Comparative Analysis of Hidradenitis Suppurativa Quality of Life (HISQOL) and Dermatology Life Quality Index (DLQI) Scores Using Data from Patients with Hidradenitis Suppurativa
Author(s)
Vadhariya A1, Coak E2, Wallinger H2, Truman I2, Teixiera B2, Garg A3, Kirby JS4
1Eli Lilly and Company, Indianapolis, IN, USA, 2Adelphi Real World, Maccelsfield, CHE, UK, 3Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, NY, USA, 4Penn State University, Hershey, PA, USA
OBJECTIVES:
The Hidradenitis Suppurativa Quality Of Life (HiSQOL) and Dermatology Life Quality Index (DLQI) are tools which can be used to assess patient quality of life (QoL) in those diagnosed with Hidradenitis Suppurativa (HS). This research aims to develop a comparative analysis of the HiSQOL to the DLQI, and vice versa, in HS patients.METHODS:
Data were drawn from the Adelphi HS Disease Specific ProgrammeTM, a survey of Dermatologists and their HS patients in the United States, France, Germany, Italy, Spain, and the UK. Dermatologists provided patient demographics, comorbidities, and clinical details. Patients reported quality-of-life measures using the validated HiSQOL and DLQI questionnaires. Model selection was assessed by comparing Coefficient of Determination (R2) from ten-fold cross-validation in a 70% randomly selected sample train set. Model performance was assessed in the remaining 30% test set. Question weights were evaluated using regression coefficients in HiSQOL to DLQI, and vice versa.RESULTS:
Data from 698 HS patients were analysed. Predicting HiSQOL using DLQI questions, the ideal model was linear regression, with an overall R2 of 0.73. DLQI questions with highest and most significant range of coefficients were related to pain, embarrassment, and sexual difficulties. Predicting DLQI using HiSQOL questions demonstrated an overall R2 of 0.68, with HiSQOL questions related to odour, exercise, sexual difficulties, and clothing choice delivering the highest and most significant range of coefficients.CONCLUSIONS:
Results show a high convergence between HiSQOL and DLQI tools highlighting the question areas most relevant to each other’s prediction, thus enabling comparisons to be made across studies. Further analysis of Psychometrics measures and interpretability might confirm results.Conference/Value in Health Info
Value in Health, Volume 25, Issue 6, S1 (June 2022)
Code
RWD32
Topic
Methodological & Statistical Research, Patient-Centered Research
Topic Subcategory
Artificial Intelligence, Machine Learning, Predictive Analytics, Patient-reported Outcomes & Quality of Life Outcomes
Disease
No Additional Disease & Conditions/Specialized Treatment Areas