Box-Jenkins modelling in medical research

Area of Science:

• Statistics
• Environmental Medicine
• Biostatistics

Background:

• Time series data, such as disease notifications or pollutant concentrations, often exhibit dependence between consecutive measurements.
• This temporal dependence, or 'inertia', is significant in environmental medicine and clinical data analysis (e.g., blood glucose levels).
• Traditional analysis methods may not adequately capture these dependencies.

Purpose of the Study:

• To review the fundamental concepts of Box-Jenkins modelling, specifically Autoregressive Integrated Moving Average (ARIMA) models.
• To illustrate the application of these time series analysis techniques.
• To present key topics including ARIMA models, transfer function models, and intervention analysis.

Main Methods:

• Explanation of the ARIMA model for modeling stochastic dependence in consecutive data.
• Introduction to transfer function models for assessing relationships between multiple time series.
• Description of intervention analysis for evaluating changes or impacts on time series.

Main Results:

• Demonstration of how ARIMA models can effectively capture the 'inertia' in sequential data.
• Illustrative applications showing the practical utility of Box-Jenkins methods.
• Methods presented are applicable to various scientific fields requiring time series analysis.

Conclusions:

• Box-Jenkins (ARIMA) modelling provides a robust framework for analyzing dependent time series data.
• These methods are valuable tools in environmental medicine, healthcare, and economics.
• The reviewed techniques offer powerful approaches for understanding temporal patterns and relationships in data.