Impact of itch and skin pain on quality of life in adult patients with atopic dermatitis in Japan: results from a real-world, point-in-time, survey of physicians and patients

Abstract

Objective

Itch is a common symptom of atopic dermatitis (AD), however, there is limited evidence of the frequency and association of skin pain alongside itch. This study assessed the incremental dual burden and impact of itch and skin pain on satisfaction, quality of life and work productivity in patients with AD in Japan.

Methods

Data were drawn from the 2020 Adelphi AD Disease Specific Programme, a point-in-time survey of dermatologists (n = 56) and their patients with history of moderate/severe AD (n = 265). Patients were grouped accordingly: no itch/skin pain (No I/SP, reference group, n = 89), itch/no skin pain (I-only, n = 71), and itch and skin pain (I + SP, n = 26). Descriptive analyses were performed alongside a range of regression models, dependent on outcome variables.

Results

I + SP patients had a 4.97-point worse POEM score (p = .005) and 14.5% more overall work impairment (p = .034) versus the reference group. I-only and I + SP patients were 8.92 and 23.5 times more likely, respectively, to experience sleep disruption on a day-to-day basis (both p < .001). I + SP patients were 4.6 times more likely to be bothered by their symptoms (p = .034), had a mean EASI score 6.7 points higher (p = .008) and had 1.39 more areas affected (p = .001). I + SP patients were 7.26 times more likely to express dissatisfaction with lack of improvement in their condition and 8 times more likely to be dissatisfied with convenience of treatment (both p < .05).

Conclusion

This dissatisfaction, alongside variations in reported symptomatic burdens, suggests that physicians could consider alternative and/or novel therapeutic approaches for the management of both itch and skin pain.

PLAIN LANGUAGE SUMMARY

Patients with atopic dermatitis experience a broad range of symptoms, including both itch and skin pain, however it is not clear how these symptoms combine to impact patients in everyday life. This survey of 56 dermatologists and 265 patients with a history of moderate or severe atopic dermatitis investigated the impact of both itch and skin pain on quality of life and treatment satisfaction in Japan. Patients were categorised into three groups depending on the presence of these symptoms; patients with either no itch or skin pain, patients with itch but no skin pain, or patients with both itch and skin pain. These three groups were then compared. Patients with both itch and skin pain reported this combination of symptoms together impacted on their daily lives more than patients with itch but no skin pain, or no itch or skin pain. In particular, patients with both itch and skin pain had more areas of their body affected and reported being more bothered by their symptoms compared to those who did not experience these symptoms, with daily work impairment, sleep disruption and quality of life all worse. Importantly, patients with both itch and skin pain were more likely to be dissatisfied with the lack of improvement in their condition and with the convenience of their treatment. These results suggest that physicians should take into consideration the presence of both itch and skin pain when making treatment decisions in atopic dermatitis, and the need to consider treatments to target both symptoms.

Keywords:

• Atopic dermatitis
• itch
• quality of life
• skin pain
• treatment dissatisfaction

Introduction

Atopic dermatitis (AD) is a chronic relapsing dermatologic disorder characterized by skin barrier disruption and inflammation^1,2, and is one of the most-commonly observed skin diseases in Japan³. AD is characterized by acute periods of aggravated symptoms, commonly referred to as flares, alternating with periods of relative quiescence following treatment⁴. Pruritus (itch) and skin pain are two of the most common and burdensome symptoms^5–7. Although itch has been the subject of much research, Maarouf et al. suggested that pain symptoms had not been widely assessed and that patients may report AD-related pain either independently or concomitantly of itch⁸. More generally it has been shown that these symptoms greatly affect the quality of life (QoL) of patients with AD⁹.

Murota et al. found that even in mild cases of AD work performance decreased by approximately 10% during working hours. This rose to approximately 25% in moderate patients and to 44% in severe cases¹⁰ but the researchers did not discuss potential reasons behind this. Symptoms including skin pain and itch have independently been shown to be significantly worse in patients with AD than in those with psoriasis, particularly in severe disease, although the incremental burden of both itch and skin pain was not assessed¹¹. In addition, Yano et al. reported that the overall impact of AD on work productivity appeared to be greater than in other skin diseases such as psoriasis¹². They speculated that this could be linked with the higher frequency and severity of itch in AD but no research has focused on either this or the potential impact of AD skin pain on work loss.

Aside from itch and skin pain, other prevalent symptoms in AD include sleep disturbance^13,14, and symptoms relating to reduced mental and physical health^6,10,15. Compared with matched controls, adult AD patients in Japan have reported higher prevalence of concomitant conditions including arthritis, asthma, nasal allergies, anxiety, depression and sleep disorders, which are also associated with higher healthcare resource utilization and significantly poorer health related QoL¹⁵.

In addition to use of emollients, common AD treatments include topical steroids, calcineurin inhibitors and systemic anti-inflammatory therapies including gluco-corticosteroids and cyclosporine⁵. These conventional systemics may be used to address patient flares, but this may be limited by safety concerns. Treatment options increased with approval in 2018 of the first biologic treatment for AD in Japan.

Real-world patient-reported data on the specific impact of AD itch and skin pain in Japan are sparse¹³. It is known that AD patients in Japan experience a high disease burden due to symptoms affecting sleep, QoL and work productivity¹⁴. However, there is limited evidence of the frequency that skin pain is experienced alongside itch and their combined impact on AD patients in Japan. The aim of this study was to describe the burden of itch and skin pain in adult AD patients in Japan, with a focus on patient-reported data, and to assess the incremental dual burden of itch and skin pain in terms of impact on patients’ daily activities and overall QoL.

Methods

Study design

Data were drawn from the Adelphi Real World Atopic Dermatitis (AD) Disease Specific Programme (DSP), a cross-sectional survey of physicians and their adult AD patients conducted in Japan from April to September 2019. DSPs are large, multinational surveys designed to identify current disease management, and patient- and physician-reported disease impact, supplemented by the collection of historical information from medical records regarding treatment and prescribing patterns. A complete description of the methods of the survey has been previously published and validated^16–18.

Participants

Screening and recruitment of physicians reflected nationally representative samples subject to meeting DSP inclusion criteria. To be included dermatologists had to be actively involved in AD drug management and have a minimum monthly workload of 5 adult AD patients with moderate to severe disease (with at least 1 moderate and 1 severe patient).

Each physician was instructed to recruit the next 5 adult AD patients currently with, or with a history of, moderate to severe disease and complete a detailed patient record form (PRF) for each, based on information gathered during the consultation and also from medical records.

Patients were required to be aged 15 years or older, have current or history of moderate or severe AD (as defined by the treating physician’s overall subjective assessment and information from medical records) and were not currently involved in a clinical trial for AD. There was no restriction with regards to treatment received or on current severity.

Physician-recorded data

Data captured included patient demographics: age, sex, body mass index (BMI), and work status; and clinical characteristics including concomitant conditions, AD disease duration, current symptoms, pattern of disease (whether patient experiences day-to-day symptoms and/or flares), physician’s subjective assessment of current AD severity (mild, moderate, severe), symptoms experienced on a daily basis/during a flare and the severity of symptoms experienced at these times, presence and number of flares experienced, patient’s disease status (improving, stable, changeable or deteriorating rapidly/slowly), body regions affected (head and neck, upper limbs, trunk, lower limbs; maximum of four), body surface area affected (BSA), and all components of the Eczema Area and Severity Index (EASI), which measures both the extent (area) and severity of AD (range 0–72, higher scores indicating greater severity)¹⁹. Physicians rated AD severity in response to the question: “What is your overall assessment of the severity of AD symptoms in this patient currently based on your own definitions of mild, moderate and severe?”. Physicians also recorded their perspective of the impact of AD on the patient including emotional status and impairment of day-to-day activities and provided details of any currently and any previously prescribed AD therapies.

Patient-reported data

At the time of consultation, patients were invited to complete a voluntary patient self-completion form (PSC), independently of the physician. In order to assess real-world outcomes, PSCs collected data about patients’ condition including satisfaction with current treatment and presence of day-to-day symptoms (including itch and skin pain), followed by the frequency of symptoms they selected. For each symptom, patients reported whether the symptom was present on a day-to-day basis, “never,” “rarely,” “sometimes,” “regularly” or “all the time,” with patients categorised as having the symptom if they indicated experiencing it either “all the time” or “regularly.”

QoL was evaluated using the Dermatology Life Quality Index (DLQI; range 0–30, with higher scores indicating greater impact on QoL)²⁰. The Patient-Oriented Eczema Measure (POEM) evaluated the presence of AD signs and symptoms in the past week, and their impact on sleep (range 0–28, higher scores indicating greater severity)²¹. The Work Productivity and Activity Index (WPAI) evaluated AD effects on productivity during the past 7 days (range 0–100, higher scores indicating greater severity)²². Each patient’s questionnaire was completed in a single session, typically in the practice waiting room, immediately after the consultation and returned to the physician or receptionist in a sealed envelope. This ensures the physician does not see the patient’s responses and thus future interactions between the physician and patient are not influenced.

Ethics

All responses captured on the data collection forms were anonymized to preserve respondent (physician and patient) confidentiality and as such no personal identifiable information was requested or collected. All participating physicians and patients were assigned a study number to aid anonymous data collection, and to allow linkage of data during data collection and analysis. Using a check box, patients provided informed consent for use of their anonymized and aggregated data for research and publication in scientific journals. The Japanese data were drawn from a multi-country data source, collection of which received ethics waiver from the Western Institutional Review Board in the United States (study protocol number: AG-8382). Furthermore, data collection was undertaken in line with European Pharmaceutical Marketing Research Association (EphMRA)²³ guidelines and is fully consistent with the United States Department of Health and Human Services (HIPAA) privacy legislation²⁴. In line with these requirements no personal identifiable information was collected either from participating patients or physicians. All data were fully de-identified prior to analysis.

Analysis groups

Patients were classified in the following groups: No itch or skin pain (No I/SP), itch but no skin pain (I-only), skin pain but no itch (SP-only), and both itch and skin pain (I + SP). Patients were categorized by the symptom being present “all the time” or “regularly,” while, for the purpose of this analysis, symptom absence was defined by those who indicated they “never,” “rarely” or “sometimes” experienced symptoms. Analysis groups were therefore based on clinical characteristics and were independent of treatment received.

Statistical analysis

Numeric variables were described using number (n), mean, standard deviation (SD), median, interquartile range (IQR), minimum (min) and maximum (max). Categorical variables were described using n and percentage (%). Comparisons in descriptive analysis were made using bivariate tests; Analysis of Variance (ANOVA) for numeric variables, and Chi-squared (CH)/Fisher’s Exact (FE) tests for categorical variables. Pairwise comparisons were also conducted utilizing the same tests.

A range of regression models was used dependent on the outcome variable as follows: Logistic regression – where the outcome variable was dichotomous, such as presence or absence of symptoms (including sleep disturbance), and current treatment; ordered logistic regression – where the outcome variable was ordinal, for example the number of nights over the last week when the patient experienced sleep disturbance, and patient dissatisfaction with treatment; negative binomial regression – where the outcome variable was a non-negative count, for example number of body areas affected; and linear regression – where the outcome variable was numeric and was not a non-negative count, for example the patient-reported outcome (PRO) measures.

Each regression analysis comparing outcomes between the itch/skin pain groups was adjusted for age, gender, underlying disease severity, comorbid conditions measured by the Charlson Comorbidity Index (CCI) and current treatment prescribed.

For the three analysis groups, coefficients (linear regression), odds ratios (logistic regression) or incident rate ratios (negative binomial regression) were derived with associated p values and 95% confidence intervals. In addition, for linear and negative binomial regressions, regression adjusted means were calculated.

Standard errors in all regressions were adjusted to allow for intragroup correlation within reporting physician, relaxing the requirement that the observations be independent. That is, the patients were independent across physicians, but patients collected by the same physician may have correlated outcomes, so this was adjusted for in the regression models.

For regressions produced, where appropriate, the model was evaluated by considering either r-squared (linear regression) or McFadden’s pseudo r-squared (other models). Coefficients/incident rate ratios/odds ratios were examined for any unusual values/results. Residuals (for linear regression), or deviance residuals (logistic and negative binomial regression) were examined by comparing them with fitted values and producing QQ-plots. In addition, a test of the proportional-odds assumption (ordered logistic regression) was conducted. In a separate series of regression models, the above analyses were repeated but with the inclusion of the interaction term between treatment and the three analysis groups.

Data were analyzed using STATA version 16.1 (StataCorp LP, College Station, TX). The base of patients for analysis varied due to physicians or patients not answering specific questions. Any patients with missing values for a particular variable were removed from all analyses where that variable was used. However, patients were still eligible for inclusion in other analyses. Missing data was not imputed.

Results

Population/analysis groups

A total of 56 dermatologists provided PRFs for 265 patients, of whom 189 fully completed a PSC. In total 186 patients were included in the analysis: 89 No I/SP patients, 71 I-only patients and 26 I + SP patients. There were three patients experiencing SP-only, and while this number is too small to allow inclusion in a statistical analysis, the impact on such patients is covered in the discussion.

Demographics/clinical characteristics

The average age of all patients included in the analysis (n = 186) was 38.8 (SD 15.4) years old with 58.1% being male (Table 1). On average, the patients had been diagnosed with AD for 11.3 years and had a mean CCI of 0.1 (SD 0.3). The majority (95.7%) had a CCI score of 0. Approximately two thirds of patients (67.2%) were considered to have currently moderate (58.6%) or severe (8.6%) AD (in their physician’s subjective opinion). Mean BSA affected was 27.3% (SD 20.4), with a mean EASI score of 10.3 (SD 8.1) and, on average, 2.7 (SD 1.2) of the four body areas for which data were collected, were affected by AD. Just over two thirds of the patients (68.3%) were receiving topical or other non-advanced therapies, defined as topical corticosteroids, topical calcineurin inhibitors, prescribed emollients, phototherapy, wet wraps, bleach baths, antibiotics for skin infection, antihistamines or pain medication. Of the 31.7% of patients who were receiving advanced therapies (± topicals), 40.7% (i.e. 12.9% of the total sample) were receiving a systemic therapy (defined as injected/oral corticosteroids and immunosuppressants) with the remainder receiving biologic therapy (18.8% of the total sample). Note that there were no patients in the sample who were not currently receiving treatment.

Table 1. Physician-reported patient demographic and clinical characteristics.

	All patients (n = 186)	No I/SP patients (n = 89)	I-only patients (n = 71)	I + SP patients (n = 26)	p Value
Age, years, mean ± SD	38.8 ± 15.4	36.8 ± 16.2	40.9 ± 14.3	40.2 ± 15.0	.210 (AN)
Male sex, n (%)	108 (58.1)	50 (56.2)	42 (59.2)	16 (61.5)	.864 (CH)
Years since diagnosis^a	11.3 ± 11.1	11.6 ± 10.8	12.3 ± 10.3	7.6 ± 15.4	.624 (AN)
Mean ± SD, (n)	(n = 59)	(n = 30)	(n = 22)	(n = 7)
CCI score, n (%)
0	178 (95.7)	84 (94.4)	69 (97.2)	25 (96.2)
1+	8 (4.3)	5 (5.6)	2 (2.8)	1 (3.8)	.787 (FE)
Mean ± SD	0.1 ± 0.3	0.1 ± 0.4	0.0 ± 0.3	0.0 ± 0.2	.591 (AN)
Current severity, n (%)
Mild	61 (32.8)	37 (41.6)	22 (31.0)	2 (7.7)	<.001 (FE)
Moderate	109 (58.6)	44 (49.4)	48 (67.6)	17 (65.4)
Severe	16 (8.6)	8 (9.0)	1 (1.4)	7 (26.9)
Current % BSA, mean ± SD	27.3 ± 20.4	23.5 ± 19.2	28.3 ± 18.2	37.6 ± 26.0	.008 (AN)
Current EASI, mean ± SD	10.3 ± 8.1	8.3 ± 7.2	10.0 ± 5.8	18.1 ± 11.6	<.001 (AN)
Range 0–72
No. of body areas affected^b mean ± SD	2.7 ± 1.2	2.5 ± 1.2	2.7 ± 1.1	3.6 ± 0.6	<.001 (AN)
Current drug treatment
Topical/other^c, n (%)	127 (68.3)	63 (70.8)	51 (71.8)	13 (50.0)	.033 (FE)
Systemic, n (%)	24 (12.9)	8 (9.0)	7 (9.9)	9 (34.6)
Biologic, n (%)	35 (18.8)	18 (20.2)	13 (18.3)	4 (15.4)

All p values shown in Table 1 are across the three analysis groups. Pairwise comparisons between individual groups were also conducted. These are not shown in the table but included in the main text where appropriate.

^a Reduced base due to missing data (physician unaware).

^b Number of body areas affected out of a maximum of four.

^c Other included prescribed emollients, phototherapy, wet wraps, bleach baths, antibiotics for skin infection, antihistamines or pain medication.

Abbreviations. AN, ANOVA test; BSA, body surface area affected; CCI, Charlson Comorbidity Index; CH, Chi-squared test; EASI, Eczema Area and Severity Index; FE, Fisher’s Exact test; I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain; SD, standard deviation.

Patient demographics were similar across the three analysis groups, however differences in clinical and treatment characteristics were observed (Table 1). Compared with No I/SP, I + SP patients were considered to be currently more severe, had a higher mean BSA, EASI and a greater number of body parts affected (all p < .01). Compared with I-only, I + SP patients were more severe, had a higher EASI and greater number of body parts affected (all p < .01) but the difference in BSA was not significant (p = .057). No differences in comorbidities, as defined by CCI score, were observed between the groups.

Although the use of a biologic was similar across all the groups, we observed differences between the groups regarding the use of non-biologic systemic therapies (injected/oral corticosteroids and immunosuppressants). Direct pairwise comparison showed that I + SP patients were more likely to be receiving non-biologic systemic therapy than either No I/SP (p = .01) or I-only patients (p = .022). There was no difference observed between No I/SP and I-only patients (p = .945).

Symptom burden and concomitant conditions

Table 2 shows the associations of symptoms and concomitant conditions across the three analysis groups. While no differences in the incidence of skin infections were observed for baseline (day-to-day) symptoms, those with I + SP were more likely to suffer skin infection during an acute episode (p = .004). With respect to dry skin, those with I-only or I + SP were more likely to suffer dry skin than the reference group both day-to-day and during an acute episode (both p < .001), and those with I + SP were more likely to suffer infection during an acute episode than those with I-only (p < .001).

Table 2. Physician-reported symptoms and concomitant conditions.

	All patients	No I/SP patients	I-only patients	I + SP patients	p Value
Day to day symptom, n (%)	186	89	71	26
Moderate/severe skin infection	22 (11.8)	10 (11.2)	9 (12.7)	3 (11.5)	.951 (FE)
Moderate/severe dry skin (in non-lesional areas)	84 (45.2)	27 (30.3)	42 (59.2)	15 (57.7)	<.001 (CH)
Acute episode symptom, n (%)	186	89	71	26
Moderate/severe skin infection	16 (8.6)	6 (6.7)	3 (4.2)	7 (26.9)	.004 (FE)
Moderate/severe dry skin (in non-lesional areas)	45 (24.2)	21 (23.6)	10 (14.1)	14 (53.8)	<.001 (CH)
Concomitant condition, n (%)	165	77	64	24
Asthma	35 (21.2)	14 (18.2)	13 (20.3)	8 (33.3)	.277 (CH)
Allergic rhinitis	61 (37.0)	28 (36.4)	25 (39.1)	8 (33.3)	.874 (CH)
Urticaria	18 (10.9)	11 (14.3)	6 (9.4)	1 (4.2)	.380 (FE)
Angioedema	2 (1.2)	1 (1.3)	0 (0.0)	1 (4.2)	.271 (FE)
Allergic contact dermatitis	12 (7.3)	8 (10.4)	2 (3.1)	2 (8.3)	.227 (FE)
Atopic keratoconjunctivitis	16 (9.7)	9 (11.7)	2 (3.1)	5 (20.8)	.023 (FE)
Other allergic condition(s)	1 (0.6)	1 (1.3)	0 (0.0)	0 (0.0)	1.000 (FE)
Nasal polyps	4 (2.4)	1 (1.3)	2 (3.1)	1 (4.2)	.494 (FE)
Alopecia Areata	3 (1.8)	2 (2.6)	1 (1.6)	0 (0.0)	1.000 (FE)
None of these	66 (40.0)	30 (39.0)	28 (43.8)	8 (33.3)	.652 (CH)

All p values shown in Table 2 are across the three analysis groups.

With respect to concomitant conditions and comorbidities, no differences between the three analysis groups were seen for asthma, allergic rhinitis or allergic contact dermatitis, with the number of patients reported to have no comorbidities comparable across categories. Atopic keratoconjunctivitis appeared to be less likely in those with I-only (p = .023), however only 9.7% of all patients were reported to have this condition.

Impact of itch and skin pain on EASI and body areas affected

Figure 1 shows regression results regarding differences in severity for I-only and I + SP patients compared with the reference group with No I/SP (linear regression for EASI, negative binomial regression for body areas affected). I + SP patients had a mean EASI score of 15.43, on average 6.7 points higher than the reference group (p = .008).

Figure 1. I-only/I + SP patients have a higher EASI score and number of body areas affected versus reference group (No I/SP). Number of body areas affected out of a maximum of four. Abbreviations. EASI, Eczema Area and Severity Index; I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain.

In terms of number of body areas affected, I + SP patients had on average 3.5 out of 4 body areas affected, compared with 2.5 among the reference group (p = .001). While I-only patients also showed an increase in EASI and number of body areas affected versus the reference group, the differences were not statistically significant.

Impact of itch and skin pain on treatment received

Compared with the reference group, logistic regression analysis (Figure 2) showed the odds of I-only patients receiving potent topicals steroids were 2.55 times higher, and for I + SP patients 2.96 times higher (both p < .05). Similarly, the odds of I + SP patients receiving antihistamines were 3.03 times higher compared with the reference group (p = .036).

Figure 2. I-only/I + SP patients are more likely to be receiving a potent topical steroid or antihistamine versus reference group (No I/SP). Abbreviations. I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain.

Impact of itch and skin pain on POEM and WPAI

Linear regression results showed that I + SP patients had a mean POEM score of 13.42, on average 4.97 points worse than the reference group (p = .005) (Figure 3). For WPAI, I + SP patients had a mean overall work impairment of 43.1%, 14.5% higher than the reference group (p = .034). Differences between I-only and No I/SP were not significant.

Figure 3. I-only/I + SP have a higher POEM score and % work impairment versus reference group (No I/SP). Abbreviations. I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain; POEM, Patient-Oriented Eczema Measure – differences are numerical; WPAI, Work Productivity and Activity Impairment – differences show percentage.

Impact of itch and skin pain on sleep disruption

Figure 4 shows that for I-only patients the odds of them having sleep disturbance on a day-to-day basis were 8.92 times higher, and for I + SP patients 23.5 times higher, versus the reference group (both p < .001, logistic regression). The odds of I + SP patients being bothered by sleep disturbance were 4.6 times higher versus the reference group (p = .034, ordered logistic regression).

Figure 4. I-only/I + SP patients are more likely to have daily sleep disturbance and to be bothered by this versus reference group (No I/SP). Abbreviations. I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain.

Impact of itch and skin pain on patient dissatisfaction

Figure 5 shows that, compared with the reference group, the odds of I + SP patients expressing dissatisfaction in terms of lack of improvement in their condition were 7.26 times higher and the odds of them being dissatisfied with the convenience of treatment were 8 times higher (both p < .05, logistic regressions). No differences were observed between I-only and the reference group.

Figure 5. I-only/I + SP patients are more likely to be dissatisfied with improvement as a result of treatment and with inconvenience of treatment versus reference group (No I/SP). Abbreviations. I-only, itch only; I + SP, itch and skin pain; No I/SP, no itch or skin pain

Interaction between treatment and patient groups

Each of the preceding figures shows results taken from regression models without interaction effects. When we included an interaction term, we observed some significant interaction effects although these were typically associated with a low number of patients. For I + SP patients currently receiving biologic therapy (n = 4), the interaction term was associated with a decrease in EASI score of 11.74 (p = .009). For patients with I-only receiving systemic therapy other than a biologic (n = 7), the interaction term was associated with a decrease in overall WPAI of 20.90% (p = .006) and their odds of being bothered by sleep disturbance decreased by 83% (p = .029). With the introduction of interaction terms to the regression models, in some cases the p values associated with main effects became non-significant. This was the case with antihistamine treatment (p was 0.036, then p = .151), WPAI (p was 0.034, then p = .069), and being bothered by sleep disturbance (p was 0.034, then p = .308).

Discussion

Key findings of the study were that I + SP was associated with greater clinical severity in terms of number of body parts affected and increased body surface area affected, greater dissatisfaction with control and, from the patient perspective, greater sleep disruption, work impairment and poorer overall quality of life. While patients with itch only were slightly worse off than those without itch or skin pain, the differences were not generally statistically significant. As a result, we suggest that it is the skin pain component, when added to itch, that causes the greatest impact on the patient.

In order to reach this conclusion, it was important to demonstrate that the severity of patient-reported itch was similar between patients with I-only or I + SP. In this analysis, we used the frequency of patient-reported itch or pain to determine whether these symptoms were present or not. However, this did not take into consideration intensity, which may be a factor in the overall burden of disease. While we did not directly capture information relating to the intensity of the itch, we did investigate the degree to which patients were bothered by their itch, and found no statistically significant difference in the degree of bother when comparing the two groups.

The additive impact of skin pain to itch on burden of disease is consistent with findings from other studies, where there was a strong positive correlation between reported skin pain and both EASI[8] and POEM scores⁶. Maarouf et al. found that skin pain is typically reported only when signs of excoriation were present. The pain itself was described as resembling neuropathic pain^6,8, with the hands, perioral region, and toes being associated with the greatest pain burden⁸.

Our results showed that presence of itch but no skin pain was associated with patients being almost nine times more likely to experience sleep disturbance compared with those patients not experiencing itch or skin pain on a regular or continuous basis. In patients experiencing I + SP, this rose to them being more than 20 times more likely to have disturbed sleep and they were 4.6 times more likely to be bothered by this. Evidence from other studies has shown that sleep disturbance is increased among Japanese AD patients with estimates for difference in prevalence ranging from 1.38%¹³ to 8.36%¹⁵ compared with control patients without AD. However, these studies assessed itch without including the effect of concomitant skin pain. Our findings suggest that within the AD population the presence of both itch and skin pain are important determinants for sleep disturbance. Maarouf et al. also assessed the dual burden of itch and skin pain and reported sleep disturbance in 80% of their sample of AD patients. Skin pain was found to occur most frequently in the evening and overnight, helping to explain the high impact of skin pain on sleep in patients with AD⁸.

For those with I-only, while the patients report sleep disturbance, they are less likely to be bothered by this and it does not have a significant impact on work productivity or POEM. Evidence of the impact of sleep disturbance on work productivity has also been observed in a number of disease areas including nocturia²⁵, chronic obstructive pulmonary disease²⁶, overactive bladder²⁷, autoimmune blistering disease gastro-oesophageal reflux disease²⁸ and the consequences of sleep disturbance in terms of work productivity lost and associated costs has been measured in the general populations²⁹.

Finally, we found that patients with I + SP report experiencing significantly greater (14.5%) overall impairment at work. Regarding work productivity it is likely that factors other than the pain burden could be important in determining the extent of productivity loss. For example, the specific location of AD lesions, such as on the hands, may also play an important role in the degree of work impairment. The sample size in this research was not sufficiently large to investigate this so further research would be needed to explore this possibility. We did recognise the potential of an interaction effect between treatment received and the three analysis groups. However, although the effects of the interactions could be quite large, the number of patients for whom the interaction term was significant were very low.

Of the studies that have assessed patient-reported symptoms, many have focused on itch, but skin pain has not been widely assessed⁸. This is surprising given that, firstly, skin pain is a common symptom – Maarouf et al. reported that almost 4-times as many patients surveyed suffered concomitant skin pain and itch compared with itch alone – and secondly, skin pain severity is a good indicator to assess treatment success⁶. A French study indicated that more than half of patients with AD reported skin pain with an average pain intensity of almost 6/10, which was shown to impact QoL³⁰. A further study demonstrated that the presence of itch, pain, or other sensory symptoms were commonly reported by patients with classic skin diseases but also in those with acne, rosacea, or warts, with the association of itch and pain impacting psychological suffering³¹.

We note that patients with I + SP have worse PROs than those with I-only, thus it is clear from the data presented in this paper that skin pain is important. This is consistent with patho-physiology which highlights the potential independence of skin pain and itch associated with AD³². Their research suggests that although the nature of skin pain in AD has not been systematically studied and is therefore not well understood, patients reported soreness, discomfort, and tenderness that may reflect both peripheral and central pain sensitization.

As noted in the results, our sample included only 3 patients with skin pain independent of itch. This represented 1.6% of our sample in Japan, and is in line with previous research in other settings⁸. In the context of AD, while the percentage is small it still represents large numbers of patients in the total AD population.

The lack of SP-only patients in our patient sample in Japan can be considered a limitation of our study, so further research on Japanese patients would be needed to understand specific issues and challenges faced by SP-only patients and explore the impact on patients’ QoL and their ability to be confident in performing their usual activities. Additional research would also allow the possibility of assessing the implications of skin pain in relation to the incidence and implications of this in other countries. As we have done with I + SP in relation to I-only, estimating the incremental burden of I + SP versus SP-only is another area for potential future investigation.

As suggested by Maarouf et al., it remains clear that further studies are needed to elucidate mechanisms and transmission of skin pain in patients with AD. It is possible that the pain experienced in AD may be a result of factors other than itch/excoriation. There are conflicting reports on the relevance of nerve fibre density and their association with increased pain in AD, with future research required to understand any mechanisms involved⁶. A number of studies have also investigated the impact of allergies in patients with AD^33,34. While it is clear that sensitivity to allergens can trigger immune-mediated skin reactions and disease, we did not observe any differences in the presence of asthma, allergic rhinitis or allergic contact dermatitis between the I-only, I + SP or no I + SP patient groups in this study, and this is an avenue for further research.

Stander et al. also suggested that skin pain may not be adequately addressed by current therapies for AD and may be undertreated compared with other symptoms. Since skin pain can present as a neuropathic symptom independent from itch, it can pose a distinct and unique burden to patients. Recognition of the significant consequences of skin pain and discomfort should reinforce the need to assess and treat this symptom in patients with moderate-to-severe AD³². A position paper from the International Society of Atopic Dermatitis reviews the pathophysiology of itch and pain in AD, and presents expert recommendations for the management of both of these symptoms, dependent on their individual presence and persistence³⁵. The management of itch and pain is mostly based on topical therapy in patients with controlled disease, in particular given these are the main symptoms of patient complaint. A range of systemic therapeutic options are increasingly available for more severe disease presentation, and a patient-centred treatment model is recommended given the evolving treatment landscape.

We observed that patients with I + SP are more likely to be receiving potent topical steroids and/or antihistamine but they are not satisfied with either the level of improvement or the convenience of use associated with their medication. Although based on a relatively small sample, these findings are consistent with those of a two-year longitudinal study of Japanese patients published in 2019, in which Katoh et al. recorded no obvious difference in the drug treatments received by moderate versus severe AD patients at the start of the study, but reported minor differences in the use of oral immunosuppressants and oral corticosteroids over time. It was unclear whether this was ongoing treatment for chronic disease or whether these were treatments specifically for recent flares. This potential underlying treatment inertia may at least in part be driving the patient dissatisfaction we identified.

Furue et al. reported that the negative effects caused by long-term use of topical steroids has led to patients worldwide being afraid to use them – the concern is that this could lead to undertreatment in some cases³. Maarouf et al. also reported that application of steroids and other topical treatments may contribute to skin pain due to the presence of necessary excipients. In the future there is hope that treatment will be personalized depending on patient preferences, needs, clinical phenotypes, and comorbidities³⁶. The introduction of novel advanced treatments may contribute to meeting this need.

Katoh et al. found that disease severity reported by patients did not always map on to QoL – some patients with more severe disease reported experiencing a higher QoL than some with milder disease. This could indicate either that there is variation between patients in their perception and reporting of disease symptoms, or that patient-reported symptoms are independent of physician-assessed measures. However, it is also possible that individual aspects of a condition that feed into severity may have specific implications on QoL. We identify itch and skin pain as among the characteristics which may be a driver of reduced QoL and work productivity.

It has been recognized for over 20 years that patients with long term conditions modify their PRO scores over time as they adapt to their condition³⁷. This response shift, defined as a change in the meaning of self-evaluation, as a result of changes in internal standards, values, and/or conceptualization of the patient reported outcome³⁷ can be induced by health-changing events, such as falling ill or undergoing treatment³⁸. Response shifts are often a sign of adaptation and may distort the interpretation of changes in PRO scores over time. In the case of AD, it is possible that patients learn to accept certain signs and symptoms and so stop mentioning them to their doctor. This gives rise to the possibility that the overall burden reported by a patient may in fact underestimate the true burden of the condition, therefore potentially diminishing the impact of a symptom such as, in this case, skin pain. Physicians need to be aware of this potential effect and ask about it even if the patient does not report it, and tailor treatment of AD and other skin diseases according to the presence of symptomatic burdens such as itch and pain³⁵.

Similarly, some patients with mild and moderate EASI scores reported a relatively high skin pain burden, one possibility is that clinically “mild” looking lesions may still cause the patient significant skin pain and itch⁸. One potential explanation, which we were not able to assess in this study, would be that the skin pain is a relatively recent aspect of the patient’s condition; a symptom to which the patient may not have adapted. Skin pain may therefore be independent of an EASI score so would still need to be considered separately. Relying on general measures of severity such as EASI score may underrepresent the true impact of the condition on the patient. Finally, regarding the eight patients excluded due to missing data, their clinical and demographic characteristics were similar to the patients included in the analysis.

Limitations

Limitations with the Adelphi DSP include: (1) the selection of patients, whereby physicians are asked to choose a consecutive series of patients to avoid selection bias, but no formal source data verification procedures are in place; (2) diagnosis of the target patient group is based on judgement of the respondent physicians and not a formalized diagnostic checklist; (3) the point-in-time design cannot be used to demonstrate cause and effect, although associations between variables can be assessed. These limitations should be balanced with the methodological strengths, which include recruitment of a representative sample of patients across geographies, capturing patients from specialty physician practices, the ability to collect detailed information regarding patients’ demographics, clinical profile, health status and treatment/management history, ability to match patient-reported information to the physician-provided information, and standardized data collection tools across countries which will allow for a potential comparison/contrast of differences in customs and practices within and across countries.

Conclusions

We have demonstrated the extent of the additional burden associated with itch and skin pain compared with itch without skin pain (and absence of both itch and skin pain) in terms of patient QoL and work productivity, and dissatisfaction with disease control achieved. Therefore, skin pain needs to be addressed as an important and independent symptom in its own right either in combination with itch or on its own. We note that many patients with itch and skin pain are likely to be receiving potent topical steroids and/or antihistamine. Given a potential dissatisfaction with the level of improvement or convenience, combined with a need to consider the impact of potentially hidden symptoms such as skin pain, suggests that physicians could consider alternative and/or novel therapeutic approaches appropriate to the management of both itch and skin pain.

Transparency

Declaration of funding

This study was funded by Eli Lilly, Japan K.K.

Declaration of financial/other relationships

Atsushi Sakamoto and Hitoe Torisu-Itakura are employees of Eli Lilly Japan K.K., and hold stock of Eli Lilly and Company. Peter Anderson, James Piercy and James Pike are employees of Adelphi Real World, a company that received research funding from Eli Lilly, Japan K.K. Dr Kabashima has received grants, honoraria, or consulting fees from Japan Tobacco, LEO Pharma, Maruho, Mitsubishi Tanabe Pharma, Procter & Gamble, Ono Pharmaceutical, Taiho Pharmaceutical, and Torii Pharmaceutical. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

All authors were involved in (1) conception or design, or analysis and interpretation of data; (2) drafting and revising the article; (3) providing intellectual content of critical importance to the work described; and (4) final approval of the version to be published, and therefore meet the criteria for authorship in accordance with the International Committee of Medical Journal Editors (ICMJE) guidelines. In addition, all named authors take responsibility for the integrity of the work as a whole and have given their approval for this version to be published.

Acknowledgements

The authors would like to thank Victoria Davis and Gary Sidgwick, employees of Adelphi Real World, for medical writing and editorial support in preparation of this manuscript, in accordance with Good Publication Practice (GPP3) guidelines. They have no conflict of interest to report.

Data availability statement

All data that support the findings of this study are the intellectual property of Adelphi Real World. All requests for access should be addressed directly to Peter Anderson at [email protected]