The disease burden and frustration of patients with AA is high, mostly caused by limited treatment options and lack of reimbursement, limiting access to approved drugs such as Janus kinase inhibitors.

Through the mixed-methods design, we were able to combine patient experiences and quantitative data reflecting the reality of healthcare in Germany.

Online or telephone interviews were conducted between June 2022 and April 2023. We used a semistructured, pilot-tested interview guideline (online supplemental files S2 and S3) based on the research questions and the literature. If necessary, the guideline was adapted throughout the study to allow for integrating new aspects found in previous interviews and in the claims data analysis. The interviews were conducted by two female researcher associates (health scientists, M.Sc; BH, TMJ) experienced in conducting qualitative interviews through their university education and previous qualitative studies. We took a neopositivist philosophical position with a realist approach, meaning that we sought to approach objective reality.16 Data collection and analysis were not theory-based.

Patients were recruited conveniently via a dermatological outpatient clinic, patient organisations and social media. Dermatologists were recruited from a nationwide network and the dermatological societies, representing >95% of professionally working German dermatologists. All participants gave written informed consent. Participants had the right to withdraw their consent at any time until pseudonymised data were anonymised. Only members of the study team had access to interview data and information on the participants. For patients, it was stated explicitly that non-participation had no influence on their treatment. The interviewers had no personal, professional or therapeutic relation with the interviewees, except for one dermatologist working at the same institute. Interviews introduced themselves, their background and the aim of the study prior to the interview. Recruitment stopped as soon as further interviews added no additional content, meaning that thematic saturation was reached. For this, we documented which categories were newly developed per interview. We defined thematic saturation to be reached once no new categories emerged on the two highest levels of the category system.

All interviews were audiorecorded and transcribed verbatim, in the process of which the interviews were pseudonymised. Data were analysed with MAXQDA V.24 (VERBI GmbH, Berlin, Berlin, Germany) using qualitative content analysis according to Kuckartz17: two researchers (BH, TMJ) formulated research questions and derived main categories from these. Afterwards, both researchers familiarised themselves with the interview transcripts and case summaries and categorised the first transcript jointly. Based on this, they revised the main categories. One researcher categorised three subsequent transcripts, and the categorisation was checked by the other. In the further categorisation, additional (levels of) subcategories were developed, if necessary. The category system was frequently discussed by the study team (BH, CB, TMJ).

The analysis was based on an anonymised 40% sample (n=2 513 860; 58.0% women, average age 55.1 years in 2020) of all persons (aged at least 18 years) who were insured for at least 1 day with the DAK-Gesundheit (DAK-G) between 01 January 2016 and 31 December 2020. Based on the guidelines and recommendations for ensuring good secondary data (good practice statements (GPS)) and good epidemiological practice (GEP), these data from the SHI system do not require an ethical vote or the consent of the insured persons.18 19 This is because these data sets are based on existing data originally collected for other purposes. In addition, the data are anonymised, which prevents the identification of individuals and thus minimises the associated risks. The use of the project’s database is well protected by legal and institutional regulations, such as the GPS and the GEP. These regulations ensure ethical and responsible use of the data.

Prevalent insured persons were considered to have AA (ICD-10 GM L63) if they had at least one confirmed outpatient or inpatient main or secondary diagnosis. Incident cases had to have a 3-year diagnosis-free period in 2020.9

The utilisation of outpatient care for prevalent AA was determined based on all physician contacts.20 AA-related drug prescriptions (coded by the Anatomical Therapeutic Chemical definition system) and defined daily doses (DDD) or treatments (ambulatory doctors fee schedule (German: einheitlicher Bewertungsmaßstab, EBM)) were only recorded if an AA (online supplemental file S4, table S1) diagnosis was made in the same quarter.

In the sensitivity test, systemic antibiotics and their proportion of prescriptions in the total population were analysed to determine the population effect. The treatment history of AA patients was followed over four quarters using drug prescriptions. The proportion of outpatient psychotherapy (at least one prescription; online supplemental file S4, table S2) was compared between prevalent AA patients and other skin diseases (ICD-10 ‘L’ diagnosis excluding L63). Hospitalisation rates for full and partial inpatient stays with a principal diagnosis of AA were also recorded.

The administrative estimates of period prevalence and incidence (annual) were expressed as percentages with 95% CIs and were standardised according to the German federal statistics institute DESTATIS as of 31 December for the observation year. Descriptive statistics and multivariate analysis methods were used according to the data level of variables. For comparative analyses, the data were adjusted for age, sex and federal state using 1:3 nearest neighbour propensity score matching (PSM). Logistic regression was used to compute propensity scores. Relative risks (RR) and 95% CI were calculated to assess differences between the comparison cohorts. The analyses were performed with SAS V.9.4 German (SAS Institute, Cary, North Carolina, USA).

Both types of data were collected and analysed in parallel to enable integrated analysis: The study team met on a monthly basis to discuss interim results of all parts of the study. Hence, patterns found in one study part were presented and possibilities discussed to investigate from these originating hypotheses in the other study part. As an example, we detected a high use of antibiotics in the claims data analysis. Accordingly, we asked dermatologists in interviews if they prescribe antibiotics for patients with AA and, if so, in which cases, and we probed this topic in interviews with patients.

The data were synthesised into an overall conclusion and from this, recommendations for action were derived.

Based on the results of all study parts, an expert meeting was held with two dermatologists, two patient representatives and four scientists (researchers responsible for analyses of qualitative and claims data). Within the meeting, recommendations for action to reduce inequalities in the care of patients with AA were developed.

In the expert panel, members of a patient organisation participated to discuss results and develop recommendations. At the end of the project, all participants received a lay summary of the results.

In the qualitative interviews (table 1), 20 patients and 14 dermatologists participated (duration 20–85 min).

For each patient and dermatologist, distinct category systems were established through qualitative content analysis. The category system was considered final when the last two participant interviews and the three last dermatologist interviews had revealed no new categories on the two highest levels of the category system. Main categories in patient data were ‘healthcare’ and ‘life with AA’. Main categories derived from dermatologist interview were: ‘healthcare stations’, ‘treatment’ and ‘reasons for visits at dermatologist’. For the full category system, see online supplemental file S5, tables S3 and S4.

According to claims data, the standardised prevalence of AA in 2020 was 0.23% (n=4692; mean age 55.8 years (SD 19.0, median 57); 76.2% women) and incidence was 0.09% (n=1565; 2019), corresponding to about 150 000 prevalent and 53 000 incident persons across Germany. The prevalence and incidence increased with age.

Following the claims data, 57.5% of the persons with AA received at least one drug prescription in 2020. Topical and systemic non-biological drugs were prescribed to 36.8% and 35.1%, respectively; among the topical drugs, 95.3% were TGCs (mostly class III and IV) and 11.1% received SGCs (table 2). Less than 0.02% (n=9) used JAKi inhibitors, namely tofacitinib and baricitinib (each 0.06%), ruxolitinib (0.04%) and upadacitinib (0.02%). Ultraviolet light (UV) therapy received 0.53% of the insured persons with AA in 2020 (0.78% in 2016), with selective phototherapy being the most common form of treatment. With 32.0% of insurees with AA receiving psychotherapeutic treatment, they were 19% (RR 1.19, CI 1.13 to 1.25) more likely to receive psychotherapy than persons with another skin disease (according to PSM).

The total prescription volume decreased continuously. The number of DDD dispensed fell from around 288 398 in 2016 to 2 33 264 in 2020. Only systemic biologics (ie, dupilumab) showed a slight increase from 1206 DDDs in 2018 to 2580 DDDs in 2020 (figure 1).

41.1% of the AA-related drugs were prescribed by general practitioners (GPs), 32.8% by dermatologists and 4.2% by internists. GPs prescribed antibiotics (65.6%), TGCs class I (63.4%), systemic antihistamines (53.4%), azathioprine (66.7%) and SGCs (50.4%) more often than dermatologists (figure 2). Dermatologists prescribed the majority of biologics (ie, dupilumab; 75.0%). A total of 27.3% of insured persons with AA received a systemic antibiotic, compared with 21.7% of the general population (all insured persons without AA).

In the qualitative data, patients reported the use of different therapeutic options including local, systemic and further therapies or products. Some patients decided against therapy (table 3, Q1.1), mostly due to experienced lack of improvement. Some reported the therapy was too time-consuming or was not likely to be successful (Q1.2). Female patients also reported ending treatment due to pregnancy.

Patients who received topical therapies reported different therapeutic success: in some patients, hair started growing again, while others did not see improvement.

Many reported wearing headgear (Q1.3), having permanent make-up or using specific shampoos. Furthermore, patients reported lifestyle changes to improve hair growth.

For systemic therapies, patients reported the use of immunosuppressants (Q1.4), oral or intralesional systemic corticosteroids and (in few patients) JAKis and biologics through participation in a clinical study or as an off-label therapy. The patient with JAKi reported improvement of symptoms.

Most patients reported having tested various therapies and products, including prescribed therapies and over-the-counter products or home remedies. They also mentioned critical aspects regarding therapy: most patients were frustrated because of few therapeutic options. When receiving treatment, many felt a lack of shared decision-making (Q1.5). Many reported scepticism towards medication and side effects. This scepticism includes a general scepticism not only against conventional medicine but also towards JAKi. Some patients criticised the exclusively symptomatic treatment and that no long-term improvement was achieved. Some wanted their practitioner to provide more information about the disease and therapy options (Q1.6). The most discussed aspects regarding therapy options were costs and external conditions. Patients reported major differences in cost reimbursement between different SHIs and criticised that their insurance did not pay for the necessary equipment (eg, wigs).

The dermatologists reported that they usually start treatment with local therapies, mostly TGCs. Most dermatologists also mentioned therapies with diphenylcyclopropenone or platelet-rich plasma as treatment options; however, only a few reported using these in practice. Also, both therapies could only be offered as privately paid treatments. Other local therapies were TCI and UV therapy.

Dermatologists also reported offering new therapies to patients, namely JAKi or immunomodulating therapies as off-label therapies. Study inclusion was reported as another possibility to offer persons with AA access to modern treatments (Q1.7). Regarding systemic therapies, SGCs were described as a short-term, acute treatment. Dermatologists described JAKi as a very effective treatment (Q1.8), but as only a theoretical therapeutic option in Germany. As a treatment to ‘simply stop this acute phase of the disease’ (dermatologist, male, 60–69 years), immunosuppressants such as cyclosporine and methotrexate, as well as minoxidil, were reported. Finally, dermatologists described that patients were often content with using wigs and other supportive devices (Q1.9), and some patients did not require or want any additional treatment.

Dermatologists also reported issues regarding the therapy, notably the overall limited therapy options, especially for those working in office-based practice. The second major aspect was the lack of reimbursement by the health insurances (Q1.10). Further issues were recurrence after treatment end (Q1.11) and non-respondence to therapies.

Almost all dermatologists reported spontaneous remission in many patients with mild AA. They encouraged these patients to wait and document the symptoms (Q1.12).

An important aspect of AA treatment was patient involvement, which could be achieved through patient education (Q1.13) or by considering patient preferences. Dermatologists reported overall high levels of adherence (Q1.14).

Patients reported their symptoms describing different severity of hair loss ranging from a few bald spots to total hair loss; most had severe AA, including some with a long disease history. Few patients had experienced a spontaneous remission. Some reported co-morbidities, including allergies, bronchial asthma and other autoimmune diseases or a family history of AA.

Few patients reported no obvious trigger; most patients assumed their AA to be triggered by different factors such as stress, for example through work (table 4, Q2.1), private life or relationships. Few patients had experienced their first symptoms after an infection. Several female patients reported changes in their symptoms (both increased and reduced) in relation to physical or hormonal changes such as pregnancy, miscarriage or breast feeding. Some patients experienced a seasonal influence on symptoms (Q2.2). Of these, all but one reported hair growth in spring/summer and hair loss in autumn/winter.

Information and exchange with other patients were very important for the participants, making use of social media, patient support groups or other ways of connecting with other patients. Some reported searching for new research results online. Few patients said they did not want or need exchange with other patients.

Patients have found different ways of handling their disease. Some patients reported having accepted the disease (Q2.3, Q2.4). Others talked about the mental burden, ranging from “of course, mentally you're not doing so great” (female patient) to severe depression and suicidal thoughts. The visibility of the disease was a severe burden for many patients. They described resignation/frustration, feeling uncomfortable and desperate or repressing thoughts about their AA.

Some patients described wanting to understand the causes of their disease and how to influence its progression (Q2.5); some hoped for a spontaneous remission.

In the qualitative interviews, for the initial examination of the AA, generally including blood tests, some patients visited a dermatologist while others first saw a GP. After this, the GP was mostly no longer a relevant provider for AA treatment. Visits at the dermatologist mostly happened as needed (table 4, Q3.1), while some patients saw their dermatologist regularly for the monitoring of the disease/therapy progression. Some patients visited specific dermatological consultations for hair loss. Pharmacies were only used by patients to pick up prescribed therapies (Q3.2). Most patients had made use of alternative options, including visits to homoeopaths or acupuncture therapy. Most of these patients reported no improvement in their symptoms, and some mentioned the high costs of alternative providers; however, some appreciated that, “you are perceived quite differently. Of course, they have a completely different time register for you” (female patient). Some patients attained psychological care, mostly looking for help in handling their disease (Q3.3) but also hoping to reduce stress and improve the symptoms.

Most patients criticised a lack of empathy or felt like they were not taken seriously by the physician (Q3.4). However, some reported being taken seriously and appreciating the dermatologists’ empathy and honesty (Q3.5). Some patients experienced a lack of competencies or expert knowledge (Q3.6) and therefore saw multiple different dermatologists. Patients mostly reported long waiting times until their appointment and difficulties finding good physicians.

Dermatologists reported that patients usually first visit GPs or dermatologists. GPs would usually do routine blood tests to exclude other diagnoses and then refer the patient to dermatologists. Patients with a long or complicated disease history are often referred to university clinics or specialised outpatient clinics (Q3.7). Inpatient clinic visits were reported to play a minor role. Furthermore, dermatologists reported that some patients used over-the-counter products from the pharmacy, for example, shampoos against hair loss. Few mentioned suggesting the option of psychological support to patients.

Alternative medical options were mentioned by some dermatologists but reported that patients with AA generally make less use of these compared with other patient groups. Dermatologists also reported that information and exchange with other patients was important for many patients.

In 2019, 53.3% of the 1565 patients with AA initially received prescriptions for AA. In the first quarter, 70.5% were treated by dermatologists and 31.5% by GPs, with 47.7% receiving prescriptions, falling to 39.6% in the third quarter. In the fourth quarter, the proportion of patients treated by GPs rose to 53.5%, while the proportion treated by dermatologists fell to 27.0% (figure 3).

The proportion of persons with AA having received systemic non-biologic drugs increased by age (table 5). The user rate of topical treatment was significantly higher in women (26.2%; p<0.001) and those from urban areas (26.5%; p =0.002). Dermatologists prescribed topical and biological drugs more frequently (63.7%; 75.0%; each p <0.001), whereas GPs prescribed systemic non-biologics, especially GCs more frequently (60.3%; 48.1%; p <0.001).

Not applicable.

This study involves human participants and the study has been approved by the Local Ethics Committee at the University Medical Center Hamburg-Eppendorf (LPEK-0427). Participants gave informed consent to participate in the study before taking part.