Cost-effectiveness of treatment strategies for populations from strongyloidiasis high-risk areas globally who will initiate corticosteroid treatment in the USA

Abstract

Abstract Background The risk of developing strongyloidiasis hyperinfection syndrome appears to be elevated among individuals who initiate corticosteroid treatment. Presumptive treatment or treatment after screening for populations from Strongyloides stercoralis-endemic areas has been suggested before initiating corticosteroids. However, potential clinical and economic impacts of preventative strategies have not been evaluated. Methods Using a decision tree model for a hypothetical cohort of 1000 individuals from S. stercoralis-endemic areas globally initiating corticosteroid treatment, we evaluated the clinical and economic impacts of two interventions, ‘Screen and Treat’ (i.e. screening and ivermectin treatment after a positive test), and ‘Presumptively Treat’, compared to current practice (i.e. ‘No Intervention’). We evaluated the cost-effectiveness (net cost per death averted) of each strategy using broad ranges of pre-intervention prevalence and hospitalization rates for chronic strongyloidiasis patients initiating corticosteroid treatment. Results For the baseline parameter estimates, ‘Presumptively Treat’ was cost-effective (i.e. clinically superior with cost per death averted less than a threshold of $10.6 million per life) compared to ‘No Intervention’ ($532 000 per death averted) or ‘Screen and Treat’ ($39 000 per death averted). The two parameters contributing the most uncertainty to the analysis were the hospitalization rate for individuals with chronic strongyloidiasis who initiate corticosteroids (baseline 0.166%) and prevalence of chronic strongyloidiasis (baseline 17.3%) according to a series of one-way sensitivity analyses. For hospitalization rates ≥0.022%, ‘Presumptively Treat’ would remain cost-effective. Similarly, ‘Presumptively Treat’ remained preferred at prevalence rates of ≥4%; ‘Screen and Treat’ was preferred for prevalence between 2 and 4% and ‘No Intervention’ was preferred for prevalence <2%. Conclusions The findings support decision-making for interventions for populations from S. stercoralis-endemic areas before initiating corticosteroid treatment. Although some input parameters are highly uncertain and prevalence varies across endemic countries, ‘Presumptively Treat’ would likely be preferred across a range for many populations, given plausible parameters.