A test for detection of clustering over time

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Clustering of health events in time can indicate underlying environmental or biological factors.
• Accurate statistical methods are crucial for identifying and validating such clusters.
• Spontaneous abortions with trisomies represent a significant public health concern.

Purpose of the Study:

• To discuss the application of the scan statistic for temporal disease cluster detection.
• To analyze a specific reported cluster of trisomies in spontaneous abortions in New York City using this method.

Main Methods:

• The scan statistic is defined as the maximum observed cases within a sliding time interval.
• The methodology involves applying this statistic to time-series data.
• A distribution table for the scan statistic is provided for lower tail analysis.

Main Results:

• The scan statistic effectively identifies periods of increased event occurrence.
• The study provides a framework for assessing the significance of temporal clusters.
• The distribution table aids in interpreting the observed scan statistic values.

Conclusions:

• The scan statistic is a valuable tool for investigating temporal disease clustering.
• This method can be applied to analyze specific health events like trisomies in spontaneous abortions.
• Further research can utilize this statistic to monitor and understand disease patterns over time.