A Causal Investigation of the Effects of Advanced Lines of Sequential Therapy in Psoriatic Arthritis

Psoriatic arthritis (PsA) is a chronic inflammatory musculoskeletal condition frequently associated with psoriasis,
impacting multiple systems such as the joints, tendons, and spine. The management of this complex disease often
necessitates a sequential treatment approach, typically beginning with non-steroidal anti-inflammatory drugs
(NSAIDs) to alleviate pain and inflammation. As patients' responses to these initial therapies diminish over time,
treatment strategies may progress to more advanced options, such as disease-modifying anti-rheumatic drugs
(DMARDs) and biologics. Given the extensive range of available medications, numerous potential therapeutic
sequences can address the varied manifestations of PsA. This diversity allows for personalised treatment plans but
complicates the determination of the optimal sequence for achieving long-term disease control.
This research project aims to examine the direct effects of advanced lines of therapy on disease activity measures in
PsA patients, with the goal of improving and optimising the decision-making process for treating and managing this
condition. However, factors such as previous lines of treatment, age, and comorbidities significantly influence both
treatment choices and disease activity levels, potentially obscuring the true effects of therapy on patient outcomes.
To address this challenge, this research will employ causal inference methodologies designed to isolate and
eliminate the impact of these confounding factors. These approaches include inverse probability weighting with
marginal structural models, the parametric g-formula, and longitudinal targeted maximum likelihood estimation,
among others. Such methodologies enable a more accurate estimation of the true average causal effect of
treatments on disease activity in PsA patients.
By utilising retrospective data from PsA patients collected by the Bath Royal United Hospital and the Royal National
Hospital for Rheumatic Diseases, this project seeks to explore the implementation of various causal methodologies
and compare the resulting estimates, making adjustments where necessary. To facilitate this comparison, the
performance of each method in estimating treatment effects will be evaluated and their relative strengths and
weaknesses will be assessed. The optimal approach may then be explored among other observational studies to
investigate its generalisability This approach will allow for a more precise estimates of the average causal effect of
subsequent lines of therapy on disease activity to be obtained and will help identify the most effective methodologies
for application in clinical settings.
By effectively controlling for these confounding variables, this research aims to provide valuable insights that can
inform clinical practice, ultimately leading to enhanced management strategies for patients with PsA and improved
patient outcomes. The findings from this study have the potential to contribute to more tailored and effective
therapeutic approaches, helping healthcare professionals make informed decisions that enhance the quality of life for
those affected by this debilitating condition.