Cost-effectiveness of procalcitonin-guided antibiotic use in community acquired pneumonia

Abstract

Background: Although prior randomized trials have demonstrated that procalcitonin-guided antibiotic therapy effectively reduces antibiotic use in patients with community-acquired pneumonia (CAP), uncertainties remain regarding use of procalcitonin protocols in practice.

Objective: To estimate the cost-effectiveness of procalcitonin protocols in CAP.

Design: Decision analysis using published observational and clinical trial data, with variation of all parameter values in sensitivity analyses.

Patients: Hypothetical patient cohorts who were hospitalized for CAP.

Interventions: Procalcitonin protocols vs. usual care.

Main measures: Costs and cost per quality adjusted life year gained.

Key results: When no differences in clinical outcomes were assumed, consistent with clinical trials and observational data, procalcitonin protocols cost $10-$54 more per patient than usual care in CAP patients. Under these assumptions, results were most sensitive to variations in: antibiotic cost, the likelihood that antibiotic therapy was initiated less frequently or over shorter durations, and the likelihood that physicians were nonadherent to procalcitonin protocols. Probabilistic sensitivity analyses, incorporating procalcitonin protocol-related changes in quality of life, found that protocol use was unlikely to be economically reasonable if physician protocol nonadherence was high, as observational study data suggest. However, procalcitonin protocols were favored if they decreased hospital length of stay.

Conclusions: Procalcitonin protocol use in hospitalized CAP patients, although promising, lacks physician nonadherence and resource use data in routine care settings, which are needed to evaluate its potential role in patient care.

Figure 1.

Procalcitonin protocol decision tree for low-risk pneumonia. Legend: The tree diagram depicts the decision to use or not use a procalcitonin protocol in low-risk hospitalized patients with community-acquired pneumonia (CAP). Ab = antibiotic, QOL = quality of life.

Figure 2.

Three-way sensitivity analysis. Graphs depict areas where strategies are favored in the base case analysis when no differences in patient outcomes or hospital length of stay are assumed. Points for parameters on the x- and y-axes that fall within shaded areas depict where procalcitonin protocol use is cost saving compared to usual care. Graphs in row a show results when procalcitonin levels are only drawn on admission in low-risk patients, row b depicts when procalcitonin levels are drawn on admission and every second day thereafter in low-risk patients, and row c shows procalcitonin protocol use in high-risk patients. Columns depict scenarios where physician nonadherence is held at base case levels (36.3 %) or at high levels (60 %). Rx = prescription, PCT = procalcitonin.

Figure 3.

Probabilistic sensitivity analysis. Cost-effectiveness acceptability curves show the likelihood that strategies will be favored over ranges of willingness to pay (or acceptability) thresholds. These analyses include the possibility of positive or negative impact on quality of life through procalcitonin protocol use. In row a, procalcitonin levels are only drawn on admission in low-risk patients; in b, procalcitonin levels are drawn on admission and every second day thereafter in low-risk patients; and c depicts procalcitonin protocol use in high-risk patients.