Identifying Longitudinal Patterns for Individuals and Subgroups: An Example with Adherence to Treatment for...

Area of Science:

• Behavioral Science
• Medical Adherence Research
• Sleep Medicine

Background:

• Improving complex behaviors like adherence to medical recommendations requires understanding behavior change over time.
• Adherence to the most common treatment for obstructive sleep apnea (OSA) is notably poor, necessitating further investigation.

Purpose of the Study:

• To investigate temporal patterns of treatment adherence in patients with obstructive sleep apnea (OSA).
• To develop a typology of temporal patterns (TTP) in adherence using time series and cluster analysis.

Main Methods:

• A sample of 161 participants with approximately 180 nights of data each.
• Individual time series analysis (mean, level, slope, variance, autocorrelation) followed by dynamic cluster analysis.
• Four distinct adherence subgroups were identified and labeled: Great Users, Good Users, Low Users, and Slow Decliners.

Main Results:

• Four homogenous subgroups of adherence patterns were identified: Great Users (17.2%), Good Users (32.8%), Low Users (22.7%), and Slow Decliners (27.3%).
• Subgroups differed in mean usage, level, slope, and variance, indicating distinct adherence trajectories.
• External validity was established by comparing participants across identified subgroups.

Conclusions:

• The Typology of Temporal Patterns (TTP) approach, combining time series and dynamic cluster analysis, effectively evaluates longitudinal adherence patterns.
• This methodology provides a nuanced understanding of individual and subgroup adherence behaviors in OSA patients.
• Identifying these distinct patterns can inform targeted interventions to improve long-term treatment adherence.