View Record

TitleFolliculitis keloidallis nuchae severity score: development and reliability assessment
AuthorNyika, Dennias Toderai
TypeMaster Thesis
AbstractBackground: Folliculitis keloidalis nuchae (FKN) is a chronic inflammatory condition that targets the hair follicle, leading to keloidal scarring and alopecia. The absence of a severity scoring tool for FKN limits objective assessment of disease progression and response to treatment. Objectives: To develop and test the reliability of a severity scoring tool for FKN. Methods: The tool was developed based on lesion type, number, size and distribution on the scalp. An initial pilot period with 2 assessors was followed by the main study that used 78 anonymised and standardised clinical photographs of the back of the scalp. The participants were selected from an ongoing case control study of FKN. The assessors could allocate disease severity in one of 14 categories (with/without inflammation). However, inflammation (especially erythema) can be missed in photographs of pigmented skin. Thus, two groups of analysis were conducted first with all 14 and again with 8 categories (i.e. excluding inflammation). Assessors were 4 dermatology consultants and 7 registrars, who all independently scored the same anonymised and standardised photographs on two separate occasions, 2 weeks apart. Results: Inter-observer standard errors were higher with the 14-category compared to the 8- category analysis for both consultants and registrars. The intraclass correlation coefficient for registrars improved from poor [0.46 (0.36 -0.56)] to good [0.74 (0.68- 0.80)] with 14 compared to 8-categories, but stayed the same for consultants [0.82 (0.76 – 0.88) versus 0.81 (0.75 – 0.87)]. Limitations of the study were the use of clinical photographs instead of live participants and the problem that the signs of inflammation may be particularly difficult to judge in pigmented skin. Conclusion: We developed a severity scoring tool with poor to good reliability which also highlighted the difficulty of perceiving inflammation from clinical photographs. This improved with the seniority of the observer. The 8-category analysis has good reliability for clinical photographs for both junior and senior staff. For live patient care and clinical trials the 14-category version is likely to be more useful, but requires validation.
PublisherFaculty of Health Sciences
PublisherDepartment of Medicine