Table 5 Challenges in the choice of outcomes and outcome measures for CSTs in oncology

Communication skills and the outcomes of communication encounters between health care professionals and their patients are related to many internal and external variables.

HCPs communication is influenced by trait factors such as extraversion, state variables such as current stress level and work satisfaction as well as personal knowledge. The same is true for patients, who also have different personality factors and information bases as well as emotional needs and may be at different stages in the illness trajectory. A specific communication encounter will be additionally influenced by external factors that shape the communication situation, such as availability of time and its implementation in clinical routine.

It is hard to define ‘correct’ communication behaviour.

HCPs communication styles and patients’ needs addressable by communication differ widely, both across patients and during the course of disease. Communication often takes unpredictable turns and miscommunication is frequent; this does not necessarily imply that the outcome of a miscommunication is bad.

Targeting of CST can be improved.

Highly motivated HCPs with good communication skills are more likely to take part in CSTs than HCPs with bad communication styles. Therefore, ceiling effects, both in actual effects and their measurement, have been frequently observed. Patients’ needs must be adequately addressed in the conceptualization of the training.

Learning objectives of CST vary widely.

CSTs differ widely in their specificity (generic communication training, such as active listening and expressing empathy vs. training tailored to specific communication tasks such as breaking bad news). If a CST is focused on a specific communication task, consideration needs to be given to all the skills required to satisfactorily deal with the situation.

Communication affects many different outcomes.

CSTs target many different outcome parameters. Some of them are closely connected to the content of the CST (proximal outcomes), others are influenced by many other factors as well (distal outcomes). While proximal outcomes are more likely to reflect changes after a CST, there are known problems. For example, measures of satisfaction of CST participants have frequently exhibited ceiling effects. Additionally, empathy was considered an important construct by experts but difficult to measure in an objective way. It seems to be difficult to define the appropriate measurement to capture proximal outcomes, such as clinician skill in expression of empathy. Distal outcomes such as Anxiety, Distress and Quality of Life are influenced by many other factors besides communication and the effect of a communication training on such distal outcomes has often been limited.

Validated measures are not available for specific outcomes of interest.

The limited availability of validated scales for proximal outcomes was identified by experts as a considerable barrier. This also implies that it is unclear what minimal important differences are on such scales. Scales measuring generic, broadly applicable outcomes are more likely to be used and validated. Most outcomes for which validated measures exist are distal. The imperative in research to employ validated scales might influence researchers to select generic outcomes, which may not be optimally aligned with the goals of a particular CST.