An Economic Model to Compare Linezolid and Vancomycin for the Treatment of Confirmed Methicillin-Resistant Staphylococcus aureus Nosocomial Pneumonia in Germany

Abstract

Objectives: To evaluate economic impact of linezolid (LZD) versus vancomycin (VAN) for treatment of confirmed methicillin-resistant Staphylococcus aureus(MRSA) nosocomial pneumonia (NP) in German health care system. MethOds: A 4 week decision model was developed capturing 1 st and 2 nd line therapy. Published literature (primarily Wunderink 2012, reported 54.8% clinical efficacy for LZD vs. 44.9% for VAN in modified Intent-to-Treat population at End of Study for treatment of MRSA NP) and expert opinion provided clinical and resource use data, such as efficacy, mortality, adverse events (AEs), treatment duration, and length of hospital/ICU stay. German cost data was obtained from published literature. Base-case analysis used 10-day treatment duration. In event of treatment failure/ severe AEs on 1st-line therapy, drug was switched after 7 days. Costs were reported in 2012 Euros. Results: LZD was associated with minimally lower costs (€ 16,119 vs. € 16,144), and greater overall treatment success compared to VAN, resulting in LZD 'dominating' VAN. About 80% of treatment costs were related to hospital stay, primarily ICU (72%). Drug therapy, physician visits, laboratory tests and AEs/ treatment failure each account for ≤ 5% of total costs. Several scenarios were tested by varying treatment duration (7 or 14 days), and varying discontinuation/ switch of therapy (at 5 or 10 days). In cases of shorter treatment duration (7 day) or delayed switch of therapy (at 10 day), linezolid continues to 'dominate'. However, with 14 day treatment duration or early switch (at 5 day) linezolid becomes more costly, but still has greater effectiveness resulting in a relatively low Incremental Cost Effectiveness Ratio (ICER) of € 8,207 and € 2,343 per successfully treated patient. cOnclusiOns: LZD is a cost-effective alternative to VAN for treatment of MRSA-confirmed NP, owing primarily to its higher clinical response rate. Future analyses should use other country costs/resource use data to test result generalizability and assess empiric treatment phase.

Review of health economic models exploring and evaluating treatment and management of hospital-acquired pneumonia and ventilator-associated pneumonia

“…All but two [38,39] of the nine studies considered MRSA HAP/VAP: four found linezolid to be cost-effective [33,40,43,46]; and three found linezolid to be less costly and more effective [25,34,45]. These conclusions for treating MRSA HAP accord with those reached by Zhang et al [49], who include vancomycin and linezolid amongst a number of comparators.…”

“…Linezolid and vancomycin are only active against Gram-positive infections and five studies focused solely on cases where these organisms were confirmed [25,34,43,45,46]; four studies drew on clinical trials where patients received aztreonam for Gram-negative coverage, but either did not consider its costs in their model [38][39][40] or excluded patients with Gram-negative infections [33].…”

“…Only one was published before 2000 [32]; seven [24, [33][34][35][36][37][38] were published between 2001 and 2006, and twelve [25,[39][40][41][42][43][44][45][46][47][48][49] between 2009 and 2019. The USA was the most-represented country (n=13) [24, 25, 32, 35-39, 42, 44, 47-49], with two studies in Germany [40,45], and one in each of: Brazil [34]; China [46]; Spain [33]; Taiwan [43] and the UK [41]. [33,38,39,41,47]; and life years gained [33,38,40,47,49].…”

“…A range of outcomes were considered across studies, and some studies used multiple outcomes. The most commonly used outcomes included: survival[34-36, 38, 40, 42, 47]; clinical cure rate[24, 25,40,43,45,46,48]; QALYs…”

“…Six studies considered HAP/VAP broadly[24,33,36,39,42,44], whereas the remaining studies focused on HAP and/or VAP caused by specific pathogens, especially Methicillin-resistant Staphylococcus aureus (MRSA). Six only considered VAP[24,35,36,43,45,47].The most common interventions evaluated (n=15) were simple comparisons between pairs of antibiotics. In seven studies, antibiotics were used from the empiric treatment phase[34,39,40,43,44,46,47]; three after the empiric phase [25,41,45]; for five it was unclear[33,35,38,42,48].…”