Recovery Patterns: Longitudinal Cluster Analysis of Physical Function Following Abdominal Surgery

Area of Science:

• Surgical recovery research
• Health informatics
• Patient-reported outcomes

Background:

• Increasing day surgery rates necessitate enhanced home recovery support.
• Blended eHealth offers potential for remote patient supervision.
• Lack of data-driven characterization of recovery heterogeneity hinders optimal care.

Purpose of the Study:

• To identify distinct patterns of physical function recovery after abdominal surgery.
• To analyze patient characteristics associated with different recovery trajectories.

Main Methods:

• Secondary analysis of two randomized controlled trial datasets.
• Utilized k-medoids clustering and Growth Mixture Modelling.
• Analyzed longitudinal Patient Reported Outcome Measurement Information System (PROMIS) physical function (PF) t-scores from 649 patients.

Main Results:

• Identified three consistent recovery patterns: fast, intermediate, and uneven.
• A 'relapse' cluster emerged with heavy surgery.
• Additional 'low gain' and 'high gain' clusters were observed with k-medoids clustering.

Conclusions:

• Physical function recovery post-abdominal surgery is heterogeneous.
• Distinct, clinically meaningful recovery patterns can be identified.
• These patterns can inform shared decision-making and serve as targets for predictive research.