Five types of clinical difference to monitor in practice.

A fundamental part of professional practice is to monitor case progress to inform basic clinical decisions about when to discontinue interventions, when to adjust interventions, and when to proceed as planned. When interpreting observed change, there are at least five distinct types of clinical difference that can occur independently, and mistaking one for another can lead to misinformed clinical decisions. We introduce a distinction between observed difference, detected difference, predicted difference, attainment difference, and induced difference, and use these to analyze current systems for routine outcome monitoring (ROM) in clinical practice. Contrary to what supporting evidence of current ROM systems suggests, we find that—by design—these systems fail to detect and predict potentially harmful treatment and fail to detect and predict likely treatment responders. We discuss implications of the presented classification for professional practice and further development of ROM systems. We argue that clinical practice would benefit from monitoring difference instead of change, as these are not equivalent. We further argue that future research and development efforts should focus on the development of a working approach to monitoring induced difference, improving how to monitor predicted difference, exploring statistical models that better discriminate between various types of clinical cases, and better communicate what can and cannot be interpreted from the clinical differences that are actually monitored, as guided by the presented classification of clinical differences to monitor in practice. (PsycINFO Database Record (c) 2018 APA, all rights reserved)