Modeling the economic impact of linezolid versus vancomycin in confirmed nosocomial pneumonia caused by methicillin-resistant Staphylococcus aureus

Table 1 Model inputs on clinical and resource use data ^a

	Linezolid base case value (range^b)	Vancomycin base case value (range^b)	Distribution^b	Source
Efficacy and safety end points^c,%
Efficacy (in survivors)	54.8 (49.8^d to 66.7)	44.9 (35.5 to 52.9)	β	[16, 20]
Mortality^e	27.2	27.2	–
SAEs leading to discontinuation^f	1.8 (0^d to 5.2)	3.1 (0^d to 6.5)	β	[14, 16]
Failure leading to discontinuation^g	16.2	24.8	–
Resource use
Total days in hospital^h	17.9 (13.9^d to 18.8)	18.6 (14.6^d to 20.1)	γ	[20, 21]
Days of treatmentⁱ	10.0 (7 to 14)	10.0 (7 to 14)	Uniform	[16]
Days in ICU^h	10.1 (6.1^d to 12.2)	10.6 (6.6^d to 16.2)	γ	[20, 21]
Days on mechanical ventilation^h	8.3 (4.3^d to 10.4)	8.1 (4.1^d to 14.3)	γ	[20, 21]
Additional days in hospital due to SAE	1.7 (0 to 5)^d	1.7 (0 to 5)^d	γ	[21]
Additional days in hospital due to treatment failure	2 (0 to 5)^d	2 (0 to 5)^d	Uniform	Expert input
Number of days until switch to second-line after treatment failure/SAE with first- line	7 (5 to 10)^d	7 (5 to 10)^d	Uniform	Expert input
Days receiving IV antibiotic^h	10.0	10.0	–	[16]
Antibiotic IV doses/day^h	2.0	2.0	–	Product label
Physician/attending/intensivist visit (inpatient)/dayⁱ	1.0	1.0	–	Expert input
Lab work/wk^i,j	7.0	8.0	–	Expert input

^aIV, Intravenous; SAE, Serious adverse event. Clinical response rate for modified intent-to-treat population at end-of-study time point was used [16, 21, 23]. ^bRanges and distributions are provided for variables that were used in sensitivity analyses. ^cSame clinical data were used for second-line treatment. ^dThis was an arbitrary assumption that was validated with expert opinion. ^eWeighted average, because model assumes equal mortality due to lack of significant mortality difference (linezolid = 63 of 224; vancomycin = 59 of 224). ^fLinezolid = 4 of 224; vancomycin = 7 of 224. ^gBecause this is a decision tree model, this probability was derived as [1 – (probability of efficacy + probability mortality + probability of SAEs leading to discontinuation)]. ^hData input for first line treatment only. ⁱData input for first and second line treatment. ^jDaily serum creatinine levels and complete blood count for both antibiotics and once-weekly serum vancomycin levels for vancomycin.

ISSN: 1364-8535