Abstract
Introduction
Diphenylcyclopropenone (DPCP) is the medication of choice for the treatment of severe alopecia areata (AA) according to AA treatment guidelines. Precise initiation and application are important factors for successful treatment. However, it is difficult for patients who live far away to visit their doctor weekly.
Methods
We conducted a retrospective cohort study to assess the cost, effectiveness, and side effects of DPCP treatment between office-use DPCP (O-DPCP) and home-use DPCP (H-DPCP) in severe AA patients. A cost-effectiveness analysis was performed from the perspective of healthcare providers and patients using real-world data and the national cost statistics for hospital services comparing O-DPCP and H-DPCP in patients with severe AA at 24, 36, and 48 weeks.
Results
Two groups of 41 patients treated with O-DPCP and H-DPCP were enrolled. There was no significant difference in the proportion of patients who showed a favorable outcome (≥50% improvement) with minimal side effects between both groups at 24 (O-DPCP 43.9% vs. H-DPCP 26.8%, p = 0.11), 36 (O-DPCP 58.5% vs. H-DPCP 43.9%, p = 0.19), or 48 weeks (O-DPCP 63.4% vs. H-DPCP 56.1%, p = 0.49). The cost of H-DPCP was half of the cost of O-DPCP.
Discussion/Conclusion
H-DPCP is a cost-effective and time-efficient alternative treatment option for severe AA patients.
Keywords: Alopecia areata, Alopecia totalis, Alopecia universalis, Diphenylcyclopropenone, Home use, Cost-effectiveness
Introduction
Alopecia areata (AA) is an autoimmune disorder that causes hair loss from the body, mainly involving the scalp [1]. This condition occurs in males and females at any age with varying severity [1]. Several guidelines recommend contact immunotherapy, Janus kinase (JAK) inhibitors, potent topical corticosteroids, and short-contact anthralin. Diphenylcyclopropenone (DPCP) is commonly used for contact immunotherapy and has been effective in the treatment of severe AA based on standard AA treatment guidelines [2, 3, 4]. DPCP causes allergic contact dermatitis by modulating the CD4+/CD8+ T-lymphocyte ratio, resulting in the elimination of autoreactive lymphocytes and modification of cytokine production [5]. The average hair regrowth with DPCP treatment is 53% [6]. DPCP treatment requires weekly application and dose adjustments until the correct strength is achieved. Home-use DPCP (H-DPCP) is a treatment approach offered to patients unable to comply with weekly doctor visits due to time and financial constraints. In previous studies [7, 8], H-DPCP was shown to be an effective treatment option for patients without the time-consuming and costly limitations of the standard protocol. However, there is no previous study comparing the effectiveness and safety between H-DPCP and office-use DPCP (O-DPCP). This study compared the effectiveness and economic impact between O-DPCP and H-DPCP in extensive AA.
Materials and Methods
This study was comprised of 2 parts:
A retrospective cohort study to compare O-DPCP and H-DPCP in severe AA.
Cost-effectiveness analysis of H-DPCP using all data from the retrospective study.
Retrospective Cohort Study
The study was carried out from January 2011 to December 2016 at the Institute of Dermatology (IOD), Bangkok, Thailand, following its approval by the Institutional Review Board and the Department of Medical Services, Ministry of Public Health, Thailand. Subjects have given their written informed consent. In this study, O-DPCP patients received weekly DPCP treatments at the hospital as recommended in the standard guideline. In contrast, H-DPCP patients received weekly DPCP treatments at the hospital for 1–2 months to adjust to DPCP concentrations and were given training on self-applying DPCP at home. Then, H-DPCP patients were required to have monthly doctor visits for DPCP concentration evaluation and prescription refills. The retrospective cohort study design was required to determine the efficacy of DPCP for the 24-week treatment period. It was the patient's choice to select either O-DPCP or H-DPCP based on their time and financial constraints.
Patients
All eligible subjects were severe AA patients (>50% scalp hair loss) who received DPCP treatments at the study site for a minimum of 24 weeks. We excluded patients with unstable underlying conditions affecting hair loss. We obtained baseline characteristics and clinical data by performing a chart review. Furthermore, costs from the patient perspective were obtained from phone/in-person interviews. All patients were independently examined by 2 dermatologists for the type of AA and area involvement, with routine 3-month scalp pictures.
Effectiveness and Safety Assessments
The area of improvement measured using Severity of Alopecia Tool (SALT) scores was routinely recorded in all AA patient charts to determine the effectiveness of DPCP treatments. The SALT scores ranged from 0 to 100, with higher scores indicating more significant hair loss areas [9]. The delta SALT score (SALTbaseline − SALT24/36/48 weeks) represents the area of improvement. If the improvement area was >50% compared with the baseline, it was defined as a favorable outcome. Safety issues regarding the adverse side effects of DPCP were compared between the 2 groups. The effectiveness and safety of O-DPCP and H-DPCP treatments were evaluated at 24, 36, and 48 weeks.
Statistical Analysis
Categorical variables were reported as numbers and percentages. Continuous variables were reported as the mean ± SD, and the disease duration was reported as the median and range. Statistical comparisons of the mean width between 2 treatment groups were performed with a paired t test, and the Mann-Whitney test was applied to analyze the median width data. Pearson's χ2 test or Fisher's exact test was used to compare the categorical data as appropriate. All calculations were performed using Stata version 14.2.0 (Stata Corp., College Station, TX, USA). A 2-tailed p value of <0.05 was considered statistically significant.
Cost-Effectiveness Analysis
Model Structure
A simple decision tree was built by including 2 options for treatment selection: O-DPCP and H-DPCP (Fig. 1).
Fig. 1.
Simple decision tree model.
Model Parameters
All necessary data were obtained, including the effectiveness of H-DPCP and O-DPCP from the retrospective cohort study and direct and indirect medical costs from phone/in-person interviews. Costs were calculated on the basis of healthcare provider perspectives. Cost discounting was considered in the model.
Cost Analysis
Treatment costs between H-DPCP and O-DPCP were compared using both healthcare provider and patient perspectives. Inclusive healthcare provider perspectives were direct medical costs, such as labor, material, and capital costs. Generally, while the doctor visit expenses for H-DPCP patients were equivalent to those in the O-DPCP group, they saved money due to fewer visits. For the evaluation of patient perspectives, both direct nonmedical and indirect costs were considered. Direct nonmedical costs refer to travel, food, and ancillary employer expenses on the hospital visitation day. Indirect costs refer to the loss of daily income. Following the assessment of their treatment effectiveness, patients subsequently underwent phone/in-person interviews, regarding the direct nonmedical and indirect costs. Reported costs of all items occurring in different years were inflated using the Consumer Price Index. Standard medical visit fees were obtained from the lists of Health Intervention and Technology Assessment Program (HITAP) and used in the cost comparisons between O-DPCP and H-DPCP treatments in addition to the patient's actual out-of-pocket expenses. All costs were calculated in the US dollar ($).
Analysis
The cost per success/failure of O-DPCP and H-DPCP was calculated in the decision tree model. The incremental cost-effectiveness ratio (ICER) was formulated as the ratio of the cost differences to the differences in the effectiveness between O-DPCP and H-DPCP to determine the cost required to achieve additional success (ICER = cost of O-DPCP − cost of H-DPCP/effectiveness of O-DPCP − effectiveness of H-DPCP). A sensitivity analysis was performed to determine the degree of efficacy of H-DPCP, which affects the outcome.
Results
Retrospective Cohort
Over a 6-year study period, the study included 82 patients (41 patients in the O-DPCP group and 41 patients in the H-DPCP group). The baseline characteristics of all patients are shown in Table 1. Approximately, two-thirds of patients in both groups were female, with a mean (±SD) age of 33.5 ± 13.6 years in the O-DPCP group and 26.4 ± 12.0 years in the H-DPCP group (p = 0.01). The primary types and percentages of alopecia disorders in patients undergoing O-DPCP and H-DPCP were as follows: alopecia totalis/alopecia universalis with 23 (56.1%) and 17 (41.5%) and multifocal patchy AA with 14 (34.2%) and 20 (48.8%), respectively. There were no significant differences between the 2 groups in terms of median duration, medical complications, occupations, living environments, transportation, and direct nonmedical costs.
Table 1.
Baseline characteristics
| Variables | Office group (N = 41) | Home group (N = 41) | p value |
|---|---|---|---|
| Mean age ± SD, years | 33.54±13.62 | 26.41±12.02 | 0.014 |
| Gender, n (%) | |||
| Male | 11 (26.83) | 16 (39.02) | 0.240 |
| Female | 30 (73.17) | 25 (60.98) | |
| Duration of disease, median (range), day | 365 (30–6,209) | 365 (30–6,940) | 0.630 |
| Duration of disease, median (range), month | 11.99 (0.986–204.13) | 11.99 (0.986–228.16) | 0.629 |
| Subtypes, n (%) | |||
| Multiple patchy | 14 (34.15) | 20 (48.78) | |
| Ophiasis | 1 (2.44) | 1 (2.44) | |
| Acute diffuse total | 3 (7.32) | 3 (7.32) | 0.680 |
| Alopecia totalis | 10 (24.39) | 9 (21.95) | |
| Alopecia universalis | 13 (31.71) | 8 (19.51) | |
| Area involved, % | |||
| Mean±SD | 79.00±21.87 | 80.61±20.09 | 0.729 |
| Median (range) | 85 (50–100) | 87 (30–100) | 0.989 |
| Underlying diseases, n (%) | |||
| None | 36 (87.80) | 34 (82.93) | 0.532 |
| Thyroid | 1 (2.44) | 3 (7.32) | 0.616 |
| Vitiligo | 0 (0) | 1 (2.44) | 1.000 |
| Atopy | 4 (9.76) | 4 (9.76) | 1.000 |
| Connective tissue diseases | 0 (0) | 1 (2.44) | 1.000 |
| Others | 0 (0) | 1 (2.44) | 1.000 |
| Occupation | |||
| Officer − employee | 11 (26.83) | 5 (12.20) | |
| Government employee | 2 (4.88) | 2 (4.88) | |
| Business − self-employed | 8 (19.51) | 1 (2.44) | |
| Student | 3 (7.32) | 13 (31.71) | |
| Unemployed | 6 (14.63) | 3 (7.32) | 0.007 |
| Contractor | 4 (9.76) | 3 (7.32) | |
| State enterprise employee | 0 (0) | 2 (4.88) | |
| Gardener | 0 (0) | 2 (4.88) | |
| Unknown | 7 (17.07) | 10 (24.39) | |
| Domicile, n (%) | |||
| Bangkok | 11 (26.83) | 4 (9.76) | |
| Suburban | 2 (4.88) | 8 (19.51) | |
| Central | 5 (12.20) | 4 (9.76) | |
| Northeast | 6 (16.63) | 3 (7.32) | |
| North | 3 (7.32) | 5 (12.20) | 0.116 |
| South | 4 (9.76) | 1 (2.44) | |
| East | 2 (4.88) | 3 (7.32) | |
| West | 0 (0) | 2 (4.88) | |
| Foreign country | 1 (2.44) | 1 (2.44) | |
| Unknown | 7 (17.07) | 10 (24.39) | |
| Current living location, n (%) | |||
| Bangkok | 20 (48.78) | 8 (19.51) | |
| Suburban | 4 (9.76) | 6 (14.63) | |
| Central | 7 (17.07) | 6 (14.63) | |
| Northeast | 0 (0) | 0 (0) | |
| North | 1 (2.44) | 3 (7.32) | 0.131 |
| South | 0 (0) | 1 (2.44) | |
| East | 2 (4.88) | 5 (12.20) | |
| West | 0 (0) | 1 (2.44) | |
| Foreign country | 0 (0) | 1 (2.44) | |
| Unknown | 7 (17.07) | 10 (24.39) | |
| Education, n (%) | |||
| Primary school | 5 (12.20) | 3 (7.32) | |
| Secondary school | 14 (34.15) | 12 (29.27) | |
| Associate degree | 4 (9.76) | 3 (7.32) | 0.895 |
| Bachelor's degree | 10 (24.39) | 11 (26.83) | |
| Master's degree | 1 (2.44) | 2 (4.88) | |
| Unknown | 7 (17.07) | 10 (24.39) | |
| Mode of transportation, n (%) | |||
| Car | 8 (19.51) | 8 (19.51) | 0.858 |
| Bus | 19 (46.34) | 10 (24.39) | 0.110 |
| Taxi | 1 (2.44) | 3 (7.32) | 0.475 |
| Motorcycle | 3 (7.32) | 1 (2.44) | 0.543 |
| Sky train | 7 (17.07) | 4 (9.76) | 0.589 |
| Shuttle | 5 (12.20) | 8 (19.51) | 0.495 |
| Plane | 1 (2.44) | 2 (4.88) | 0.696 |
| Patients' view travel cost/visit, mean±SD, USD | 7.65±15.14 | 20.32±51.55 | 0.135 |
| Patients' view food cost/visit, mean±SD, USD | 3.11±3.14 | 3.11±3.03 | 1.000 |
| Patients' view miscellaneous cost*/visit, mean±SD, USD | 1.8±9.87 | 0.76±2.50 | 0.517 |
| Patients' view income loss/visit, mean±SD, USD | 8.79±17.04 | 5.28±9.27 | 0.251 |
| Patients' view opportunity loss/visit, mean±SD, USD | 0.93±2.98 | 0.76±2.86 | 0.798 |
| Patients' view total cost/visit, mean±SD, USD | 22.28±33.34 | 30.24±53.54 | 0.419 |
| Patients' view total cost (adjusted)/visit, USD | |||
| Mean ± SD | 22.67±34.22 | 30.50±53.54 | 0.432 |
| Median (range) | 15.48 (0–165.63) | 13.95 (0–324.38) | 0.699 |
Miscellaneous cost refers to extra expense due to patient disappearance from the ordinary day such as baby sit and parents care.
Comparisons of the favorable clinical outcome (FCO) and side effects between O-DPCP and H-DPCP treatments at 24, 36, and 48 weeks are shown in Table 2. At 24 weeks, O-DPCP patients showed a significant increase in improvement areas compared with H-DPCP patients (34% vs. 13%, p = 0.037). At 36 and 48 weeks, areas of improvement were not significantly different between the 2 groups. Interestingly, the FCO at 24, 36, and 48 weeks was not significantly different between the 2 groups. Improvements in O-DPCP treatment responses at 24 weeks increased from 44% to 63% at 48 weeks, whereas H-DPCP treatment outcomes over the same period increased from 27% to 56%. Additionally, an increased incidence of side effects was found in the O-DPCP group at 24, 36, and 48 weeks (p < 0.05).
Table 2.
Effectiveness and safety of O-DPCP and H-DPCP at weeks 24,36, and 48
| Variables | Week 24 office group (N = 41) | Week 24 home group (N = 41) | p value | Week 36 office group (N = 35) | Week 36 home group (N = 36) | p value | Week 48 office group (N = 32) | Week 48 home group (N = 33) | p value |
|---|---|---|---|---|---|---|---|---|---|
| Area improvement* | |||||||||
| Mean±SD,% | 34.07±44.69 | 12.96±45.44 | 0.037 | 36.92±50.32 | 22.20±48.33 | 0.213 | 38.62±55.60 | 21.72±52.55 | 0.212 |
| Median (range) | 28.57 (−98-100) | 0 (−127.97-100) | 0.027 | 50 (−90-100) | 15.66 (−120.93-100) | 0.182 | 46.43 (−90-100) | 10.00 (−120.93-100) | 0.133 |
| Good outcome**, n (%) | 18(43.90†) | 11 (26.83†) | 0.106 | 24 (58.54†) | 18(43.90†) | 0.185 | 26 (63.41†) | 23(56.10†) | 0.499 |
| Side effects, n (%) | |||||||||
| None | 18(43.90) | 31 (75.61) | 0.003 | 10(47.62) | 24 (63.16) | 0.011 | 6(46.15) | 21 (61.76) | 0.001 |
| Vesicles/bullaes | 15(36.59) | 7(17.07) | 0.080 | 9(25.71) | 6(16.67) | 0.396 | 5(15.63) | 6(18.18) | 1.000 |
| Generalized eczema | 1 (2.44) | 0(0) | 1.000 | 1 (2.86) | 0(0) | 0.493 | 1 (3.13) | 1 (3.03) | 1.000 |
| Urticaria | 0(0) | 0(0) | 0(0) | 0(0) | 0(0) | 0(0) | |||
| Fever | 0(0) | 0(0) | 0(0) | 0(0) | 0(0) | 0(0) | |||
| Hyperpigmentation | 9(21.95) | 3 (7.32) | 0.116 | 9(25.71) | 5(13.89) | 0.245 | 11 (34.38) | 4(12.12) | 0.042 |
| Hypopigmentation | 0(0) | 0(0) | 0(0) | 1 (2.78) | 1.000 | 0(0) | 0(0) | ||
| Lymphadenopathy | 1 (2.44) | 1 (2.44) | 1.000 | 0(0) | 0(0) | 0(0) | 0(0) | ||
| Folliculitis | 1 (2.44) | 0(0) | 1.000 | 1 (2.78) | 0(0) | 1.000 | 1 (3.13) | 3 (3.03) | 1.000 |
O-DPCP, office-use diphenylcyclopropenone; H-DPCP, home-use diphenylcyclopropenone.
Area improvement when comparing to baseline.
A good outcome means the improvement was > 50% compared to the baseline.
These parameter values were used in the decision tree analysis.
The total cost in the O-DPCP group at 24, 36, and 48 weeks was significantly higher than that in the H-DPCP group based on healthcare provider and patient perspectives in all categories except direct and indirect nonmedical costs (Table 3). Total direct costs calculated from standard average spending by Thai patients (listed by HITAP) were similar to our actual costs presented in Table 4. At 24 weeks, the cost of O-DPCP was USD 368.56/person compared with USD 175.05/person for H-DPCP. At 48 weeks, the cost of O-DPCP was USD 683.52/person compared with USD 303.67/person for H-DPCP.
Table 3.
The economic outcome of O-DPCP and H-DPCP at weeks 24, 36, and 48
| Variables | Week 24 office group (N = 41) | Week 24 home group (N = 41) | p value | Week 36 office group (N = 35) | Week 36 home group (N = 36) | p value | Week 48 office group (N = 32) | Week 48 home group (N = 33) | p value |
|---|---|---|---|---|---|---|---|---|---|
| Healthcare providers' view, USD | |||||||||
| Material cost, mean±SD | 74.32±42.22 | 45.59±20.47 | <0.001 | 126.43±52.70 | 68.70±37.49 | <0.001 | 189.83±82.08 | 105.05±53.93 | <0.001 |
| Office material cost, mean ± SD | 9.91±1.05 | 4.36±1.50 | <0.001 | 13.53±1.94 | 5.51±1.84 | <0.001 | 16.62±3.64 | 6.67±1.93 | <0.001 |
| Take-home material cost, mean ± SD | 0.30±1.59 | 5.24±1.44 | <0.001 | 0.84±1.32 | 7.84±1.78 | <0.001 | 1.93±2.77 | 10.90±2.02 | <0.001 |
| Office medicine cost, mean ± SD | 62.28±41.50 | 21.35±15.09 | <0.001 | 100.60±54.92 | 28.22±19.71 | <0.001 | 145.26±90.80 | 39.02±25.61 | <0.001 |
| Take-home medicine cost, mean ± SD | 0.51±2.30 | 14.29±7.16 | <0.001 | 4.53±8.32 | 25.61±17.62 | <0.001 | 12.04±20.52 | 42.18±26.36 | <0.001 |
| Side-effect medicine cost, mean ± SD | 0.34±0.77 | 0.13±0.49 | 0.160 | 1.05±3.49 | 0.18±0.60 | 0.145 | 1.55±6.66 | 0.51±1.11 | 0.382 |
| Labor cost, mean ± SD | 172.39±22.91 | 74.55±4.05 | <0.001 | 235.27±42.77 | 93.41±30.36 | <0.001 | 288.79±72.74 | 114.25±32.16 | <0.001 |
| Capital cost, mean ± SD | 12.01±1.27 | 5.28±1.82 | <0.001 | 16.40±2.36 | 6.67±2.23 | <0.001 | 20.15±4.42 | 8.09±2.34 | <0.001 |
| Total direct cost*, mean ± SD | 258.73±52.84 | 125.42±42.46 | <0.001 | 378.86±81.88 | 169.11±59.54 | <0.001 | 498.77±136.89 | 227.39±72.10 | <0.001 |
| Total indirect cost**, mean ± SD | 51.75±10.57 | 25.08±8.49 | <0.001 | 75.77±16.38 | 33.82±11.91 | <0.001 | 99.75±27.38 | 45.48±14.42 | <0.001 |
| Total cost***, mean ± SD | 310.47±63.41 | 150.51±50.95 | <0.001 | 454.64±98.26 | 202.93±71.44 | <0.001 | 598.52±164.27 | 272.87±86.53 | <0.001 |
| Patients' view, USD | |||||||||
| Travel cost, mean ± SD | 176.76±362.12 | 224.45±696.49 | 0.698 | 222.66±452.38 | 257.08±701.89 | 0.793 | 254.73±501.51 | 289.67±713.66 | 0.798 |
| Median (range) | 90 (0–2250) | 67.50 (0–4,484.38) | 0.716 | 116.25 (0–2812.50) | 112.50 (0–4,484.38) | 0.689 | 140.63 (0–3,093.75) | 112.50 (0–4,484.38) | 0.678 |
| Food cost, mean ± SD | 70.76±71.46 | 31.40±29.09 | 0.002 | 88.48±83.01 | 39.53±37.37 | 0.001 | 102.29±93.53 | 46.36±45.12 | 0.001 |
| Median (range) | 75 (0–375) | 28.13 (0–100) | 0.002 | 78.75 (0–421.88) | 34.38 (0–125.00) | 0.002 | 90.00 (0–468.75) | 34.38 (0–162.50) | 0.001 |
| Miscellaneous cost, mean ± SD | 43.17±236.99 | 8.99±32.02 | 0.363 | 51.04±269.49 | 11.05±39.12 | 0.349 | 58.90±302.70 | 12.88±45.79 | 0.339 |
| Median (range) | 0(0–1,500) | 0(0–140.63) | 0.447 | 0(0–1,687.50) | 0(0–168.75) | 0.447 | 0(0–1,875.00) | 0(0–196.88) | 0.447 |
| Income loss cost, mean ± SD | 208.25±407.98 | 56.06±117.39 | 0.024 | 253.03±481.92 | 69.28±137.70 | 0.021 | 286.97±533.85 | 82.00±160.85 | 0.021 |
| Median (range) | 0(0–1,500) | 0 (0–500) | 0.183 | 0(0–1,875.00) | 0 (0–562.50) | 0.181 | 0(0–2125) | 0 (0–656.25) | 0.176 |
| Opportunity loss cost, mean ± SD | 22.28±71.60 | 6.71 ±30.38 | 0.197 | 24.62±80.14 | 9.30±36.04 | 0.268 | 25.01 ±80.78 | 11.43±44.24 | 0.348 |
| Median (range) | 0 (0–300) | 0(0–150) | 0.626 | 0 (0–375.00) | 0(0–187.50) | 0.626 | 0 (0–375) | 0 (0–225.00) | 0.632 |
| Total cost, mean ± SD | 530.68±822.71 | 329.87±716.75 | 0.242 | 650.72±976.75 | 388.98±730.98 | 0.173 | 740.13±1,078.91 | 445.76±755.71 | 0.156 |
| Median (range) | 371.62 (0–3,975) | 146.63 (0–4,541.33) | 0.085 | 371.62 (0–4,968.75) | 173.42 (0–4,541,033) | 0.082 | 418.07 (0–5,465.63) | 189.69 (0–4,541.33) | 0.101 |
| Healthcare providers' view (HITAP), USD | |||||||||
| Total cost, mean ± SD | 368.56±56.49 | 175.05±58.10 | <0.001 | 528.97±92.86 | 230.96±79.46 | 0.0000 | 683.52±166.80 | 303.67±87.56 | <0.001 |
HITAP, Health Intervention and Technology Assessment Program; O-DPCP, office-use diphenylcyclopropenone; H-DPCP, home-use diphenylcyclopropenone.
Total direct cost = material cost + labor cost + capital cost.
Total indirect cost = 20% total direct cost.
Total cost = total direct cost + total indirect cost.
Table 4.
Cost-effectiveness results
| Treatment choice | Week 24 |
Week 36 |
Week 48 |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| cost (USD) | effectiveness (chance) | total cost of the choice (USD) | IC (USD) | IE | ICER (USD) | cost (USD) | effectiveness (chance) | total cost of the choice (USD) | IC (USD) | IE | ICER (USD) | cost(USD) | effectiveness (chance) | total cost of the choice (USD) | IC(USD) | IE | ICER (USD) | |
| Healthcare provider's view | ||||||||||||||||||
| O-DPCP | 311.81 | 160.75 | 0.17 | 945.59 | 452.56 | 248.44 | 0.15 | 1,656.25 | 597.78 | 329.38 | 0.07 | 4,705.34 | ||||||
| Success | 291.31† | 0.4390† | − | − | − | − | 440.47 | 0.5854 | − | − | − | − | 594.94† | 0.6341† | − | − | − | − |
| Failure | 325.47† | 0.5610† | − | − | − | − | 469.63 | 0.4146 | − | − | − | − | 602.59† | 0.3659† | − | − | − | − |
| H-DPCP | 151.06 | − | − | − | − | − | 204.132 | − | − | − | − | − | 268.41 | − | − | − | ||
| Success | 141.81† | 0.2683† | − | − | − | − | 212.94 | 0.4390 | − | − | − | − | 249.97† | 0.5610† | − | − | − | − |
| Failure | 153.69† | 0.7317† | − | − | − | − | 197.25 | 0.5610 | − | − | − | − | 291.94† | 0.4390† | − | − | − | − |
| Healthcare provider's view (HITAP) | ||||||||||||||||||
| O-DPCP | 368.56 | 193.53 | 0.17 | 1,138.41 | 528.47 | 294.84 | 0.15 | 1,965.63 | 685.16 | 384.34 | 0.07 | 5,490.63 | ||||||
| Success | 352.16† | 0.4390† | − | − | − | − | 523.53 | 0.5854 | − | − | − | − | 691.03† | 0.6341† | − | − | − | − |
| Failure | 381.41† | 0.5610† | − | − | − | − | 532.63 | 0.4146 | − | − | − | − | 675.00† | 0.3659† | − | − | − | − |
| H-DPCP | 175.03 | 233.63 | 300.81 | |||||||||||||||
| Success | 169.44† | 0.2683† | − | − | − | − | 252.69 | 0.4390 | − | − | − | − | 289.13† | 0.5610† | − | − | − | − |
| Failure | 177.09† | 0.7317† | − | − | − | − | 218.69 | 0.5610 | − | − | − | − | 315.78† | 0.4390† | − | − | − | − |
O-DPCP, office-use diphenylcyclopropenone; H-DPCP, home-use diphenylcyclopropenone; HITAP, Health Intervention and Technology Assessment Program; IC, incremental cost; IE, incremental effectiveness; ICER, incremental cost-effectiveness ratio.
These parameter values were used in the decision tree analysis.
Cost-Effectiveness Analysis
Figure 1 shows the decision tree with 2 potential options for DPCP treatments, including O-DPCP and H-DPCP. At 24 weeks, the probability of the FCO in O-DPCP and H-DPCP groups was 0.4390 and 0.2683, respectively. At 36 weeks, the FCO probability was 0.5854 in the O-DPCP group and 0.4390 for the H-DPCP option. At 48 weeks, the FCO probability for the O-DPCP option was 0.6341 and that for the H-DPCP option was 0.5610. Table 4 shows the cost of O-DPCP and H-DPCP categorized into success and failure groups and the differences in the ICER between O-DPCP and H-DPCP at 24, 36, and 48 weeks.
At 24 weeks, the simple decision tree showed O-DPCP costs at $311.81/person and H-DPCP costs at USD 151.06/person. Incorporating the HITAP data, these numbers correspond to USD 368.56 versus 175.03 baht/person, respectively.
The ICER (O-DPCP vs. H-DPCP) was USD 945.59 per 1 FCO (or USD 1,138.41 per 1 FCO if the HITAP cost data were used), as shown in Table 4. At 36 weeks, the differences in the ICER between O-DPCP and H-DPCP groups revealed that $1,656.25 was saved by each H-DPCP patient (USD 1,965.63 calculated from the HITAP data). At 48 weeks, the difference in the ICER between both choices showed a savings of USD 4,705.34 for each patient in the H-DPCP group compared with the O-DPCP group (USD 5,490.63 using the HITAP data).
The 1-way sensitivity analysis presenting the difference in patient costs at 48 weeks is shown in Figure 2. Because the cost-effectiveness threshold (willingness to pay) used to measure the underlying demand and valuation for nonmarketed social goods, such as healthcare, was less than the value of the ICER (USD 8,857.60 savings per each patient's similar treatment outcome), O-DPCP was not as cost-effective compared with H-DPCP. Under these circumstances, the H-DPCP protocol provided a better cost-effective treatment choice when the ICER was lower than the USD 8,857.60 treatment cost associated with patient office visits, which was <USD 294.38 over 48 weeks.
Fig. 2.
One-way sensitivity analysis. ICER, incremental cost-effectiveness ratio.
Discussion
AA is an autoimmune disease that can be difficult to treat in patients who suffer for a long term from this condition. DPCP is an excellent choice for most patients because of its efficacy and economic value, and it can be used as a monotherapy or in combination with anthralin [7, 10, 11]. O-DPCP is the standard protocol, while H-DPCP is sometimes used as an alternative option for patients who have difficulty visiting hospitals every week to improve patient compliance [12]. Systematic reviews showed that the average response rate of O-DPCP for alopecia totalis and alopecia universalis was 47%–55% [6, 13]. Although some studies have demonstrated the efficacy and safety of H-DPCP [7, 8], further investigation is required to support the use of H-DPCP worldwide [14, 15]. Furthermore, the efficacy and safety of O-DPCP compared with H-DPCP remain a debate [14].
In our study, O-DPCP showed better results than H-DPCP after 24 weeks of treatment, with a similar treatment response at weeks 36 and 48. However, there was no significant difference in the number of good responders (≥50% improvement) between O-DPCP and H-DPCP at 24, 36, or 48 weeks. Although a previous study showed an overall therapeutic response of 45% for both O-DPCP and H-DPCP [7, 8], our study showed that the effectiveness of O-DPCP and H-DPCP was lower at 24 weeks (44% vs. 27%) but higher at 48 weeks (63% vs. 56%). One explanation is a slow response to DPCP, which has been previously reported [7]. We observed a similar result of hair regrowth in patients receiving both O-DPCP and H-DPCP at 48 weeks, which may indicate that the treatment cutoff time should be extended from the commonly used 24-week treatment protocol to 48 weeks for both O-DPCP and H-DPCP. To our knowledge, there are no publications regarding H-DPCP and its cost-effectiveness. Therefore, the data presented in this study provide practical information for policymakers to consider when making their informed treatment decisions.
This study was performed to determine the effectiveness and cost-effectiveness of O-DPCP and H-DPCP in extensive AA patients from the perspective of both healthcare providers and patients. The SALT scores were used to establish standard effectiveness outcomes, with cost assessments for the medication (direct medical cost) and financial burdens related to both direct nonmedical and indirect costs of DPCP treatment obtained by interviewing patients. In our study, both O-DPCP and H-DPCP showed similar effectiveness, but the expenses associated with H-DPCP were significantly lower than those of O-DPCP. Side effects of DPCP occurred less often in the H-DPCP group than in the O-DPCP group, possibly due to less frequent DPCP concentration adjustments or fewer DPCP applications in the H-DPCP group. Additionally, the actual cost assessments from the view of healthcare providers that we collected were consistent with the costs calculated on the basis of national cost statistics from HITAP.
The accepted values for the cost-effectiveness threshold in Thailand were 300,000 baht (USD 9,375) or <USD 50,000 per quality-adjusted life-year gained in the USA. In our study, ICER analysis showed that during the first 24 weeks, H-DPCP patients needed to pay an additional USD 945.59 to earn 1 additional treatment success event. However, for both 36- and 48-week intervals, H-DPCP patients needed to pay an additional USD 1,656.25 and USD 4,705.34, respectively, to earn the same success rate as O-DPCP patients. There are other typical treatments for extensive AA by the standard guideline, including intralesional corticosteroid and JAK inhibitors [2, 3, 4]. In Thailand, the cost of intralesional corticosteroids in conjunction with topical corticosteroids is USD 212.5 for 6 months, which is more expensive than H-DPCP (USD 150.5) and gives the poorer result (25–50%) [16]. JAK inhibitors show a better outcome (58–62%) [17, 18, 19] but have much higher costs of USD 5,625 for 6 months. According to the general comparative effectiveness of both DPCP delivery methods and respective cost savings, H-DPCP has advantages over O-DPCP due to the convenience of home treatments. Following adequate training given to the H-DPCP group, patient application of DPCP was an acceptable method with a low rate of side effects. One additional benefit of H-DPCP is that the number of DPCP applications can be increased to more than once a week, which may have better efficacy. However, further studies are needed to confirm this prediction.
This study documents the relative effectiveness of 48-week DPCP treatment protocols. We focused on the perspectives of the healthcare providers and patients. Furthermore, we used extensive records of all patient visits from the IOD hair clinic over 6 years. SALT scores and pictures were routinely recorded in every visit. Additionally, the expenses related to side effects were determined in each group besides treatment costs. We avoided the overestimation of patient costs by performing a sensitivity analysis to evaluate the association between the ICER and the cost of patient visits.
The limitations of this study were its retrospective single-center design and small patient numbers. It was necessary to use retrospective chart reviews for DPCP efficacy evaluations in this study because DPCP is a time-consuming treatment that may not suit policymakers desiring prospective data. A multicenter study offers a future design option to evaluate larger patient groups.
Conclusion
This study not only provides essential cost-effective information for health policy decisions regarding H-DPCP therapy but also offers the potential for effective clinical treatment of extensive AA in expanded provincial regions. H-DPCP is a cost-effective and time-efficient alternative treatment option for severe AA patients. We suggest that the use of H-DPCP can have a substantial impact on the treatment of extensive AA patients.
Statement of Ethics
This study was reviewed and approved by Institute of Dermatology IRB; approval # 012/2559. Subjects have given their written informed consent.
Conflict of Interest Statememt
The authors have no conflicts of interest to declare.
Funding Sources
The authors have received no external funding.
Author Contributions
Chinmanat Lekhavat contributed to the concept and design, acquisition of data, analysis, interpretation of data, drafting of the manuscript, critical revision of the paper for important intellectual content, provision of study materials and patients, administrative technical or logistical support, and supervision. Pinyo Rattanaumpawan contributed to statistical analysis and interpretation of data, critical revision of the paper for important intellectual content. Isree Juengsamranphong contributed to provision of study materials and patients and administrative technical or logistical support.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available (due to state restrictions such as privacy or ethical restrictions).
Acknowledgments
We thank Melissa Crawford, PhD, from Edanz Group (https://en-author-services.edanz.com/ac) for editing a draft of this manuscript and Kanonrat Boochangkool, MD, and Penny Pimonrat for assistance with data collection.
References
- 1.Alkhalifah A, Alsantali A, Wang E, McElwee KJ, Shapiro J. Alopecia areata update: part I. Clinical picture, histopathology, and pathogenesis. J Am Acad Dermatol. 2010;62((2)):177–90. doi: 10.1016/j.jaad.2009.10.032. [DOI] [PubMed] [Google Scholar]
- 2.Gupta AK, Carviel JL, Foley KA, Shear NH, Piraccini BM, Piguet V, et al. Monotherapy for alopecia areata: a systematic review and network meta-analysis. Skin Appendage Disord. 2019;5((6)):331–7. doi: 10.1159/000501940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Strazzulla LC, Wang EHC, Avila L, Lo Sicco K, Brinster N, Christiano AM, et al. Alopecia areata: an appraisal of new treatment approaches and overview of current therapies. J Am Acad Dermatol. 2018;78((1)):15–24. doi: 10.1016/j.jaad.2017.04.1142. [DOI] [PubMed] [Google Scholar]
- 4.Ohlmeier MC, Traupe H, Luger TA, Böhm M. Topical immunotherapy with diphenylcyclopropenone of patients with alopecia areata: a large retrospective study on 142 patients with a self-controlled design. J Eur Acad Dermatol Venereol. 2012;26((4)):503–7. doi: 10.1111/j.1468-3083.2011.04114.x. [DOI] [PubMed] [Google Scholar]
- 5.Herbst V, Zöller M, Kissling S, Wenzel E, Stutz N, Freyschmidt-Paul P. Diphenylcyclopropenone treatment of alopecia areata induces apoptosis of perifollicular lymphocytes. Eur J Dermatol. 2006;16((5)):537–42. [PubMed] [Google Scholar]
- 6.Jang YH, Jung HJ, Moon SY, Lee WJ, Lee SJ, Lee WK, et al. Systematic review and quality analysis of studies on the efficacy of topical diphenylcyclopropenone treatment for alopecia areata. J Am Acad Dermatol. 2017;77((1)):170–2 e1. doi: 10.1016/j.jaad.2017.03.015. [DOI] [PubMed] [Google Scholar]
- 7.Chiang KS, Mesinkovska NA, Piliang MP, Bergfeld WF. Clinical efficacy of diphenylcyclopropenone in alopecia areata: retrospective data analysis of 50 patients. J Investig Dermatol Symp Proc. 2015;17((2)):50–5. doi: 10.1038/jidsymp.2015.28. [DOI] [PubMed] [Google Scholar]
- 8.Lee S, Lee WS. Home-based contact immunotherapy with diphenylcyclopropenone for alopecia areata is as effective and safe as clinic-based treatment in patients with stable disease: a retrospective study of 40 patients. J Am Acad Dermatol. 2018;78((3)):599–e1. doi: 10.1016/j.jaad.2017.09.037. [DOI] [PubMed] [Google Scholar]
- 9.Olsen EA, Hordinsky MK, Price VH, Roberts JL, Shapiro J, Canfield D, et al. Alopecia areata investigational assessment guidelines part II. National alopecia areata foundation. J Am Acad Dermatol. 2004;51((3)):440–7. doi: 10.1016/j.jaad.2003.09.032. [DOI] [PubMed] [Google Scholar]
- 10.Durdu M, Özcan D, Baba M, Seçkin D. Efficacy and safety of diphenylcyclopropenone alone or in combination with anthralin in the treatment of chronic extensive alopecia areata: a retrospective case series. J Am Acad Dermatol. 2015;72((4)):640–50. doi: 10.1016/j.jaad.2015.01.008. [DOI] [PubMed] [Google Scholar]
- 11.Murad A, Bergfeld W. Treatment for facial alopecia areata: a systematic review with evidence-based analysis. J Am Acad Dermatol. 2018;78((3)):601–5. doi: 10.1016/j.jaad.2017.09.054. [DOI] [PubMed] [Google Scholar]
- 12.Kim BJ, Lee S, Lee CH, Lee WS. Home-based contact immunotherapy with diphenylcyclopropenone improves compliance with the recommended follow-up for patients with alopecia areata: A retrospective cohort study. J Am Acad Dermatol. 2020;82((5)):1223–5. doi: 10.1016/j.jaad.2019.10.043. [DOI] [PubMed] [Google Scholar]
- 13.Lee S, Kim BJ, Lee YB, Lee WS. Hair regrowth outcomes of contact immunotherapy for patients with alopecia areata: a systematic review and meta-analysis. JAMA Dermatol. 2018;154((10)):1145–51. doi: 10.1001/jamadermatol.2018.2312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Karanovic S, Harries M, Kaur MR. Diphenylcyclopropenone for alopecia areata: a U.K. survey. Br J Dermatol. 2018;179((2)):514–5. doi: 10.1111/bjd.16489. [DOI] [PubMed] [Google Scholar]
- 15.Meah N, Wall D, York K, Bhoyrul B, Bokhari L, Sigall DA, et al. The alopecia areata consensus of experts (ACE) study: results of an international expert opinion on treatments for alopecia areata. J Am Acad Dermatol. 2020;83((1)):123–30. doi: 10.1016/j.jaad.2020.03.004. [DOI] [PubMed] [Google Scholar]
- 16.Tan E, Tay YK, Goh CL, Chin Giam Y. The pattern and profile of alopecia areata in Singapore: a study of 219 Asians. Int J Dermatol. 2002;41((11)):748–53. doi: 10.1046/j.1365-4362.2002.01357.x. [DOI] [PubMed] [Google Scholar]
- 17.Yu DA, Kim YE, Kwon O, Park H. Treatment outcome of oral tofacitinib and ruxolitinib in patients with alopecia areata: a systematic review and meta-analysis. Indian J Dermatol Venereol Leprol. 2021 Sep;87((5)):621–7. doi: 10.25259/IJDVL_975_19. [DOI] [PubMed] [Google Scholar]
- 18.Liu LY, Craiglow BG, Dai F, King BA. Tofacitinib for the treatment of severe alopecia areata and variants: a study of 90 patients. J Am Acad Dermatol. 2017;76((1)):22–8. doi: 10.1016/j.jaad.2016.09.007. [DOI] [PubMed] [Google Scholar]
- 19.Zheng C, Tosti A. Alopecia areata: new treatment options including janus kinase inhibitors. Dermatol Clin. 2021;39((3)):407–15. doi: 10.1016/j.det.2021.03.005. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available (due to state restrictions such as privacy or ethical restrictions).