Multidisciplinary integration in the context of integrated care - results from the North West London Integrated Care Pilot

Background: In the context of integrated care, Multidisciplinary Group meetings involve participants from diverse professional groups and organisations and are potential vehicles to advance efficiency improvements within the local health economy. We advance a novel method to characterise the communication within Multidisciplinary Group meetings measuring the extent to which participants integrate and whether this integration leads to improved working. Methods: We purposively selected four Multidisciplinary Group meetings and conducted a content analysis of audio-recorded and transcribed Case Discussions. Two coders independently coded utterances according to their ‘integrative intensity’ which was defined against three a-priori independent domains - the Level (i.e. Individual, Collective and Systems); the Valence (Problem, Information and Solution); the Focus (Concrete and Abstract). Inter- and intra-rater reliability was tested with Kappa scores on one randomly selected Case Discussion. Standardised weighted mean integration scores were calculated for Case Discussions across utterance deciles, indicating how integrative intensity changed during the conversations. Results: Twenty-three Case Discussions in four different Multidisciplinary Groups were transcribed and coded. Inter- and intra-rater reliability was good as shown by the Prevalence and Bias-Adjusted Kappa Scores for one randomly selected Case Discussion. There were differences in the proportion of utterances per participant type (Consultant 14.6%; presenting general practitioner 38.75%; Chair 7.8%; nonpresenting general practitioner 2.25%; Allied Health Professional 4.8%). Utterances were predominantly coded at low levels of integrative intensity; however, there was a gradual increase (R2 = 0.71) in integrative intensity during the Case Discussions. Based on the analysis of the minutes and action points arising from the Case Discussions, this improved integration did not translate into actions moving forward. Interpretation: We characterise the Multidisciplinary Groups as having consultative characteristics with some trend towards collaboration, but that best resemble Community-Based Ward Rounds. Although integration scores do increase from the beginning to the end of the Case Discussions, this does not tend to translate into actions for the groups to take forward. The role of the Chair and the improved participation of non-presenting general practitioners and Allied Health Professionals seems important, particularly as the latter contribute well to higher integrative scores. Traditional communication patterns of medical dominance seem to be being perpetuated in the Multidisciplinary Groups. This suggests that more could be done to sensitise participants to the value of full participation from all the members of the group. The method we have developed could be used for ongoing and future evaluations of integrated care projects