PIN63 A Cost-Minimization Model to Evaluate the Impact of Contact information: Wajeeha Ansari for the Treatment of Complicated Skin and Soft Tissue Infections Pfizer Inc., 235 E 42nd Street, in Hospitalized Adults in Spain New York, NY10017, USA Alex Soriano1, Santiago Grau2, Simone Rivolo3, Edit Remak4, Carmen Peral5, Michal Kantecki6, Wajeeha Ansari7, [email protected] Claudie Charbonneau6, Jennifer Hammond8, Mark Wilcox9 1Hospital Clínic de Barcelona, Barcelona, Spain; 2Hospital del Mar Universitat Autònoma de Barcelona, Barcelona, Spain; 3Evidera, London, UK; 4Evidera, Budapest, Hungary; 5Pfizer, Madrid, Spain; 6Pfizer, Paris, France;7 Pfizer, New York, NY, USA;8 Pfizer, Collegeville, PA, USA;9 University of Leeds, Leeds, UK.

Introduction Clinical cure rates for ceftaroline fosamil were sourced from an integrated analysis Results • of the CANVAS 1 and 2 Phase III, randomized controlled trials.6 Complicated skin and soft-tissue infections (cSSTI) are associated with considerable Ceftaroline fosamil demonstrated similar costs and percentage of early discharged • In the absence of head-to-head data, the clinical cure rates for other treatments • morbidity and high healthcare resource use and costs, with most of the economic • patients (PDE) compared with the comparators investigated, as shown in Figure 2. burden driven by prolonged hospital stay.1–3 were indirectly estimated using results from a network meta-analysis (NMA) in the clinically evaluable (CE) population7 or sourced from the literature if a comparator Most patients admitted to hospital with cSSTI are treated empirically, with the 8 • was not included in the NMA. Figure 2. Total costs per patient and percentage of early discharges predicted by selection being guided by disease severity, local patterns of bacterial the model (based on 1,000 simulations) Day 3 clinical response (sourced from the literature) was defined as:9 resistance, the frequency of cSSTI-associated pathogens and individual antibiotic 8,000 35 4 • safety profiles. –– Cessation of infection spread (no increase in baseline lesion width or length 7,500 30

Initial antibiotic treatment failure has been shown to prolong the duration of hospital measurement). ges (%) • 5 7,000 stays, increase antibiotic usage and add to hospital expenditure. –– Absence of fever (temperature ≤37.6°C). 25 6,500 Therefore, the choice of initial antibiotic therapy is pivotal to achieve an early It is important to note that cloxacillin early response criteria utilized in the literature • response and possibly early discharge, thus reducing hospital expenditure. 20 • were significantly different to those for ceftaroline fosamil, thus hampering the 6,000 15 Objective possibility of direct comparison. However, threshold analysis was performed. 5,500 Costs and resource use inputs 5,000 10 To evaluate the use of ceftaroline fosamil (600 mg q12h) as an alternative to other otal costs per patient ( € ) T • antibiotic therapies ( 1 g q12h, linezolid 600 mg q12h, Healthcare resource use and costs used in the base-case analysis are reported 4,500 centage of early discha r 5

• Pe r 4–10 mg/kg q24h, cloxacillin 1–2 g q4–6h and tedizolid 200 mg q24h) for the empiric in Table 2 and Table 3. Following clinical specialists’ consultation, 4,000 0 treatment of hospitalized patients with cSSTI, from the perspective of the Spanish clavulanate (vial formulation) was chosen as second-line antibiotic, while amoxicillin Ceftaroline Vancomycin Linezolid Daptomycin Tedizolid Ceftaroline Vancomycin Linezolid Daptomycin Tedizolid fosamil 1g 600 mg 4–10 mg/kg 200 mg fosamil 1g 600 mg 4–10 mg/kg 200 mg National Health System. clavulanate (oral formulation) was chosen as oral step-down antibiotic. 600 mg q12h q12h q24h q24h 600 mg q12h q12h q24h q24h q12h q12h Following clinical specialists’ consultation, it is assumed that all comparators except Methods • for ceftaroline fosamil may require pairing with a Gram-negative antibiotic, and this Model description Gram-negative antibiotic would be used in 33% of patients.2 The Gram-negative • Specifically, the 95% confidence intervals generated from the probabilistic sensitivity A cost-minimization decision-tree model was developed in Microsoft Excel 2016® antibiotic used in the model is 1 g q8h. analyses overlap (1,000 simulations) suggested similar total costs per patient and • (Microsoft Corporation, Redmond, WA, USA). The model simulates a hypothetical similar PDE with ceftaroline fosamil as with the other considered in cohort of adult patients admitted to hospital with cSSTI that require empiric Daptomycin 4 mg/kg q24h and cloxacillin 1 g q6h dosages were used for the base- the analysis. • case, given that these were the dosages used in the clinical trials from which the intravenous (IV) antibiotic therapy. Ceftaroline fosamil resulted in average PDE of 24.6% (CI 19.49–30.2%) with clinical inputs were derived and on which the NMA was based. • Each branch of the decision tree represents a clinical pathway (Figure 1). average total cost per patient of €6,763 (CI €6,268–€7,219). • –– However, additional scenarios with daptomycin 4–10 mg/kg q24h and cloxacillin –– The square in the schematic of the model represents a decision node that indicates Vancomycin, linezolid, daptomycin and tedizolid resulted in average PDE of 22% where the choice to use ceftaroline fosamil or other comparators is made. 1–2 g q4–6h were simulated (without varying the clinical inputs) as, following • (17.3–27.1%), 26.4% (20.5–32.3%), 28.6% (22.1–35.8%) and 26.5% (20.4–33.3%), clinical experts’ consultation, these are the dosages most commonly used in –– The circles represent chance nodes, at which probabilities generated by the respectively, for a total cost per patient of €6,619 (€5,902–€6,929), €6,394 model determine the pathway that a trial simulation will travel. clinical practice. (€5,881–€6,904), €6,855 (€5,800–€7,410) and €7,173 (€6,608–€7,763), respectively. –– The triangles represent terminal nodes, at which the total direct costs are • The costs of the drugs and of the hospital resources (Table 3) were obtained from Threshold analysis indicated that total cost neutrality with respect to ceftaroline determined for that specific branch of the model. the Spanish National Price List. • fosamil was achieved with cloxacillin, with an early response rate of 50.2–53.3% (depending on dosage). Figure 1. Schematic visualization of the decision-tree model The model demonstrated that antibiotic costs represent a small proportion of overall • costs (5–15% of total costs), the cost associated with a hospital stay had the highest Patients considered unsuitable for early hospital discharge because impact (85–95% of total costs). of factors unrelated to cSSTI Pathway A: these patients remain in hospital to complete their Day 3 clinical initial IV antibiotic course (6.6 days based on8) and are discharged Finally, the deterministic sensitivity analyses demonstrated that early response to, response YES 5 Patients considered eligible for switching to a 5-day course of from hospital after an average stay of 13.4 days • and duration of, antibiotic treatment are the key drivers of total costs, as shown in oral antibiotics and discharged from hospital early Treat with Pathway B: Figure 3 for ceftaroline fosamil with respect to vancomycin. ceftaroline fosamil these patients are discharged from hospital on Day 4 Similar results were obtained with ceftaroline fosamil versus other comparators Patients switched to a second-line antibiotic (7 days) assumed to achieve • (data not shown). 100% cure rate Pathway C: these patients receiving a second-line antibiotic are Adult patients Day 3 clinical discharged from hospital after an average stay of 20.3 days5 admitted to hospital response NO with cSSTI Patients showing signs of improvement Figure 3. Tornado diagram showing, in decreasing order, the key parameters Continue and complete ceftaroline fosamil therapy Pathway D: delayed responders remain on their initial antibiotic Day 3 clinical therapy (6.6 days based on8) and are discharged from hospital that generate the most variation in the total incremental cost per patient, when response YES after an average stay of 13.4 days5 comparing ceftaroline fosamil with vancomycin

Treat with Total cost difference per patient between ceftaroline fosamil and vancomycin for the treatment of cSSTI (€) comparator Same pathways as for ceftaroline fosamil –1.57 48.43 98.43 148.43 198.43 248.43 298.43 Day 3 clinical response NO Duration of ceftaroline fosamil (days) for responders not switched to an oral antibiotic Over each branch, a brief description of the corresponding patient pathway has been summarized. Day 3 clinical response vancomycin Day 3 clinical response ceftaroline fosamil q12h Cost per day of tests/monitoring for vancomycin This model structure was considered the most appropriate, as costs and clinical Table 2. Summary of resource use along with SE and distribution used for base- • Percentage of Day 3 responders switched to outcomes (namely, resolution of signs and symptoms of the infection) occur over a case and sensitivity analysis oral antibiotics 1 relatively short period of time, typically <3 weeks from hospital admission to discharge. Cost of a hospital day stay on a medical ward Base- Source Description SE Distribution Model assumptions case (reference) Risk ratio vancomycin Hospital length of stay: Initial antibiotic In building the decision-tree model, several simplifying assumptions were required Mean hospital LOS for initial antibiotic 13.4 0.03 Lognormal 5 treatment success • to generalize the clinical pathways and allow the integration of data from multiple success (days) Hospital length of stay: Initial antibiotic treatment failure sources. In Figure 1, the modelling assumptions for each patient pathway have Mean hospital LOS for initial antibiotic 20.3 0.06 Lognormal 5 Clinical cure rate ceftaroline fosamil q12h been summarized. failure (days)

Costs associated with microbiological tests performed to identify the pathogens Proportion of patients eligible for 0.333 0.045 Beta 2 Result obtained using the parameter low value • causing the infection were the same for both treatment arms and were therefore not IV-to-oral therapy Result obtained using the parameter high value considered in the model. Percentage of patients receiving Gram- 0.333 2 negative antibiotic (all comparators • The financial consequences associated with the management of treatment-related except ceftaroline fosamil) Further analyses are needed to address the large heterogeneity in early response adverse events are not incorporated within the cost-minimization model. Biochemistry tests per day 0.37 12 • criteria found in the literature. (all comparators except Model inputs ceftaroline fosamil) Clinical inputs Haemogram tests per day 0.46 12 The clinical inputs used in the cost-minimization analysis are summarized in Table 1. (all comparators except Conclusions • ceftaroline fosamil) C-reactive protein tests per day 0.00 12 • The model results demonstrated that early response to, and duration of, Table 1. Summary of clinical inputs along with standard error (SE) and distribution (all comparators) antibiotic treatment are the key drivers of total costs for managing cSSTI. used for base-case and sensitivity analysis Therapeutic drug monitoring test 0.00 12 per day (all comparators except Given its clinical and safety profile, ceftaroline fosamil is an Base- Source (reference)/ • Description SE Distribution vancomycin) case assumption acceptable choice for cSSTI empiric therapy, providing costs and Therapeutic drug monitoring test 0.34 12 percentage of patients discharged early that are similar to other per day (vancomycin) Probability of clinical cure 0.916 0.0112 Beta 6 relevant antibiotic options. with ceftaroline fosamil Duration of antibiotic treatment (days) 6.6 1 Normal 13 Relative risk of ceftaroline 1.0 0.02 Lognormal 7 Duration of second-line antibiotic 7.0 1 Normal Assumption fosamil versus vancomycin treatment (days) plus Gram-negative Disclosures antibiotic LOS, length of stay. This study was funded by Pfizer. S.R and E.R. are employees of Evidera, who were paid Relative risk of ceftaroline 0.94 0.02 Lognormal 7 consultants to Pfizer in connection with the study and the development of this poster. C.P., M.K., fosamil versus linezolid W.A., C.C. and J.H. are employees of and shareholders in Pfizer. A.S. is an employee of Hospital with/without Gram-negative Table 3. Summary of cost inputs sourced from the Spanish Drug Official List coverage Clínic of Barcelona, S.G. is an employee of Universitat Autònoma de Barcelona, and M.W. is an Relative risk of ceftaroline 0.94 0.05 Lognormal 7 Description Base-case (2017 €) employee of The University of Leeds, each of which received research funding from Pfizer. fosamil versus daptomycin Relative risk of ceftaroline 1.04 0.12 Lognormal 7 Hospital stay (per day) 553.90 References fosamil versus cloxacillin Drug acquisition costs (unit cost/day cost) Relative risk of ceftaroline 1.05 0.02 Lognormal Relative risk chosen to 1. Garau J, et al. Clin Microbiol Infect. 2013;19:E377–E385. 10 Ceftaroline fosamil 600 mg q12h 60.00/120.00 fosamil versus tedizolid achieve 87% cure rate 2. Garau J, et al. BMC Infect Dis. 2015;15:78. Day 3 clinical response 0.740 0.0219 Beta 9 Vancomycin 1 g q12h 6.90/13.80 3. Ostermann H, et al. Poster P690. Presented at the 22nd European Congress of Clinical ceftaroline fosamil Linezolid 600 mg q12h 35.77/71.54 Microbiology and Infectious Diseases, London, UK. 31 March–3 April 2012. Day 3 clinical response 0.662 0.0237 Beta 9 Daptomycin 4 mg/kg q24h 100.00/100.00 vancomycin plus Gram- 4. Bassetti M, et al. Clin Microbiol Infect. 2014;20 Suppl 4:3–18. Cloxacillin 1 g q24h 1.11/1.11 negative antibiotic 5. Tarricone R, et al. J Med Econ. 2008;11:265–279. Tedizolid 200 mg q24h 198.67/198.67 Day 3 clinical response 0.794 0.0221 Beta 8 6. Corey GR, et al. Clin Infect Dis. 2010;51:641–650. linezolid with/without Gram-negative antibiotic aztreonam 1 g q8h 9.50/28.50 Gram-negative coverage 7. Rao N, et al. Poster PIN7. Presented at the ISPOR 19th Annual European Congress, Oral step-down antibiotic 0.44 per day Day 3 clinical response 0.860 0.0319 Beta 11 Vienna, Austria, 29 October–2 November 2016. daptomycin Second-line antibiotic 2.75 per day 8. Prokocimer P, et al. JAMA. 2013;309:559–569. Day 3 clinical response NA NA NA Different values tested Biochemistry test cost 2.05 9. Friedland HD, et al. Antimicrob Agents Chemother. 2012;56:2231–2236. cloxacillin Haemogram test cost 5.03 Day 3 clinical response 0.795 0.0222 Beta 8 10. Moran GJ, et al. Lancet Infect Dis. 2014;14:696–705. tedizolid C-reactive protein test cost 7.94 11. Kauf TL, et al. BMC Infect Dis. 2015;15:503. Therapeutic drug monitoring test cost 106.62 NA, not available. 12. Schürmann D, et al. Eur J Health Econ. 2009;10:65–79. 13. File TM Jr, et al. Clin Infect Dis. 2010;51:1395–1405.

Poster typesetting support provided by Mark Waterlow, BSc, of Prime, Knutsford, Cheshire, UK, funded by Pfizer, consisted solely of poster production and formatting with no contribution to editorial content. Presented at the International Society for Pharmacoeconomics and Outcomes Research Europe Congress, 10–14 November 2018, Barcelona, Spain.