Collecting close-contact social mixing data with contact diaries: reporting errors and biases

Area of Science:

• Epidemiology
• Network Science
• Public Health

Background:

• Understanding disease spread dynamics relies heavily on contact network analysis.
• Empirical contact data is commonly gathered via contact diaries, which depend on self-reported information.
• Validation of self-reported contact data is typically not incorporated into studies.

Purpose of the Study:

• To analyze the reporting accuracy of contact diary studies.
• To quantify the extent of non-reporting in empirical contact data collection.
• To identify factors influencing contact reporting accuracy.

Main Methods:

• A complete network study design was employed.
• Contact data was collected from employees across three research groups.
• Data was gathered over a one-week period.

Main Results:

• Over one-third of reported contacts were unilateral, meaning only one partner recorded the contact.
• Non-reporting was most prevalent for short (≤5 minutes) and less intense contacts.
• The probability of forgetting a contact lasting 5 minutes or less exceeded 50%.

Conclusions:

• Contact diary studies may significantly underestimate the true extent of social contacts.
• The accuracy of contact diaries is particularly compromised for brief interactions.
• Disease spread models relying on contact diary data may require adjustments to account for under-reporting.