Our primary aim was to study current AA treatment practices across age groups and stages of progressive disease severity. Our results suggest that children receive treatment less frequently than adults. The most commonly used AA treatments overall are corticosteroids and minoxidil. The frequency of AA treatment decreases with progressive disease severity, and AT appears to be the "critical point" at which there is a drop in rates of certain treatments.
Our survey respondents were largely general dermatologists in suburban, solo/group private practice. The gender distribution of respondents was 60% male/40% female. Most respondents saw five or fewer new AA patients per month. Interestingly, the majority of our respondents had been in practice for over 10 years, and a relatively high number (40%) had practiced for 21 years or more. The American Academy of Dermatology (AAD) reported the gender distribution of its members in 2007 to be 62% male/38% female, and the 2007 AAD Practice Profile Survey (sponsored by the AAD) reported the practice settings of a random sample of AAD members practicing dermatology in the U.S. to be 54.8% suburban and 44.2% solo practice/33.1% dermatology group practice. Thus, the demographics of our study's respondents appear to be generally consistent with the overall demographics of U.S. dermatologists.
We observed that respondents treated children less frequently than adults, particularly in cases of patch hair loss. Dermatologists may be less willing to treat children due to the pain or risk associated with treatment . They may also seek to avert the patient's attention from hair loss or to avoid social disruption , or to prevent a "roller coaster of improvements and setbacks" in the absence of proven treatments . As emphasized by a recent review, the primary limitation in evidence-based treatment of children with AA is the lack of randomized controlled trials of AA therapies . AA is a disease with strong potential for psychological trauma, particularly in children [10, 11], and it is important to optimize treatment in this age group. Thus, some recommend that pediatric patients receive counseling or join a support group from the onset of the disease . Our survey did not assess how often physicians recommend support groups or counseling to their patients with AA; such an assessment would be worthwhile in the future.
Our observation that the frequency of AA treatment decreases with progressive disease severity may similarly reflect physician reluctance to expose patients to the risks, costs, and side effects of treatment in the absence of proven options. It is interesting to note that while the number of respondents prescribing no treatment increases with disease severity, so does the number of respondents prescribing systemic drugs with more side effects such as systemic corticosteroids and methotrexate. This suggests that two different approaches to treatment exist among dermatologists - some see a greater need to treat with more widespread disease, while others see less value in treatment.
Topical and intralesional corticosteroids and minoxidil were the most commonly used treatments cited by our respondents; however, while these drugs have frequently been reported to be useful and successful in management of AA, they are of questionable long-term and overall benefit. A recent systematic review found no randomized, controlled trials of intralesional corticosteroids, systemic corticosteroids, or minoxidil that demonstrated clinically significant hair regrowth . The meta-analysis identified only one trial that showed a statistically significant difference between hair regrowth with betamethasone valerate foam and betamethasone dipropionate lotion; the study did not include a vehicle-control arm . As the authors noted, corticosteroids and minoxidil appear to be in wide use due to their relative safety, but their use remains unsupported by rigorous evidence. We noted that these drugs are used quite frequently even to treat more severe forms of AA (AT/AU) per our survey results, though many perceive them to be ineffective (Table 6) - again, we posit that their relative safety and the physician's desire to offer treatment leads to their use despite lack of evidence as to their efficacy.
We noted that use of wigs and eyebrow tattoos for scalp and eyebrow hair loss is limited. As it is very important to AA patients to optimize their cosmetic appearance, more widespread use of wigs and tattoos may be warranted.
The most commonly cited barriers to the use of various treatments were risk of side effects; pain associated with treatment; excessive time commitment, compliance or cost; and lack of experience with the treatment. Interestingly, lack of evidence and FDA approval were not commonly cited as barriers to the use of treatment. Our findings suggest that many physicians desire to treat despite the limited amount of available evidence, but that many treatments are simply unsatisfactory or unfamiliar.
Data from the NAAR showed that, as suggested in our survey findings, a relatively high percentage of AA patients never receive treatment even in the more advanced stages of AA. NAAR data did not reveal any consistent pattern of drug use. Even the most commonly employed treatments (topical, intralesional, and systemic corticosteroids, topical immunotherapy, and minoxidil) show no consistent pattern of increasing or decreasing use with progressive stages of AA.
Limitations of our survey study include our response rate, which was 14.6% of our target population. It is not known whether the respondents to our survey are a representative sample of the target population. Thus, substantial "survey bias" is possible, and it is reasonable to believe that some degree of "survey bias" is likely to be present. We used a paper questionnaire rather than a web-based survey format because dermatologists' e-mail addresses for an entire region could not be obtained; our choice of format may have skewed the demographics of our responders. We simply defined children as "preadolescents" in our questionnaire, leaving determination of an age cutoff to the respondent; some respondents may have defined the "preadolescent" age range differently than others. Our survey instrument divided AA into four discrete stages (first episode of patch loss vs. multiple episodes of patch loss vs. AT vs. AU) in order to simplify and categorize physicians' treatment preferences, but within each disease stage lies a spectrum of disease presentations with varied prognoses that may warrant different treatment approaches; these nuances may have been lost in our study. Our survey did not ask physicians about their diagnostic and treatment approaches for related medical conditions such as thyroid disease.
The NAAR portion of our study was limited by our use of the "mentions per patient" ratio; because we could not access individual patient treatment histories, we could not calculate percentages of patients who had used treatments from a certain drug category (as we could not account for cases in which a patient had received two or more drugs from the same category). However, we were able to determine the percentages of patients who had received no treatment over the course of their disease. The experiences of the patients in the NAAR database may not be representative of the experiences of all AA patients.
Our survey relied on physician recall of practice habits; medical records were not examined for verification of physician practices. Nonetheless, our physician-reported findings showed good congruence with patient-reported treatment data from the NAAR.