The determinants of expert opinion in the development of care pathways: insights from an exploratory cluster analysis

Abstract

Abstract Background We performed a secondary exploratory cluster analysis on the data collected from the validation phase of the study leading to the development of the model care pathway (CP) for Myasthenia Gravis (MG), in which a panel of 85 international experts were asked some characteristics about themselves and their opinion about the model CP. Our aim was to identify which characteristics of the experts play a role in the genesis of their opinion. Methods We extracted the questions probing an opinion and those describing a characteristic of the expert from the original questionnaire. We performed a multiple correspondence analysis (MCA) and a subsequent hierarchical clustering on principal component (HCPC) on the opinion variables, integrating the characteristic variables as supplementary (predicted). Results After reducing the dimensionality of the questionnaire to three dimensions we noticed that the not-appropriateness judgement of the clinical activities may overlap with the completeness one. From the HCPC it seems that the working setting of the expert may play a crucial role in determining the opinion about the setting of the sub-processes of MG: shifting from a cluster where the experts do not work in sub-specialist settings to one where the experts are working in them, the opinion changes accordingly from a mono-disciplinary setting to a multi-disciplinary one. Another interesting result is that the experience in neuromuscular diseases (NMD) measured in years and the expert typology (whether general neurologist or NMD expert) seem not to contribute significantly to the opinions. Conclusions These findings might indicate a poor ability of the expert to discriminate what is not appropriate from what is not complete. Also, the opinion of the expert might be influenced by the working setting, but not by the experience in NMD (as measured in years).

Keywords

• Care pathways
• Expert opinion
• Cluster analysis
• Myasthenia gravis
• Multiple correspondence analysis
• Factorial analysis