This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors. As such, ethics review and approval was not required.

A decision-analytic model (Fig. 1) was developed in Microsoft Excel to assess the budget impact of VOS from a US health plan perspective over a 1-year time horizon. The model first estimates the number of patients experiencing an rCDI event within a hypothetical US health plan. Patients with rCDI entered the model and were treated with SoC with or without microbiome treatment. Following treatment for the initial rCDI event, patients could either recover from the recurrence or experience a subsequent recurrence (Fig. 2). The probability of recurrence was dependent on patient characteristics and the treatment received. Each recurrence resulted in costs of treatment (pharmacy costs plus other medical costs related to hospitalization, outpatient visits, etc.).

Model structure. Figure is for illustrative purposes only. rCDI, recurrent Clostridioides difficile infection; SoC, standard of care; VOS, fecal microbiota spores, live-brpk (formerly SER-109)

Recurrence episodes. CDI, Clostridioides difficile infection; rCDI, recurrent CDI

The model compared two market mixes in the base-case analysis: (1) a market mix in which patients are treated with SoC alone for all recurrences (the status quo market mix), and (2) a market mix in which VOS use (following SoC) increases in patients with an increasing number of recurrences (anticipated market mix). Total annual and per-member per-month (PMPM) costs and outcomes were reported for each market mix. The differences in total annual and PMPM costs, and outcomes between the two market mixes, were calculated to estimate the budget impact of VOS to the health plan.

The perspective of the analysis is that of a hypothetical third-party payer in the US (e.g., Centers for Medicare & Medicaid Services, a managed care organization, or a self-insured employer) of 1 million members. The base-case assumed a commercial health plan setting; a scenario analysis was conducted assuming a Medicare setting. Patients modeled in the analysis were adults (aged ≥ 18 years) with rCDI. Adults were assumed to comprise 78% of the health plan population, and the number of individuals with rCDI was estimated on the basis of published epidemiology data for risk of experiencing an initial CDI (0.13%) [16, 17] and risk of first recurrence (22.04%) [2, 18] (Table 1). A total of 225 patients with rCDI were estimated in a commercial health plan (1533 patients in a Medicare plan).

The following treatment options were considered in the analysis: SoC alone, SoC followed by VOS, and SoC followed by RBL. For SoC, a distribution of the following antibacterial treatments was assumed: fidaxomicin (200 mg), metronidazole (500 mg), vancomycin (125 mg), and vancomycin pulse–taper. The distribution of antibacterial use was assumed to be dependent on rCDI (Supplementary Material Table S1).

The input parameter estimates used in the model were obtained from available published literature (Table 1) and are described in the following subsections.

The baseline risks of recurrence following an initial rCDI event were estimated from published literature [2, 3]. For SoC alone, no change was assumed from the baseline risk of recurrence. For VOS, a relative risk reduction of 0.69 was assumed on the basis of the VOS phase 3 clinical trial [13]. A scenario analysis in which a proportion of the patients received RBL instead of VOS was also included. For the scenario analysis including RBL, a relative risk reduction with RBL of 0.31 based on the RBL phase 3 trial was assumed [14].

With each recurrence, patients were also assumed to incur an increased risk of mortality. The relative risk of mortality was estimated on the basis of a retrospective cohort study in a US healthcare setting [19]. This risk was then applied to the general population mortality risk [20] to estimate the mortality risk for each recurrence.

Antibiotic drug costs and microbiota treatment costs (in 2023 US dollars) were obtained based on wholesale acquisition costs from IBM Micromedex Red Book [21]. Costs for each recurrence were then estimated on the basis of the assumed distribution of antibiotic treatments (Supplementary Material Table S1) and microbiota treatments (Table 1) for that recurrence. The nondrug cost of recurrence was estimated on the basis of resource utilization data (e.g., diagnostic tests, office visits, CDI-related hospitalizations, and CDI-related surgeries) from a study of patients with rCDI in a US hospital setting [6]. Unit costs for these resources were obtained from published literature [22,23,24,25]. Where necessary, costs were inflated to 2023 US dollars using the medical consumer price index [26].

The percentage use of SoC alone, SoC plus VOS, and SoC plus RBL (referred to hereafter as market share) for each recurrence was assumed for each market mix (Table 2).

For the status quo market mix, 100% use of SoC alone was assumed for all recurrences. For the anticipated market mix, a gradual increase was assumed in uptake of VOS following SoC for each subsequent recurrence, ranging from 10% in recurrence 1 up to 40% for recurrences 4+ . Two additional uptake scenario analyses were considered in which the use of VOS varied: a more aggressive uptake scenario in which VOS uptake was increased by 50% (market shares ranging from 15% in recurrence 1 up to 60% in recurrences 4+) and a hypothetical scenario in which VOS was used in 100% of recurrences. Finally, a hypothetical scenario was considered in which 100% were treated with SoC plus RBL.

The efficacy of VOS and RBL (relative risk reduction) were assumed to be the same regardless of patient population (commercial or Medicare) and number of recurrence (e.g., first, second, third, etc.). Costs of SoC were assumed to be the same regardless of whether patients were treated with SoC alone or with VOS or RBL. The most common adverse events occurring in ≥ 5% of patients were abdominal pain, abdominal distension, diarrhea, flatulence, nausea, urinary tract infection, and fatigue [12]. These events were assumed not to incur any healthcare costs and were excluded from the model.

The base-case analysis was performed for a hypothetical commercial health plan setting. In addition to this base-case analysis, several scenario analyses were conducted. Specifically, the market uptake scenarios were performed as outlined above. Finally, a scenario with a Medicare health plan perspective, assuming the base-case market mix comparison, was considered. In addition, to test the impact of individual parameter uncertainty on results, a deterministic one-way sensitivity analysis in which parameter estimates were varied by ± 20% was conducted.