TY - JOUR
T1 - Statistical analysis of Goal Attainment Scaling endpoints in randomised trials
AU - Urach, S.
AU - Gaasterland, C. M. W.
AU - Posch, M.
AU - Jilma, B.
AU - Roes, K.
AU - Rosenkranz, G.
AU - van der Lee, J. H.
AU - Ristl, R.
PY - 2019/6
Y1 - 2019/6
N2 - Goal Attainment Scaling is an assessment instrument to evaluate interventions on the basis of individual, patient-specific goals. The attainment of these goals is mapped in a pre-specified way to attainment levels on an ordinal scale, which is common to all goals. This approach is patient-centred and allows one to integrate the outcomes of patients with very heterogeneous symptoms. The latter is of particular importance in clinical trials in rare diseases because it enables larger sample sizes by including a broader patient population. In this paper, we focus on the statistical analysis of Goal Attainment Scaling outcomes for the comparison of two treatments in randomised clinical trials. Building on a general statistical model, we investigate the properties of different hypothesis testing approaches. Additionally, we propose a latent variable approach to generate Goal Attainment Scaling data in a simulation study, to assess the impact of model parameters such as the number of goals per patient and their correlation, the choice of discretisation thresholds and the type of design (parallel group or cross-over). Based on our findings, we give recommendations for the design of clinical trials with a Goal Attainment Scaling endpoint. Furthermore, we discuss an application of Goal Attainment Scaling in a clinical trial in mastocytosis.
AB - Goal Attainment Scaling is an assessment instrument to evaluate interventions on the basis of individual, patient-specific goals. The attainment of these goals is mapped in a pre-specified way to attainment levels on an ordinal scale, which is common to all goals. This approach is patient-centred and allows one to integrate the outcomes of patients with very heterogeneous symptoms. The latter is of particular importance in clinical trials in rare diseases because it enables larger sample sizes by including a broader patient population. In this paper, we focus on the statistical analysis of Goal Attainment Scaling outcomes for the comparison of two treatments in randomised clinical trials. Building on a general statistical model, we investigate the properties of different hypothesis testing approaches. Additionally, we propose a latent variable approach to generate Goal Attainment Scaling data in a simulation study, to assess the impact of model parameters such as the number of goals per patient and their correlation, the choice of discretisation thresholds and the type of design (parallel group or cross-over). Based on our findings, we give recommendations for the design of clinical trials with a Goal Attainment Scaling endpoint. Furthermore, we discuss an application of Goal Attainment Scaling in a clinical trial in mastocytosis.
UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85048971758&origin=inward
UR - https://www.ncbi.nlm.nih.gov/pubmed/29921167
U2 - https://doi.org/10.1177/0962280218777896
DO - https://doi.org/10.1177/0962280218777896
M3 - Article
C2 - 29921167
SN - 0962-2802
VL - 28
SP - 1893
EP - 1910
JO - Statistical methods in medical research
JF - Statistical methods in medical research
IS - 6
ER -