CDC sets a standard measure for flu shots

NQF measure to improve comparisons

A proposed National Quality Forum measure may standardize the way hospitals calculate their health care worker influenza immunization rates.

Currently, when hospitals report their influenza immunization rates, both the numerator and denominator may vary widely. Are they counting vaccinations among employees who have direct patient contact? Or all employees, regardless of where they work? Are they including people who worked only part of the year? Are they counting agency staff or contract workers?

The measure proposed by the Centers for Disease Control and Prevention now covers the vaccination status of three groups:

Employees who worked at least 30 days during the flu season. Previously, the measure asked hospitals to include any employee who had worked for at least one day. "That [change] is going to miss a small proportion of health care personnel, but it's going to provide something that is more feasible and something hospitals may feel is more fair," says Megan C. Lindley, MPH, epidemiologist with CDC's National Center for Immunization & Respiratory Diseases. "To try to capture somebody who is in there for one day or one part of one day is potentially extremely challenging, particularly for a very large institution where you have people coming in and out."

Licensed independent practitioners. The measure will count non-employee physicians, advanced practice nurses and physician assistants, but not all credentialed employees. Again, this will make it clearer and easier for hospitals and reduce variation, says Lindley. "We found that over 70% of the hospitals credentialed their physician assistants and advanced practice nurses, and 96% of them credential their physicians," she says. By contrast, "Fewer than 20% credentialed therapists or technicians." Counting independent practitioners who don't require credentialing could present challenges for some hospitals, she says. "You could capture the bulk of the credentialed nonemployees by restricting it to those three defined groups," she says. Nurses who are credentialed through an agency would not be counted, although the hospital could require the agency to provide nurses who have been vaccinated, she says.

Non-employees. This group would be limited to students, trainees and volunteers. It would not include sales people or vendors, contract personnel, or construction workers. The previous definition of non-employees was vague and could have led to different interpretations, says Lindley." It could potentially be very, very different from facility to facility, which is contrary to the point of having a standardized measure," she says.

The numerators would be: health care personnel vaccinated at the institution and those vaccinated elsewhere, those with medical contraindications, and those who declined vaccination for non-medical reasons.

To win endorsement from the National Quality Forum, sponsors must provide data on the feasibility of implementation and the validity and reliability of the measure, Lindley says. "With these revised definitions, this provides an extremely standardized way of measuring," she says.

Hospitals had expressed concerns, especially with measuring non-employee vaccinations, in online surveys that CDC conducted with 216 health care institutions, including 80 hospitals, in four states. About half of the hospitals said their ability to determine the vaccination status of those non-employees was a major barrier.

The revised measure represents a balance designed to make measurement easier but thorough, Lindley says. "It's better to have an extremely accurate measure of 80% of personnel than it is to have an inaccurate measure that covers 100% of personnel," she says.

One thing may not change with the definitions: The burden of collecting the information. "For hospitals, in every numerator category for the three groups, paper occupational health records were the most common data source by far," she says.

In other words, most hospitals can't obtain this vaccination information simply by querying a database. Still, she says, "we were heartened that 70% of hospitals only had one person working on this [data collection]," an indication that it didn't require multiple personnel, she says.