Hidradenitis Suppurativa International Online Community: Patient Characteristics and a Novel Model of Treatment Effectiveness

Abstract

Treatment for hidradenitis suppurativa is diverse, yet frequently unsatisfactory. The aims of this study were to create a reproducible artificial intelligence-based patient-reported outcome platform for evaluation of the clinical characteristics and comorbidities of patients with hidradenitis suppurativa, and to use this to grade treatment effectiveness. A retrospective patient- reported outcome study was conducted, based on online questionnaires completed by English-speaking patients registered to the hidradenitis suppurativa StuffThatWorks® online community. Data collected included patient characteristics, comorbidities and treatment satisfaction. These were recoded into scalable labels using a combination of machine learning algorithm, manual coding and validation. A model of treatment effectiveness was generated. The cohort included 1,050 patients of mean ± standard deviation age 34.3 ± 10.3 years. Greater severity of hidradenitis suppurativa was associated with younger age at onset (p < 0.001) and male sex (p < 0.001). The most frequent comorbidities were depression (30%), anxiety (26.4%), and polycystic ovary syndrome (16.6%). Hurley stage I patients rated topical agents, dietary changes, turmeric, and pain relief measures more effective than tetracyclines. For Hurley stage II, adalimumab was rated most effective. For Hurley stage III, adalimumab, other biologic agents, systemic steroids, and surgical treatment were rated more effective than tetracyclines. Patients with hidradenitis suppurativa often have comorbid psychiatric and endocrine diseases. This model of treatment effectiveness provides a direct comparison of standard and complementary options.

Keywords

• hidradenitis suppurativa
• patient-reported outcome
• comorbidities
• treatment effectiveness
• machine learning
• artificial intelligence