scales.

Since the observational study design is associated with a high risk of bias, we applied methods to address bias and confounding at both the design and analysis stages of the study. To reduce the risk of selection bias that might occur for patients completing the surveys, patients were blinded to the study’s research question when invited. All data were self-reported and not confirmed with medical records, which raises the question of misclassification for diagnosis and recall bias. However, questionnaires included in the research tool have well-established psychometric properties of reliability and validity and include a clear and interpretable scoring system that limits the risk of misinterpretation and misclassification. To address confounding and isolate the effect of pruritus due to AD, we excluded patients with concomitant dermatologic or systemic conditions associated with pruritus and we analyzed the impact of pruritus due to AD on various PROs controlling for the difference in patient characteristics in the multivariate regression analysis.

The study was conducted online, which, due to a potential lack of accessibility, may have limited patient participation in the survey. Nevertheless, our study reflected aspects of the general population. It showed a similar sex ratio to a previous database analysis in Japan and a similar proportion of patients with AD with no treatment received as in the Study on the Necessity and Large-scale Epidemiological Research for Measures against Allergic Diseases^22,54.

To assure an equal proportion of participants with different severity levels of pruritus to show the impact of pruritus due to AD, stratification of patients was conducted at the stage of patients screening and based on the WP-NRS scale. Therefore, the distribution of pruritus severity level in the study population does not reflect the general population; however, within severity strata, the study population may represent each level of pruritus severity in Japanese patients with AD since the results showed that greater severity presented a higher burden consistently.

This study population did not include AD patients who are unable to work due to AD or other reasons. Hence, the impact on these patients could not be evaluated. The disease burden in these patients may be greater and should be investigated in future studies.

Lastly, the survey of this study was conducted during the middle of the COVID-19 pandemic at the end of October 2020, and it was difficult to assume how the unprecedented situation affects the study. Proportions of patients with AD who responded that they have changes in working conditions and healthcare resource use were, however, only 21.9% and 12.7%, respectively. Even if some patients were affected by the COVID-19 pandemic, the impact of the crisis on the study results was considered limited.

Conclusions

This is the first study evaluating the relationship between pruritus severity and work productivity and the impacts of pruritus due to AD in patients with AD in Japan, thus providing a comprehensive characterization on several aspects of pruritus burden in Japanese patients with AD.

Transparency

Declaration of funding

AbbVie sponsored the study; contributed to the study design; participated in the interpretation of data; in reviewing, and approval of the final version. No honoraria or payments were made for authorship.

Declaration of financial/other interests

HM received consulting fees and/or speaker honoraria from Japan Tobacco, Maruho, Kaken Pharmaceutical, Sanofi, Mitsubishi Tanabe Pharma, Kyowa Kirin, Torii Pharmaceutical, Lily, AbbVie, Taiho Pharma, Bristol‐Myers Squibb, Novartis, Sato Pharmaceutical, Nippon Zoki Pharmaceutical, Sun Pharma, and Teikoku Seiyaku. YK has no conflict of interest to declare. KI and IK are employees of AbbVie GK and may own AbbVie stock. BMC is an employee of AbbVie Inc and may own AbbVie stock. ML and MT are employees of Creativ-Ceutical, which conducted analyses and received consultancy fees from AbbVie GK.

Peer reviewers on this manuscript have received an honorarium from JME for their review work but have no other relevant financial relationships to disclose.

Author contributions

HM, YK, KI, BMC, ML, MT and IK contributed to the study design, the acquisition, analysis, or interpretation of data for the work, drafted the work or revised it, approved the submitted version, and agreed to be accountable for all aspects of the work.

Acknowledgements

The authors acknowledge the technical writing assistance of Sylvaine Barbier and Marwa Mezghani, employees of Creativ-Ceutical in the preparation of this manuscript, which was funded by AbbVie.