Absenteeism has been of interest as a potential early indicator of outbreaks in a population. Among the earliest studies demonstrating this potential was work by Costagliola et al. (1991) on modeling of data surrounding French influenza outbreaks. They found that increased absenteeism historically preceded influenza outbreaks. Later, Quenel et al. (1998) confirmed this finding and compared absenteeism to other data sources, identifying it as a leading (temporal) indicator of influenza outbreaks. About this time, Lenaway and Ambler (1995) also noted links between school absenteeism and influenza outbreaks.

Absenteeism is also of interest because it offers windows into symptoms in particular populations that may be at special risk for outbreaks. If the outbreak is related to some process or toxin in the workplace or the learning institution, absenteeism provide a very early sign that an outbreak is occurring in that population. For example, absenteeism among postal workers might be an early sign of some mail-borne outbreak. Both the time and the subpopulation focus of absenteeism data make them one of the more potentially important and interesting types of data for health surveillance.

One of the primary reasons to monitor absenteeism data is that many individuals with illnesses avoid contact with the healthcare system. Johnson et al. (2005) studied the antecedent health behaviors of pediatric patients reporting to the emergency room with viral-like syndromes. In syndromes with fever, absenteeism preceded emergency room visits in 44-50% of patients by a mean of 1.3 days. This study is somewhat limited by its focus on the emergency room. Metzger et al. (2004) examined the relationship between emergency room visits and illnesses in the community using a telephone survey. They found that every emergency department visit for an influenza-like illness represented approximately 60 cases of such illnesses in the community. Of these, approximately 32.6% were accompanied by absenteeism from work or school. This extrapolates to 20 potentially measurable absences for every emergency room visit. The only activity that occurred more frequently in this population was a purchase of over-the-counter medication (53.2%).

Because of the desirable properties of absenteeism as an early indicator in many monitoring systems, including ESSENCE II in the nation's capital region (Lombardo et al., 2003), the New York City Public Health Department systems (Heffernan et al., 2004), and the Bionet system in San Diego, all use absenteeism data for health monitoring. This chapter will describe how systems for absenteeism work, review the present data for evidence of the effectiveness of this information source, and speculate about future models for surveillance that use improved methods to take advantage of the social structures that create absenteeism data.

Was this article helpful?

0 0

Post a comment