Repeat self-harm: application of hurdle models.

Area of Science:

• Public Health
• Epidemiology
• Biostatistics

Background:

• Self-harm is a significant public health concern with high rates of repetition.
• Analyzing repeat self-harm presentations is crucial for understanding risk and developing interventions.
• Current statistical methods may not fully capture the complexities of repeat self-harm events.

Purpose of the Study:

• To introduce hurdle models as a statistical approach for analyzing the number of repeat self-harm presentations.
• To demonstrate the application and interpretation of hurdle models using real-world emergency department data.
• To provide a framework for future research on self-harm repetition.

Main Methods:

• Hurdle models were employed to analyze self-harm repetition.
• Emergency department data from the Province of Ontario, Canada, were utilized for illustration.
• Statistical analysis focused on the number of repeat self-harm presentations.

Main Results:

• Hurdle models offer a flexible and informative approach to analyzing count data with excess zeros, common in self-harm studies.
• The implementation demonstrated the practical utility of these models in identifying patterns of self-harm repetition.
• Interpretation of results provides insights into factors influencing both the occurrence and frequency of repeat self-harm.

Conclusions:

• Hurdle models are a valuable statistical tool for researchers studying self-harm repetition.
• The findings support the use of advanced statistical methods to better understand and address self-harm.
• This approach can enhance the analysis of observational and intervention studies on self-harm.