And now for something completely different...health economics of food allergy
Published online: May 8, 2018
If we know anything about peanut allergy, we know two guiding principles: strict allergen avoidance including precautionary allergen labeling (PAL, e.g., “may contain”) and very prompt epinephrine administration, which of late has even included injecting epinephrine immediately for known or suspected ingestion, even before symptoms have developed (e.g., pre-emptive use) to make sure it is given promptly and without delay. Right? These are long-established tenants of management. But while this may be going dogma, are these policies aligned with providing value for such care, and producing the best outcomes in managing food allergy? What if what we accepted as dogma may need some realignment?
In an article recently published in The Journal of Allergy and Clinical Immunology: In Practice, Shaker and Greenhawt tackle these issues from a health economics perspective. They explored the cost-effectiveness of pre-emptive epinephrine use as well as strict avoidance of products with PAL for peanut using Markov modeling, a technique that allows for health state transitions and recurrent probabilistic risk of two scenarios over a 20-year horizon. Model A explored pre-emptive epinephrine use compared to use only after symptoms of a moderate to severe reaction . Model B explored a strategy of allowing consumption of goods with peanut allergy precautionary labeling vs. strict avoidance of such goods in all subjects. Differences in the two strategies were modeled at a 10-fold presumed risk between the approaches (i.e. the more aggressive approach was associated with a 10-fold increased risk of a food allergy fatality occurring), but also explored presuming 100, 500, and 1000-fold increased risks between the approaches for sensitivity analysis, and explored differences in quality of life. Model B also explored an alternative strategy of deliberately offering a low dose “one-shot” challenge of 1.5mg peanut protein to all peanut allergic patients and permitting PAL consumption if this is tolerated, a recently published strategy associated with improved quality of life. Key outcomes were total costs, rates of fatality, and quality of life adjusted years.
This study revealed two important and policy relevant conclusions. First, at the presumed 10-fold risk reduction, pre-emptive use of epinephrine (immediately administer epinephrine if the suspected allergen was “definitely eaten,” even if no symptoms are present) is not cost-effective—it is associated with $110,270,820 spent per death prevented, and a risk reduction of <0.0001 per-patient fatalities over the 20 year horizon). Even exaggerated to a 1000-fold risk reduction, pre-emptive use remained >$100,000 per quality of life adjusted years (QALY), the threshold for cost-effective care but was dominated by use only when moderate to severe symptoms emerged (dominated=the alternative strategy produced superior health benefits at a lower cost). Secondly, strict avoidance of PAL was associated with a similar minute reduction in fatalities and an incremental cost per death prevented of $182,434,277. However, if this base case accounted for QoL improvement associated with PAL ingestion, allowing PAL ingestion dominated. Most interestingly, if a “one shot” challenge was incorporated into routine clinical practice prior to PAL consumption, this strategy dominated all others.
With respect to peanut allergy, we provide a lot of advice, but we rarely consider the value of such decisions. This study takes an evolved look at potential peanut allergy management strategies, and challenges two commonly advised practices in pre-emptive epinephrine use and strict avoidance of PAL. At the low rates of food allergy fatality and actual peanut contamination of PAL products, these are costly strategies with a minimal impact on preventing fatal peanut allergy. Their alternative strategies, waiting for symptoms to emerge before using epinephrine and either allowing patient-preference sensitive PAL consumption or performing a “one shot” 1.5mg peanut protein challenge in all peanut allergic patients prior to permitting PAL consumption, were associated with both lower costs and superior health outcomes. The use of a low dose single peanut challenge could prove to be of high utility, as it can greatly improve the child’s QoL, while allowing the diet to be safely expanded. While further studies clarifying rates of actual PAL contamination are needed, these simulated scenarios help us to understand situations where an abundance of caution has often has driven our approach. In the case of these two policies, these data show we can do better by better understanding costs, risks, and patient preferences.
The Journal of Allergy and Clinical Immunology: In Practice is an official journal of the AAAAI, focusing on practical information for the practicing clinician.