A graph-based approach for characterizing resident and nurse handoff conversations

Area of Science:

• Healthcare communication
• Network science
• Clinical informatics

Background:

• Handoff communication is crucial in healthcare but challenging to analyze due to its dynamic nature.
• Existing methods for studying handoffs have limitations in capturing conversational structure and content.
• Characterizing interactive conversations like handoffs requires innovative analytical approaches.

Purpose of the Study:

• To develop and apply a graph-based approach to model handoff communication as conversation networks.
• To compare the structural properties of resident-resident and nurse-nurse handoff conversations.
• To identify differences and similarities in the structure and content of handoff communication between residents and nurses.

Main Methods:

• A graph-based approach was used to represent handoff conversations as networks.
• Resident (n=149) and nurse (n=126) handoff conversations were coded using a validated clinical content framework.
• Conversation networks were analyzed using 11 network measures and transition probabilities to identify common sub-networks.

Main Results:

• Significant differences were found in 10 out of 11 network measures between resident and nurse conversation networks.
• Nurse conversations showed more branching and switching between fewer clinical content categories compared to resident conversations.
• Both resident and nurse handoffs exhibited clinically relevant, organic relationships in the sequence of clinical content presentation.

Conclusions:

• A graph-based approach offers a valuable alternative for characterizing interactive conversations like healthcare handoffs.
• Network-based analysis can reveal significant structural differences in communication patterns between different healthcare professional groups.
• Future research should further explore network-based methods for analyzing handoff communication to enhance patient safety and care quality.