The cost-effectiveness of procalcitonin for guiding antibiotic prescribing in individuals hospitalized with COVID-19: part of the PEACH study

Abstract

Background: Many hospitals introduced procalcitonin (PCT) testing to help diagnose bacterial coinfection in individuals with COVID-19, and guide antibiotic decision-making during the COVID-19 pandemic in the UK.

Objectives: Evaluating cost-effectiveness of using PCT to guide antibiotic decisions in individuals hospitalized with COVID-19, as part of a wider research programme.

Methods: Retrospective individual-level data on patients hospitalized with COVID-19 were collected from 11 NHS acute hospital Trusts and Health Boards from England and Wales, which varied in their use of baseline PCT testing during the first COVID-19 pandemic wave. A matched analysis (part of a wider analysis reported elsewhere) created groups of patients whose PCT was/was not tested at baseline. A model was created with combined decision tree/Markov phases, parameterized with quality-of-life/unit cost estimates from the literature, and used to estimate costs and quality-adjusted life years (QALYs). Cost-effectiveness was judged at a £20 000/QALY threshold. Uncertainty was characterized using bootstrapping.

Results: People who had baseline PCT testing had shorter general ward/ICU stays and spent less time on antibiotics, though with overlap between the groups' 95% CIs. Those with baseline PCT testing accrued more QALYs (8.76 versus 8.62) and lower costs (£9830 versus £10 700). The point estimate was baseline PCT testing being dominant over no baseline testing, though with uncertainty: the probability of cost-effectiveness was 0.579 with a 1 year horizon and 0.872 with a lifetime horizon.

Conclusions: Using PCT to guide antibiotic therapy in individuals hospitalized with COVID-19 is more likely to be cost-effective than not, albeit with uncertainty.

Figure 1.

Economic evaluation model. Patients who did and did not receive a PCT test followed the same pathway. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 2.

Weighted histograms of ward/ICU stays and antibiotic days for patients receiving a PCT test at baseline, and associated QALY losses and costs. Crosses indicate the five highest values. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 3.

Weighted histograms of ward/ICU stays and antibiotic days for patients not receiving a PCT test at baseline, and associated QALY losses and costs. Crosses indicate the five highest values. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 4.

Weighted histograms of total cost and QALYs. Crosses indicate the five highest values. (a) Participants with PCT tested at baseline. (b) Participants without PCT tested at baseline. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 5.

Cost-effectiveness planes. The x-axis and y-axis show, respectively, the QALY and cost differences between patients given and not given PCT tests at baseline with a 1 year (left) and lifetime (right) horizon. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 6.

Cost-effectiveness acceptability curve. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.