Pandemic plan: How many HCWs will be home sick?

Software predicts pandemic absenteeism impact

To prepare for an influenza pandemic, you've counted your ventilators and calculated your patient surge, sought options to add bed space and assessed the capacity of the emergency department. But have you estimated how many of your employees will show up for work?

FluWorkLoss software from the Centers for Disease Control and Prevention in Atlanta enables hospitals to estimate employee absenteeism based on the severity and length of a pandemic. The default program is based on a moderate pandemic similar to the one in 1968, with a 15% to 35% clinical attack rate. To test the consequences of a more severe pandemic, software users can change the rate of deaths, hospitalizations and outpatient visits.

Employers create scenarios. How long with the pandemic last? A short time frame (four weeks) means a higher peak of absences; a longer pandemic (up to 12 weeks) means the illnesses will be spread out.

You can alter assumptions about how many days employees will be out of work with the flu or how long some might need to be hospitalized.

"We really do encourage you go in and play endless 'What if's?'" says Martin Meltzer, PhD, senior health economist at CDC. During seasonal flu, sick employees typically stay home less than one day. During a pandemic, they may be sicker and employers may want them home for 24 hours or more after symptoms subside, for a total absence of at least five days. "The employees and the employers can see the impact of these [scenarios]," he says.

Employees also may stay home to care for ill family members or to take care of children who have been sent home as schools are closed. To estimate the impact of that, you may want to get some information on family caregiving responsibilities by identifying how many employees listed dependents on their health insurance or by conducting an anonymous survey, Meltzer advises.

The FluWorkLoss software includes a measure for "cohabiting" adults, who may need to take care of a spouse or partner. However, it doesn't account for employees who may not show up for work out of fear of contracting the illness. Surveys of health care workers indicate that 50% may be not show up because of fear of risk to themselves or their family members.1 Meltzer advises employers to add in a factor for that additional absenteeism.

While the software was based on data from previous pandemics, there are many unknowns and there's little data on how many days a worker might be absent while recovering from a pandemic strain of influenza or caring for a sick family member. For example, the software has a default value of 40 days to replace a health care worker who died from pandemic influenza. Users can change that to another value.

Despite its limitations, the software can be useful in providing a framework for various scenarios. "I hope people will start thinking of how they would replace [workers], how they would stretch out the force, what kind of cross-training they'll do," says Meltzer. "[They should consider] what are the most essential activities that need to carry on and what can be shut down."

(Editor's note: FluWorkLoss is available at


1. Qureshi KA, Gershon RRM, Staub T, et al. Healthcare workers' willingness to report to duty during catastrophic disasters. J Urban Health 2005; 82:378-388.