Potential health and economic impacts of dexamethasone treatment for patients with COVID-19

Abstract

Dexamethasone can reduce mortality in hospitalised COVID-19 patients needing oxygen and ventilation by 18% and 36%, respectively. Here, we estimate the potential number of lives saved and life years gained if this treatment were to be rolled out in the UK and globally, as well as the cost-effectiveness of implementing this intervention. Assuming SARS-CoV-2 exposure levels of 5% to 15%, we estimate that, for the UK, approximately 12,000 (4,250 - 27,000) lives could be saved between July and December 2020. Assuming that dexamethasone has a similar effect size in settings where access to oxygen therapies is limited, this would translate into approximately 650,000 (240,000 - 1,400,000) lives saved globally over the same time period. If dexamethasone acts differently in these settings, the impact could be less than half of this value. To estimate the full potential of dexamethasone in the global fight against COVID-19, it is essential to perform clinical research in settings with limited access to oxygen and/or ventilators, for example in low- and middle-income countries.

Fig. 1. Patient pathway diagram for the calculation of COVID-19 mortality for a given exposure level.

We assign two categories of patients who are eligible for dexamethasone treatment. Those who require non-invasive oxygen (but never require ventilation) and those who ultimately require mechanical ventilation (where we assume their use of non-invasive oxygen during this pathway is negligible in terms of time and cost). Each group has two survival rates associated with them, depending on whether they receive dexamethasone treatment or not. A percentage of patients will not progress to requiring ventilation if they receive dexamethasone treatment early, and this is reflected in the “Oxygen”:“Ventilator” ratio of the “Dexamethasone” branch of the pathway.

Fig. 2. Expected impact of dexamethasone in the UK from July to December 2020 for the range of scenarios explored.

The value and range quoted for each outcome represent the median and 5th and 95th percentiles of the sensitivity analysis outcomes, i.e., 90% double-sided confidence intervals. The letter m represents million.

Fig. 3. Uncertainty regarding access to oxygen therapy.

a A cost-effectiveness acceptability curve constructed to quantify and graphically represent uncertainty in the economic evaluation. The x-axis represents ʎ, the cost-effectiveness threshold, while the y-axis reflects the estimated probability of dexamethasone being cost-effective. We highlighted the probability of dexamethasone being cost effective at the ICER threshold value corresponding to the upper end of the 90% confidence interval for the ICER obtained in the sensitivity analysis. The shaded areas are defined by the ICER threshold values corresponding to the mean and 90% confidence intervals obtained in the probabilistic sensitivity analysis: red, < 5% of the distribution area; orange, 5 –50%; green, 50–95%; blue, 95–100%. b Sensitivity of the evaluation of dexamethasone’s cost effectiveness to assumptions in access to oxygen therapy and efficacy of dexamethasone on people that do not receive oxygen. Boxplots were drawn to represent the distribution of 1 million predicted incremental cost per life-years gained values for each combination of dexamethasone efficacy when not receiving oxygen and the probability of receiving oxygen treatment if required. The middle line is the median, the lower and upper hinges correspond to the first and third quartiles, and the whiskers extend to the 5th and 95th percentiles.