Cost-Effectiveness of Influenza Treatment

Abstract & Commentary

Synopsis: For unvaccinated or high-risk vaccinated patients during the influenza season, empirical oseltamivir treatment is cost-effective. For other patients, rapid diagnostic testing followed by treatment with oseltamivir is cost-effective. Empirical amantadine treatment offers a low-cost alternative if patients cannot afford oseltamivir.

Source: Rothberg MB, et al. Ann Intern Med. 2003;139:321-329.

Using decision analysis, the cost-effectiveness of rapid diagnostic testing and antiviral therapy for influenza-like illness was evaluated in older adults (older than 65). The model was based on several key assumptions, including that neuraminidase inhibitors would decrease hospitalizations by 33%, along with a comparable reduction in complications and antibiotic use, although zanamivir would be somewhat less effective than oseltamivir because ~50% of older adults would have difficulty loading and administering the medication. In addition, it was assumed that amantadine and rimantadine were less effective because they would not reduce hospitalization and are only effective against influenza A virus. The model was not sensitive to the prevalence or severity of medication side effects.

Not surprisingly, in adults at greatest risk for influenza—unvaccinated, institutionalized, nursing home residents, etc—empiric treatment with oseltamivir without diagnostic testing was the most cost-effective strategy. In patients at lower risk (eg, those who have been vaccinated), rapid diagnostic testing followed by appropriate therapy with oseltamivir was the most cost-effective approach. Empirical amantadine was less cost effective in either circumstance, but could be used if patients had to pay for drug out of their own pocket and couldn’t afford the more expensive agent. (This sounds a bit like backward logic to me: Oseltamivir is cost-effective as long as someone else is paying for it?)

Comment by Carol A. Kemper, MD

While these conclusions largely make sense, one wonders how the model would perform in the current year when the predominant circulating strain of influenza virus (eg, A/Fujian/411/2002) is not included in (although may be partially covered by) the current vaccine.

Dr. Kemper is Clinical Associate Professor of Medicine, Stanford University, Palo Alto, Calif.