Previously, a large number of economic models have been developed to explore to cost-effectiveness of plaque psoriasis treatments in the UK, including those developed to inform submissions to NICE [11,12,13,14,15,16,17,18]. In general, a consistent structure has been developed for these submissions, namely, a Markov model that aimed to quantify the cost and health implications of interventions by tracking movement between health states defined by psoriasis area and severity index (PASI) score. Additionally, these models generally explored the impact of incorporating interventions as part of treatment sequences. An Excel-based economic model from the perspective of the UK NHS was developed for the purpose of this study. However, given the aim of the study was to explore the cost implications of drug delivery/administration options only, a simple cost analysis model was developed as a more complex structure was judged to be unnecessary. Therefore, no consideration was given to health outcomes or the impact of sequences, and a Markov structure was not required. Additionally, the wider cost implications, such as the impact of each intervention in the context of non-delivery/administration costs, were not considered. In particular, the analysis did not calculate the impact of the cost of the interventions themselves. The intervention costs were excluded from the analysis because confidential patient access scheme (PAS) discounts have been agreed for a large number of the interventions included in the analysis (e.g. secukinumab, brodalumab, ixekizumab). As these discounts are confidential, it is not possible to know the true price of these interventions in the UK currently.

The total drug delivery costs were estimated over a 2-year time horizon for one hypothetical patient receiving each included intervention. The model started at the point at which the intervention is first administered, and therefore, the induction period for each drug was captured. To simplify the analysis, it was assumed that the patient remained on the intervention for the full 2-year period. In reality, a proportion of patients will discontinue from biological therapies each year, in particular if the intervention does not achieve an adequate response after the initial induction period. However, if a patient does achieve an initial response, these biologic interventions are generally well tolerated, and therefore, patients are likely to remain on the treatment for an extended period of time. Therefore, it is plausible for someone to receive a biologic intervention for a 2-year period.

Two separate scenarios were considered in the analysis: administration of the intervention in hospitals and home delivery paid by the NHS. Further details are provided below.

A key input within the model was the frequency of administration for each intervention, as this determines how often drug delivery is required (i.e. a greater frequency of administration was expected to lead to higher delivery costs). The frequency of administration for each intervention was sourced from the British National Formulary (BNF) [19].

In terms of home delivery, it was necessary to make a small number of assumptions in order to estimate the total cost of this option. First, it was assumed that a nurse would travel to a patient’s home to administer the first dose of the intervention, and this would, on average, require a total of 1 h of nurse time (including travel). The total cost of such a nurse visit was estimated to be £140.Footnote 1 Second, it was assumed that all patients would self-administer the drug after the first dose and, therefore, subsequent nurse time would not be required. The unit cost of delivering the drug without a nurse was estimated to be £65, to account for the transfer of the intervention to the patient’s home (see Footnote 1). Therefore, the total cost of the drug delivery was estimated to be £140 for the first dose and £65 for each subsequent dose (i.e. when the nurse is no longer required). The total cost of home delivery for each intervention was, thus, calculated by multiplying the dosing frequency over 2 years with the delivery costs just outlined.

For the administration in hospital delivery method, it was assumed a patient would travel to hospital and be given the intervention by a nurse during a 20-min outpatient appointment. The unit cost of this activity was sourced from the Personal Social Services Research Unit (PSSRU) and equated to a value of £47 for 1 h with a band 6 nurse [20]. This unit cost was multiplied by the time required for administration, giving a total cost of £15.51 per administration in hospital. Additionally, the NHS is required to pay value-added tax (VAT) on all drugs that are administered within a hospital setting (VAT is not incurred for home deliveries). Therefore, these additional costs were also incorporated in the analysis. The VAT rate in the UK in 2020 was 20%, and this was applied to the relevant drug prices for each intervention to estimate the total VAT costs associated with drug administration. As noted previously, PAS discounts are in place for a number of interventions, but as these are confidential, it was necessary to use the list price for all interventions, as reported in the BNF [19].

As with home delivery, the total cost of drug administration for each intervention in the hospital was determined by the dosing frequency for each intervention. All inputs are detailed in Table 1. All costs were based on the 2019/2020 cost year, where possible. An illustration of the model is provided in Fig. 1.

An overview of the steps undertaken to estimate the total cost of the two options for biologic drug delivery (i.e. delivery at a patient’s home or administration within a hospital setting)

All data in the analysis are based on previously conducted studies, and the analysis does not contain any new studies with human participants or animals performed by any of the authors.