Measurement, Implementation and Interpretation
By Joseph C. Cappelleri, Kelly H. Zou, Andrew G. Bushmakin, Jose Ma. J. Alvir, Demissie Alemayehu, Tara Symonds
Chapman and Hall/CRC – 2013 – 351 pages
Advancing the development, validation, and use of patient-reported outcome (PRO) measures, Patient-Reported Outcomes: Measurement, Implementation and Interpretation helps readers develop and enrich their understanding of PRO methodology, particularly from a quantitative perspective. Designed for biopharmaceutical researchers and others in the health sciences community, it provides an up-to-date volume on conceptual and analytical issues of PRO measures.
The book discusses key concepts relating to the measurement, implementation, and interpretation of PRO measures. It covers both introductory and advanced psychometric and biostatistical methods for constructing and analyzing PRO measures. The authors include many relevant real-life applications based on their extensive first-hand experiences in the pharmaceutical industry. They implement a wealth of simulated datasets to illustrate concepts and heighten understanding based on practical scenarios. For readers interested in conducting statistical analyses of PRO measures and delving more deeply into the analytic details, most chapters contain SAS code and output that illustrate the methodology. Along with providing numerous references, the book highlights current regulatory guidelines.
"The book is well written and well organized. Good examples contribute to a better understanding of statistics. The focus is quantitative and includes psychometric and statistical methods relevant to the vast majority of patient-reported outcome measures, those based on multi questions scales or summed assessment scales. The content is also relevant for individualized and preference-based instruments. … The book is recommended for health care professionals and health researchers who wish to further their understanding of psychometric and statistical methods necessary for the development, testing and use of patient-reported outcome measures."
—A Garratt, in Journal of the Norwegian Medical Association, April 2015
"This book includes 11 chapters on 330 pages. The first six chapters cover topics in psychometrics, and the remaining five cover topics in statistics. … the book is a quick, easy, and understandable introduction to qualitative and quantitative psychometricmethods for the inexperienced reader."
—John Brodersen, University of Copenhagen, in Statistics in Medicine, 2015, 34 1799-1800
"The recent book, Patient-Reported Outcomes, Measurement, Implementation and Interpretation, by Cappelleri and his colleagues is a timely synthesis of the current thinking and practice of psychometrics applied to patient reported outcomes (PROs), and PRO data analysis. Written primarily by statisticians drawn from the same large multinational pharmaceutical company, the book is squarely aimed at health economists, outcomes researchers and statisticians in the industry who may require a foundation in psychometrics and the analysis of PRO data. That said it may also be of benefit and utility to advanced researchers in academia and consultancy wanting to increase their knowledge in this field. … In summary, this book attempts to cover a lot of complex material and does so successfully. It effectively teaches core concepts related to the measurement, analysis and interpretation of PROs. There is something for readers of all levels of familiarity with PROs, from the novice to the more experienced researcher."
—Book review appearing in Qual Life Res, (2014) 23:2405–2406
"… a quick, easy, and understandable introduction to qualitative and quantitative psychometric methods for the inexperienced reader."
—Statistics in Medicine, 2015
"This is a timely publication that provides comprehensive information on an important subject. … there are many who can benefit from reading this book, including health care professionals, health services researchers, policy makers, regulators, patients, caregivers, and other health-related stakeholders. … This book includes many excellent simulated and real-life examples. … For many practitioners, the sample codes provided in the book provide an excellent template for them to carry out their own analyses. This book is well organized and clearly written. Accessible language and articulate writing make this a very readable book. Ample examples with step-by-step illustrations of the analysis of data and interpretation of the results will be particularly helpful to readers who are new to PROs. It can also serve as a valuable reference for those who have experience working with PROs."
—Journal of Biopharmaceutical Statistics, 2014
Patient-Reported Outcomes in Perspective
Patient-Reported Outcomes in Clinical Research
Terms and Definitions
Psychometrics vs Clinimetrics
Selection of a PRO Questionnaire
Development of a Patient-Reported Outcome
Simulated Example Using SAS: Convergent and Divergent Validity
Factors Affecting Response
Intraclass Correlation Coefficient for Continuous Variables
ICC Simulated Example
ICC in Context
Bland and Altman Plot for Continuous Variables
Simple Kappa and Weighted Kappa Coefficients for Categorical Variables
Internal Consistency Reliability: Cronbach’s Alpha Coefficient
Simulated Example of Cronbach’s Alpha
Exploratory and Confirmatory Factor Analyses
Exploratory Factor Analysis
Confirmatory Factor Analysis
Causal Indicators vs Effect Indicators
Simulated Examples Using SAS: Exploratory Factor Analysis
Simulated Examples Using SAS: Confirmatory Factor Analysis
Item Response Theory
Classical Test Theory Revisited
Assumptions of IRT
Item Characteristic Curves
Item Fit and Person Fit
Differential Item Functioning
Example: Rasch Model Implementation
Types of PRO Data and Exploratory Methods
Comparing Two or More Samples
Repeated Measures Model
Random Coefficient Model
Single Mediator Model
Bootstrapping Methodology Implementation
Study Design to Minimize Missing Data
Missing Data Patterns and Mechanisms
Approaches for Missing Items within Domains or Measures
Approaches for Missing Entire Domains or Entire Questionnaires
Simulated Example Using SAS: Pattern Mixture Models
A Summary and References appear at the end of each chapter.