That’s the title of a great review. Article by Salah Ghabri, Dalia Dawoud and Michael Drummond. This article notes that guidance on incorporating adverse events (AEs) into economic models has not been standardized. Nevertheless, this article provides useful guidance. Specifically, the article explores: (1) incorporating AEs into economic decision models, (2) the types of AEs that should be included, (3) estimating their cost and quality of life consequences, and (4) exploring the uncertainty related to their impact on the ICER. I summarize each topic in turn.
1. Ways to incorporate EA into economic models:
- Included in the structure of the analytical modelFor example, AEs could be an independent health condition. The authors cite an article by Stevenson et al. (2016) which evaluates the cost-effectiveness of new
Rheumatoid arthritis treatments. Their model included patient health states defined as having a moderate to good response at 6 months with no serious adverse effects or patients with no response or experiencing one serious adverse effect. This approach is most useful if adverse effects are severe and relatively common. - Included as input parametersIn this approach, AE input parameters are incorporated into each health state. For example, parameterization with respect to incidence rates, time to onset, quality of life loss, and costs of each AE would be applied to each relevant health state. This approach is the most widely used in the literature. AEs are often modeled from clinical trial data and are only incorporated in the first model cycle or the first year; the long-term consequences of AEs are often ignored.
2. Terminology and types of EA to be included
Adverse events are often referred to as adverse events, but other terms are sometimes used, such as toxicity, side effects, adverse drug effects, drug effects, serious infection, or adverse reaction. A key difference between adverse events is that some are chronic in nature and others are acute.
Although adverse effects are often identified from clinical trials, real-world data can be used. For pharmaceuticals, post-marketing safety surveillance is mandatory, but for medical devices this is not always the case (depending on the country).
3. Estimation of the impact of AEs on quality of life and costs
Some guidance has been developed on how to estimate the impact of AEs on quality of life. This is most often done using utility decrements that subtract a fixed amount of utility from each relevant health state. However, this subtraction of AE impacts on quality of life should only occur if the utility impact is not already reflected in the health state’s quality of life. Furthermore, to estimate the utility decrement of AEs, the authors write:
To estimate the decrease in utility corresponding to an EA, Pullenayegum and others. Conditions were shown under which methods such as Tobit or censored least absolute deviation models should be considered with caution. Alternative regression approaches (e.g. linear mean utility models) should be used instead.
Another challenge is the frequency with which AEs occur. Often, clinical trials report the proportion of patients who have experienced a given AE, but not how frequently that AE occurs (or how long it lasts). In terms of costs, most models focus on hospitalization costs for acute and severe events (often defined as AE grade ≥3).
4. Uncertainty surrounding the impact of AEs on the ICER
Since adverse effects observed during clinical trials cannot be extrapolated to estimate the impact on long-term adverse effects or overall health outcomes, scenario analyses around adverse effects are rarely performed. However, in many cases, long-term adverse effects may be important. For example:
…[an] Economic evaluation related to smoking cessation (Keeney et al.) showed that excluding AEs associated with depression and self-harm affected the determination of the most cost-effective health strategies.
Despite these challenges, the authors offer some best practices to help researchers address them.
A very interesting article in its entirety. You can read the full article here.
