Cost-effectiveness Analysis of Antiviral Treatment in the Management of Seasonal Influenza A

Point-of-care Rapid Test Versus Clinical Judgment

Léon Nshimyumukiza; Xavier Douville; Diane Fournier; Julie Duplantie; Rana K. Daher; Isabelle Charlebois; Jean Longtin; Jesse Papenburg; Maryse Guay; Maurice Boissinot; Michel G. Bergeron; Denis Boudreau; Christian Gagné; François Rousseau; Daniel Reinharz


Influenza Resp Viruses. 2016;30(2):113-121. 

In This Article

Abstract and Introduction


Background A point-of-care rapid test (POCRT) may help early and targeted use of antiviral drugs for the management of influenza A infection.

Objective (i) To determine whether antiviral treatment based on a POCRT for influenza A is cost-effective and, (ii) to determine the thresholds of key test parameters (sensitivity, specificity and cost) at which a POCRT based-strategy appears to be cost effective.

Methods An hybrid « susceptible, infected, recovered (SIR) » compartmental transmission and Markov decision analytic model was used to simulate the cost-effectiveness of antiviral treatment based on a POCRT for influenza A in the social perspective. Data input parameters used were retrieved from peer-review published studies and government databases. The outcome considered was the incremental cost per life-year saved for one seasonal influenza season.

Results In the base-case analysis, the antiviral treatment based on POCRT saves 2 lives/100 000 person-years and costs $7600 less than the empirical antiviral treatment based on clinical judgment alone, which demonstrates that the POCRT-based strategy is dominant. In one and two way-sensitivity analyses, results were sensitive to the POCRT accuracy and cost, to the vaccination coverage as well as to the prevalence of influenza A. In probabilistic sensitivity analyses, the POCRT strategy is cost-effective in 66% of cases, for a commonly accepted threshold of $50 000 per life-year saved.

Conclusion The influenza antiviral treatment based on POCRT could be cost-effective in specific conditions of performance, price and disease prevalence.


Influenza causes over than 4000 deaths in Canada annually, the large majority of which are attributable to type A strains.[1] While vaccination is the cornerstone of prevention, antiviral drugs are the only specific medication for influenza infection. They are most effective at reducing complications when used early (within 48 hour of illness onset).[2] Management of influenza infections remains a challenge, mainly because of the difficulty in making a rapid and accurate diagnosis. Clinical diagnostic criteria lack accuracy compared with laboratory methods because influenza causes a wide spectrum of disease that is often clinically indistinguishable from other respiratory infections.[3] However, results of traditional microbiological tests are not available to practitioners in a clinically relevant timeframe, obliging clinicians to use an empirical approach when suspecting influenza A infection. Furthermore, currently available rapid diagnostic tests for influenza A have a low sensitivity.[4] Thus, there is some interest in the development of new diagnostic tools that are simple enough to be used at the bedside (i.e., point-of-care [POC]) for rapid and reliable diagnosis. Such rapid tests would allow an early and targeted use of antiviral drugs for patients with influenza infection, thus improving their outcomes.

Economic studies comparing rapid testing to clinical diagnosis of influenza remain inconclusive. Indeed, some studies suggested that, in most cases, clinical judgment combined with antiviral treatment is the most cost-effective strategy,[5] while others suggested that testing may be the most cost-effective strategy.[6] In addition, even if studies agreed that the cost-effectiveness of rapid tests is sensitive to their accuracy and costs, to the prevalence of influenza and its complications as well as to the vaccination status, no one has, to our knowledge, specifically estimated the thresholds of accuracy and costs from which the rapid testing strategy could be considered as cost-effective compared to clinical judgment.

Considering that developing such a test is time- and resource-consuming, it is therefore relevant to define the thresholds of accuracy and costs from which a new rapid POC test is expected to be cost-effective and to do it before its development and its implementation in general practice.

Using a hybrid transmission and decision analytic economic model, the objective of this study was to address two principal questions: (i) Is an antiviral treatment based on a rapid POC test for influenza A cost-effective compared to the empirical treatment based on clinical judgment? and (ii) At what thresholds of POC test accuracy (sensitivity and specificity) and cost is the testing strategy cost-effective?

We aimed to provide generic information from an economic point of view regarding the optimal characteristics of POC tests (accuracy and cost) in order to help industries to decide about the interest of developing such assays, and to help decision-makers to establish the relevance of implementing such assays within a healthcare system.