Penny Watson1, Alan Brennan2, Helen Birch2, Hong Fang2, Michelle Petri2. 1. School of Health and Related Research, University of Sheffield, Sheffield, UK, Global Health Outcomes, GlaxoSmithKline, Uxbridge, UK and Department of Medicine-Rheumatology, Johns Hopkins University School of Medicine, Baltimore, MD, USA. p.r.watson@sheffield.ac.uk. 2. School of Health and Related Research, University of Sheffield, Sheffield, UK, Global Health Outcomes, GlaxoSmithKline, Uxbridge, UK and Department of Medicine-Rheumatology, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
Abstract
OBJECTIVE: The aim of this study was to develop an SLE disease model that simulates long-term outcomes of SLE to estimate the long-term effectiveness and cost-effectiveness of SLE treatments. METHODS: Longitudinal data from 1354 patients from the Hopkins Lupus Cohort were included in the analysis. Statistical models were created to estimate disease activity [Safety of Estrogen in Lupus Erythematosus National Assessment (SELENA) SLEDAI scale] and prednisone dose over time using linear regression. Survival models for organ damage and mortality were created. The models were combined in a predictive simulation of SLE organ damage and mortality. Predictions were assessed against the Hopkins Lupus Cohort data. RESULTS: The analyses found that change in the annual average SLEDAI score was associated with the previous annual average SLEDAI score, renal involvement, age, male gender, African American ethnicity, anaemia, haematological involvement, increased DNA binding and low complement. The annual average prednisone dose increased for every unit increase in annual average SLEDAI. Organ damage and mortality modelling demonstrated that adjusted mean SLEDAI and binary SLEDAI organ involvement indicators predicted mortality, cardiovascular, renal, neuropsychiatric, pulmonary, gastrointestinal, ocular and skin damage. The cumulative average prednisone dose was associated with risk of cardiovascular, ocular, musculoskeletal, neuropsychiatric and gastrointestinal damage, gonadal failure and diabetes mellitus. The simulation reproduced mean SLEDAI and organ damage scores from the Hopkins Lupus Cohort. CONCLUSION: Longitudinal modelling of an SLE cohort confirmed relationships between risk factors and long-term outcomes in SLE. The models serve to estimate the probability of SLE outcomes over time and can be used to estimate the effectiveness and cost-effectiveness of new treatments.
OBJECTIVE: The aim of this study was to develop an SLE disease model that simulates long-term outcomes of SLE to estimate the long-term effectiveness and cost-effectiveness of SLE treatments. METHODS: Longitudinal data from 1354 patients from the Hopkins Lupus Cohort were included in the analysis. Statistical models were created to estimate disease activity [Safety of Estrogen in Lupus Erythematosus National Assessment (SELENA) SLEDAI scale] and prednisone dose over time using linear regression. Survival models for organ damage and mortality were created. The models were combined in a predictive simulation of SLE organ damage and mortality. Predictions were assessed against the Hopkins Lupus Cohort data. RESULTS: The analyses found that change in the annual average SLEDAI score was associated with the previous annual average SLEDAI score, renal involvement, age, male gender, African American ethnicity, anaemia, haematological involvement, increased DNA binding and low complement. The annual average prednisone dose increased for every unit increase in annual average SLEDAI. Organ damage and mortality modelling demonstrated that adjusted mean SLEDAI and binary SLEDAI organ involvement indicators predicted mortality, cardiovascular, renal, neuropsychiatric, pulmonary, gastrointestinal, ocular and skin damage. The cumulative average prednisone dose was associated with risk of cardiovascular, ocular, musculoskeletal, neuropsychiatric and gastrointestinal damage, gonadal failure and diabetes mellitus. The simulation reproduced mean SLEDAI and organ damage scores from the Hopkins Lupus Cohort. CONCLUSION: Longitudinal modelling of an SLE cohort confirmed relationships between risk factors and long-term outcomes in SLE. The models serve to estimate the probability of SLE outcomes over time and can be used to estimate the effectiveness and cost-effectiveness of new treatments.
Authors: Manuel Francisco Ugarte-Gil; Anselm Mak; Joanna Leong; Bhushan Dharmadhikari; Nien Yee Kow; Cristina Reátegui-Sokolova; Claudia Elera-Fitzcarrald; Cinthia Aranow; Laurent Arnaud; Anca D Askanase; Sang-Cheol Bae; Sasha Bernatsky; Ian N Bruce; Jill Buyon; Nathalie Costedoat-Chalumeau; Mary Ann Dooley; Paul R Fortin; Ellen M Ginzler; Dafna D Gladman; John Hanly; Murat Inanc; David Isenberg; Soren Jacobsen; Judith A James; Andreas Jönsen; Kenneth Kalunian; Diane L Kamen; Sung Sam Lim; Eric Morand; Marta Mosca; Christine Peschken; Bernardo A Pons-Estel; Anisur Rahman; Rosalind Ramsey-Goldman; John Reynolds; Juanita Romero-Diaz; Guillermo Ruiz-Irastorza; Jorge Sánchez-Guerrero; Elisabet Svenungsson; Murray Urowitz; Evelyne Vinet; Ronald F van Vollenhoven; Alexandre Voskuyl; Daniel J Wallace; Michelle A Petri; Susan Manzi; Ann Elaine Clarke; Mike Cheung; Vernon Farewell; Graciela S Alarcon Journal: Lupus Sci Med Date: 2021-12
Authors: Ronald F van Vollenhoven; Michelle Petri; Daniel J Wallace; David A Roth; Charles T Molta; Anne E Hammer; Yongqiang Tang; April Thompson Journal: Arthritis Rheumatol Date: 2016-09 Impact factor: 10.995