Quantitative COA research
PRO and other COA data is complex with multiple time points, multiple domains and missing data often related to the outcome of interest. The statistical techniques required to analyse PRO/COA data therefore require specialist expertise, particularly as there are no accepted standards for the analysis of PRO/COA data and a range of statistical methods are available. In evaluating the psychometric properties of PRO/COA instrument expert input from statisticians and psychometricians is critical to know which of an array of classical and modern psychometric methods is most appropriate for a given instrument and context. Moreover, interpretation is also not just about statistics, it has to be put into context of clinical relevance. Knowledge of the instruments, statistical techniques and clinical interpretation of scores is essential to optimize the use of patient-reported outcomes data.
We offer specialist statistical expertise to support patient centered outcomes data analysis to provide solutions for:
- Endpoint development
- Psychometric evaluation and score interpretation
- Clinical trial design
- Statistical analysis plans
- Clinical trial analysis
- Bespoke training
Our in-house experts offer specific skills and programming expertise which includes:
- Repeated-measures modelling
- Structural equation modelling
- Pattern-mixture modelling
- Latent trait modelling
- Imputation survival analysis
We perform the majority of analyses using Statistical Analysis Software (SAS), the industry standard, and prepare data tables following CDISC standards. However, where specialist software is more appropriate we also use other software programs such as Stata and R and have the expertise to do so in-house. For any analysis project an expert statistician/psychometrician will lead the development of a Statistical Analysis Plan (SAP) a priori that specifies all planned analyses to be implemented by our team of expert SAS programmers.
We also think carefully about what is the best way to present data that is clear, transparent as well as meaningful and easily interpeted by the intended audience. In particular this can include the use of novel graphics for presenting PRO data across multiple domains and time points. In our view, presenting and describing complex data in a clear manner that can be understood by regulators and other less expert stakeholders is as important as the analyses we perform.
Reck M, Taylor F, Penrod JR, DeRosa M, Morrissey L, Dastani H, Orsini L, Gralla RJ. Impact of nivolumab versus docetaxel on health-related quality of life and symptoms in patients with advanced squamous non-small-cell lung cancer: results from the CheckMate 017 study. J Thorac Oncol. 2018;13(2):194-204
C. Coon, A. Bushmakin, S. Tatlock, N. Williamson, M. Moffatt, R. Arbuckle & L. Abraham. Evaluation of a crosswalk between the European Quality of Life Five Dimension Five Level and the Menopause-Specific Quality of Life questionnaire. Climacteric 2018; 8: 1-8
Musoro ZJ, Hamel J-F, Ediebah DE, Cocks K, King MT, Groenvold M, Sprangers MAG, Brandberg Y, Velikova G, Maringwa J, Flechtner H-H, Bottomley A, Coens C. Establishing anchor-based minimally important differences (MID) with the EORTC quality-of-life measures: a meta-analysis protocol. BMUJ Open 2018;8:e019117
Nguyen A, Arbuckle R, Korver T, Chen F, Taylor B, Turnbull A, Norquist J. Psychometric validation of the dysmenorrhea daily diary (DysDD): a patient-reported outcome for dysmenorrhea. Qual Life Res. 2017; 26(8): 2041-2055