To the editor

The valid concern about the effectiveness of oseltamivir [1] is particularly pertinent to the management of influenza outbreaks in aged care facilities (ACFs). In many developed countries, guidelines recommend the prophylactic use of antivirals in confirmed institutional influenza outbreaks however there is very limited high quality publicly available data to support this practice.

Until recently the only randomised controlled trial data from the aged care setting found that seasonal prophylaxis with oseltamivir had no significant impact on overall laboratory-confirmed influenza but reported a significant reduction in symptoms [2].

The recent paper by Booy et al [3] was a welcome addition to this field, but given the limited existing evidence base, it is particularly important that any study qualifications are carefully considered.

The use of oseltamivir for both treatment and prophylaxis (T&P) in confirmed ACF influenza outbreaks was reported to significantly reduce outbreak duration and attack rate in comparison to use of oseltamivir for treatment only (T). The authors concluded that the trial provided “some support for a policy of treatment and prophylaxis with oseltamivir in controlling influenza outbreaks in ACFs”. However, differences in deaths, hospitalisations and pneumonia were not significant. Popular medical media have lauded the work as a “landmark study” and reported that the “results provide good evidence to support an active policy of treating and preventing influenza promptly, once an outbreak is declared” [4].

Unfortunately the underpowered nature of this study, with just three facilities randomised to T and six to T&P, meant that a single outbreak could substantially impact on the results.

Indeed, inclusion of a particular T outbreak (Facility B) seriously limits the interpretation of this study. Data from the paper and subsequently provided by the authors, indicated that the time from first influenza-like illness (ILI) case to intervention commencing was approximately 20 days, at which time there were already 22 ILI cases in the facility. Treatment was provided to fewer than half the ILI cases and the intervention was subsequently changed to T&P after about 10 days. In contrast, the mean time from first ILI case to outbreak intervention was 12.7 days in T facilities overall and 5.0 days in T&P facilities.

The unusual nature of the outbreak in Facility B and its management dictate that the analysis is repeated excluding this facility. It appears that the reported difference in outbreak duration would be non-significant if facility B was excluded, providing important qualification to the study conclusions. Differences in hospitalisations and deaths would be further reduced and remain non-significant.

The evidence base for guiding antiviral use in ACFs remains limited. In the absence of further compelling trial data from Roche, there is a need for additional adequately powered research before firm recommendations can be made.

Tony Merritt
Public Health Physician
Hunter New England Population Health

David Durrheim
Professor of Public Health Medicine
University of Newcastle, Australia

References
(1) Krumholz HM, Jackevicius CA, Ross JS. Tamiflu: 14 flu seasons and still questions. BMJ 2013;346:f547.
(2) Peters PH Jr., Gravenstein S, et al. Long-term use of oseltamivir for the prophylaxis of influenza in a vaccinated frail older population. Journal of the American Geriatrics Society 2001 Aug;49(8):1025-31
(3) Booy R, Lindner R, Dwyer DE, Ying M. Treating and preventing influenza in Aged Care facilities: a cluster randomised controlled trial. PLoS ONE Open Access Online 7(10): e46509. 2012.
(4) A new approach to deadly influenza outbreaks in nursing homes. 18 October 2012. [Cited 19 January 2013]. Available from URL http://www.healthcanal.com/infections/33099-new-approach-deadly-inflenza...

Competing interests: None declared