Modelling memory decay after injuries using household survey data from Khartoum State, Sudan

Area of Science:

• Epidemiology
• Public Health
• Biostatistics

Background:

• Injuries are a major global cause of illness and death.
• Retrospective injury studies often underestimate rates due to memory decay.
• The influence of external factors on memory decay in injury reporting is not well understood.

Purpose of the Study:

• To develop a statistical model to differentiate true injury rates from memory decay effects.
• To investigate how external factors influence memory decay in injury recall.
• To analyze injury data from a household survey in Khartoum State.

Main Methods:

• A parametric regression model was used to assess the relationship between apparent injury rates and time since injury.
• Effect modification was employed to model the impact of external factors on memory decay.
• Data from a 2010 retrospective household survey of 5661 individuals reporting 481 injuries were analyzed.

Main Results:

• Memory recall varied by socioeconomic status, with faster decay in the lowest tertile (18% recall at 6 months vs. 35% in others).
• Memory decay was independent of sex, age, urban/rural status, and education.
• Road traffic injuries showed less memory decay than falls, mechanical injuries, or burns.

Conclusions:

• Memory decay significantly distorts crude and relative injury rates in retrospective studies.
• Statistical models accounting for memory decay are crucial for accurate injury data analysis.
• Socioeconomic disparities influence injury recall, highlighting the need for context-specific analysis.