Chest ACCP Education Calendar
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Drummond, M. F.
Right arrow Articles by Sagnier, P.-P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Drummond, M. F.
Right arrow Articles by Sagnier, P.-P.
(Chest. 2003;124:526-535.)
© 2003 American College of Chest Physicians

An Economic Evaluation of Sequential IV/po Moxifloxacin Therapy Compared to IV/po Co-amoxiclav With or Without Clarithromycin in the Treatment of Community-Acquired Pneumonia*

Michael F. Drummond, DPhil; Debbie L. Becker, MSc; Margaret Hux, MSc; Jeremy V. M. Chancellor, MSc; Isabelle Duprat-Lomon, MD; Rolf Kubin, MD and Pierre-Philippe Sagnier, MD

* From Innovus Research (UK) Ltd (Dr. Drummond and Mr. Chancellor), High Wycombe, UK; Innovus Research Inc. (Ms. Becker and Ms. Hux), Burlington, ON, Canada; Bayer AG (Dr. Kubin), Wuppertal, Germany; Bayer Pharma (Dr. Duprat-Lomon), Puteaux, Paris, France; and Bayer plc (Dr. Sagnier), Stoke Poges, Berkshire, UK.

Correspondence to: Jeremy V. M. Chancellor, MSc, Managing Director, European Operations, Innovus Research (UK) Ltd, Suite 4, Leywood House, Denmark St, High Wycombe, Bucks, HP11 2ER, United Kingdom; e-mail: jchancellor{at}innovus.com


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objective: To evaluate costs, clinical consequences, and cost-effectiveness from a German and French health-care system perspective of sequential IV/po moxifloxacin monotherapy compared to co-amoxiclav with or without clarithromycin (AMC ± CLA) in patients with community-acquired pneumonia (CAP) who required parenteral treatment.

Methods: Costs and consequences over 21 days were evaluated based on clinical cure rates 5 to 7 days after treatment and health resource use reported for the TARGET multinational, prospective, randomized, open-label trial. This trial compared sequential IV/po monotherapy with moxifloxacin (400 mg qd) to IV/po co-amoxiclav (1.2 g IV/625 mg po tid) with or without clarithromycin (500 mg bid) for 7 to 14 days in hospitalized patients with CAP. Since no country-by-treatment interaction was found in spite of some country differences for length of hospital stays, resource data (antimicrobial treatment, hospitalization, and out-of-hospital care) from all centers were pooled and valued using German and French unit prices to estimate CAP-related cost to the German Sickness Funds and French public health-care sector, respectively.

Results: Compared to AMC ± CLA, treatment with moxifloxacin resulted in 5.3% more patients achieving clinical cure 5 to 7 days after therapy (95% confidence interval [CI], 1.2 to 11.8%), increased speed of response (1 day sooner for median time to first return to apyrexia, p = 0.008), and a reduction in hospital stay by 0.81 days (95% CI, - 0.01 to 1.63) within the 21-day time frame. Treatment with moxifloxacin resulted in savings of 266 and 381 for Germany and France respectively, primarily due to the shorter length of hospital stay. Cost-effectiveness acceptability curves show moxifloxacin has a >= 95% chance of being cost saving from French and German health-care perspectives, and higher probability of being cost-effective at acceptability thresholds up to 2,000 per additional patient cured.

Conclusion: IV/po monotherapy with moxifloxacin shows clinical benefits including increased speed of response and is cost-effective compared to IV/po AMC ± CLA in the treatment of CAP.

Key Words: community-acquired infections • costs and cost analysis • economics, pharmaceutical • pneumonia, bacterial • randomized controlled trials


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Community-acquired pneumonia (CAP), defined as pneumonia not acquired in a hospital or long-term care institution, is of significant importance to health-care systems worldwide due to the morbidity, mortality, and financial costs associated with the disease. Among those patients hospitalized, mortality ranges from 2 to 21% and rises to > 50% among patients with severe disease,1 making CAP the most common cause of death due to infectious disease.2 Since in as many as 30 to 60% of patients with CAP the responsible pathogen is not identifiable through diagnostic testing,1 an empiric approach to the treatment of CAP is usually required. The fluoroquinolone class of antibiotics has become established as an important therapeutic option in this indication.

Although the acquisition costs of fluoroquinolones are generally higher than the costs of traditional agents used in the treatment of CAP, anti-infectives represent only a small portion of the total costs of CAP treatment. The predominant cost-driver in the treatment of CAP is hospitalization.3 Formal techniques of economic evaluation, such as cost-effectiveness analysis (CEA), allow health-care decision makers to consider all relevant costs of alternative interventions in relation to the resulting health outcomes. Given the finite nature of health-care resources, this information can be valuable to inform decisions on efficient allocation of those resources. For example, the potential cost advantage of monotherapy with a fluoroquinolone over combination treatment with a ß-lactam and macrolide in CAP was confirmed in a trial-based cost analysis of sequential IV/po gatifloxacin compared to ceftriaxone and erythromycin/clarithromycin.4 This raises the question as to whether other new fluoroquinolones, such as moxifloxacin, will prove to be cost-effective in practice. Therefore, we studied the cost-effectiveness of two alternative antimicrobial strategies in CAP. The first was a strategy of sequential IV and oral moxifloxacin. The comparator was sequential IV and oral ß-lactam (co-amoxiclav), to which the investigator could add a macrolide (clarithromycin) at their discretion; this comparator regimen is commonly used in Europe.5 6

Our evaluation was based on a recent prospective, randomized, active-controlled trial (the TARGET study7 ), which compared treatment with moxifloxacin against co-amoxiclav with or without clarithromycin (AMC ± CLA) in patients with CAP requiring initial parenteral therapy. Significantly higher clinical and bacteriologic success rates were demonstrated in patients treated with moxifloxacin. In addition, the speed of response was significantly faster in patients receiving moxifloxacin, which resulted in a hospital duration of almost 1 day less than for patients in the co-amoxiclav treatment arm.

Economic evaluations designed as an integral component of randomized clinical trials are desirable as they draw on the inherent validity of the trial design; however, trial designs do not always reflect real-world use of health-care resources, so the most useful of randomized clinical trials for economic purposes are those that minimize the extent of protocol-driven deviations from actual clinical practice. In the TARGET trial,7 a number of clinical decisions were left to the discretion of the investigators, such as whether to add clarithromycin in individual cases, to switch patients from an IV to an oral formulation at any time following the first 3 days of IV treatment, or whether and when to discharge from hospital. Although the design did not permit blinded assessment, the potential for bias was minimized by using standardized criteria for determining clinical response to treatment. The bacteriologic and clinical outcomes, including objective criteria such as resolution of fever, were consistent, supporting the validity of these clinical judgments. Moreover, the unblinded nature of this trial allowed physicians to treat according to usual practice, allowing a more realistic assessment of the level of cost-effectiveness likely to occur in everyday use.8

In the present study, we considered health-care resource utilization data as well as the clinical outcomes collected in the TARGET trial7 to analyze the cost-effectiveness of moxifloxacin. The countries selected for the CEA were Germany and France, two countries that recruited a substantial number of patients for the trial.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
The TARGET Study
The TARGET study7 was a multinational, multicenter, randomized, open-label, phase III clinical trial designed to examine the safety and efficacy of sequential IV/po moxifloxacin compared to a standard regimen of sequential IV/po co-amoxiclav with or without clarithromycin (IV or oral). The study population was adult patients newly hospitalized with radiologic evidence of CAP and requiring initial parenteral therapy. Exclusion criteria included the presence of coexisting disease considered likely to affect the outcome of the study (eg, lung cancer, empyema, or severe cardiac failure) or a rapidly fatal underlying disease; known prolongation of the QT interval or the use of class IA or class III antiarrthythmics; known hypersensitivity to fluoroquinolones, ß-lactams, or macrolides; aspiration pneumonia; and pretreatment with systemic antibacterial agents for > 24 h prior to enrollment in the study. Patients who had received unsuccessful antibacterial therapy for at least 72 h for the current pneumonia episode could be enrolled unless the antibacterial regimen contained a fluoroquinolone or a ß-lactam-lactamase inhibitor combination. A secondary objective of the trial was to collect resource utilization data associated with CAP treatment for use in the economic evaluation reported here.

The trial started in February 1999 and was completed in May 2000, enrolling a total of 628 adult patients. Study criteria included radiologic evidence of CAP, temperature >= 38.5°C or leukocytosis, at least one clinical symptom of pneumonia (including cough, purulent sputum, dyspnea, rigors, pleuritic chest pain, or auscultatory findings), as well as the need for parenteral therapy. Patients were enrolled at 65 centers from the following 10 countries: Germany (19% of patients), Greece (16%), Israel (13%), South Africa (12%), and France (10%), followed by the United Kingdom (9%), Switzerland (8%), Spain (7%), Belgium (4%), and Russia (2%).

Patients were randomized to one of two treatment groups to be treated for a minimum of 7 days and a maximum of 14 days. Group 1 received moxifloxacin, 400 mg IV qd for a minimum of 3 days, followed by a switch to moxifloxacin, 400 mg po qd, at the discretion of the investigator. Group 2 received co-amoxiclav, 1.2 g IV tid for a minimum of 3 days, followed by a switch to co-amoxiclav, 625 mg po tid. In addition, at the discretion of the investigator, patients in group 2 received clarithromycin, 500 mg bid, either IV or po starting on the first day of the treatment period.

Patients enrolled in the trial were followed up until 21 to 28 days after the last dose of study medication. The primary efficacy variable for the trial was the clinical response at 5 to 7 days after the termination of study drug treatment (ie, test-of-cure [TOC] visit).

Economic Methods
An analysis of the costs and clinical consequences of treatment with moxifloxacin vs co-amoxiclav in patients with CAP was conducted. The analysis plan prespecified that results would be presented as a probabilistic estimate of the incremental cost of using moxifloxacin compared to AMC ± CLA per additional patient cured.

The patient population for economic evaluation included all patients properly randomized who received at least one dose of study medication. Economic evaluation requires a fixed period of time over which treatment costs and effectiveness are evaluated for each treatment alternative. The TOC visit, at which time the primary clinical efficacy was assessed, occurred within a window of 12 to 21 days based on the variable treatment and visit time frames. The maximum of 21 days (ie, 14 days of maximum treatment followed by 7 days of follow-up) was selected for the economic evaluation time frame since it corresponded to the primary clinical efficacy measure.

Resource Utilization and Costs
To determine if resource data were sufficiently homogeneous to be pooled across countries participating in the TARGET study,7 we compared the number of days spent in the hospital, as the most important component of that resource use. Because the number of days in hospital was nonnormal, a ranking procedure was used to compute normal scores. These ranks were then analyzed using an analysis of variance model that included the main effects (country and treatment group) and the interaction of the main effects (country-by-treatment group). Although resource use did differ significantly between countries, no consistent difference in the effect of study treatment on that resource use (country-by-treatment interaction) was observed. As a result, we concluded that pooling of all resource data was a valid basis for calculating single country-specific costs, consistent with current practice.9 10

Using pooled data, CAP-related health-care resource consumption was compared between treatment groups over the 21-day evaluation time frame from specific health-care perspectives in two countries: (1) Germany, from the perspective of the mandatory Sickness Funds (Gesetzliche Krankenkassen Versicherung); and (2) France, from the perspective of the public health insurance sector.

Resource data were then valued using 2000/2001 unit prices collected from each of Germany and France. The CAP-related cost per patient for each country analysis was determined by averaging the total cost for patients in each group. The total CAP-related cost included study medications, hospitalizations, outpatient concomitant and follow-up CAP-related medications, outpatient radiologic procedures, and outpatient therapeutic adjuncts over the 21-day evaluation time frame and as covered by the public health-care system of each country. Health resources utilized for treatment of adverse events (AEs) were not collected in the clinical trial, and were not estimated for the economic analysis, since based on the incidence of AEs the magnitude of AE-related costs would be minor.

Public prices were assumed for all CAP-related medications,11 12 including study antibiotics prescribed in hospital. The unit price for oral moxifloxacin in Germany was based on the market price of the product, while the expected price at market introduction was used for IV moxifloxacin. For France, the market price of oral moxifloxacin was obtained by converting the ex-manufacturer price to the public price using a standard conversion formula for France. Since the price of the IV form was not available at the time of the study, the German price of IV moxifloxacin was converted to Francs using conversion rates current at the time of analysis. This approach was believed to produce a conservative estimate of the price of IV moxifloxacin, as the price was markedly higher than the public price of a recently approved fluoroquinolone in France. For study comparator of medications in Germany, a weighted cost was assumed for the oral forms of co-amoxiclav and clarithromycin based on the current market share of brand and generic name products. For IV co-amoxiclav, the price of branded product was assumed. Since IV clarithromycin is not available in Germany, a shadow cost was estimated based on the cost of the po form and the corresponding oral to IV cost ratio in France. For the French analysis, the costs of branded products were used for co-amoxiclav and clarithromycin. Fully allocated hospitalization per diem costs13 14 and unit prices for outpatient radiologic procedures15 16 and outpatient therapeutic adjuncts17 were obtained from local sources.

For patients who withdrew from the study prior to the 21-day time point, for each cost component a regression equation was used to estimate their daily cost for the remaining evaluation period, based on the average daily cost for patients who completed the study. Factors included were country and clinical status at the TOC visit; for patients with unknown or indeterminate clinical status at study dropout, only country was included in the regression equation. The most important cost factor to consider in the imputation was considered to be death or clinical outcome and whether the patient was still in hospital. Patients discharged from hospital at study withdrawal were assumed to remain out of hospital, and no further CAP-related cost was assigned to patients who died or who had clinical cure and were out of hospital. Groups were reasonably equal in the proportion of patients (10% for moxifloxacin, 13% for AMC ± CLA) with these outcome factors and in the amount of time that required imputation (87.3 days for moxifloxacin, 90.5 days for AMC ± CLA). The small proportion of missing data and its relatively even distribution between groups minimizes the impact of the imputation on the estimated treatment difference in costs, and we concluded that the method we applied was not likely to give rise to treatment bias.

Clinical Consequences and Choice of Outcome Measure
Since in CAP, the immediate and overriding clinical objective is the rate of clinical cure we selected "cost per additional patient cured" as the economic parameter likely to be most meaningful to those concerned with decisions on treatment and consistent with the source of data, the TARGET clinical trial.7 Estimates of treatment effectiveness were obtained from the pooled clinical results for patients who had been properly randomized and received at least one dose of study medication. The overall treatment effectiveness measure was the proportion of patients cured based on the primary efficacy parameter: the clinical response to study drug at the TOC visit (ie, 5 to 7 days after treatment). For the economic analysis, the most conservative approach was used to estimate clinical effectiveness, considering all patients with the outcome of indeterminate or missing as clinical failures.

CEA
The incremental CAP-related cost and effectiveness of treatments were to be compared in a CEA. One-way sensitivity analyses were conducted to explore the uncertainty in several parameters including the cost of hospitalization and the cost of IV clarithromycin. To explore the variability inherent in the cost-effectiveness results, a confidence region was constructed by bootstrapping, a resampling procedure that is commonly applied in economic studies.18 19 20 21 The bootstrapped distribution of data for each group were then used to generate cost-effectiveness acceptability curves, which report the probability that moxifloxacin is cost-effective over a range of possible criteria for acceptable cost per unit of effectiveness. The process was replicated for the German and French data.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Primary Clinical Outcomes
A total of 622 study patients with clinical signs and symptoms, including radiology, confirming the CAP diagnosis were included in the economic analysis. More than half of the patients had a diagnosis of severe CAP according to the 1993 American Thoracic Society criteria.22 The mean ages of patients in the moxifloxacin and AMC ± CLA groups were 55.2 years and 55.9 years, respectively; a total of 64.1% and 64.5% of patients were male, 28% and 29% had preexisting bronchopulmonary disease, and 59% and 61% were smokers or had a history of smoking. Bacterial cultures or serology were performed in 52% of patients, and the most common organisms identified were Streptococcus pneumoniae and Haemophilus influenzae, accounting for 55% and 20% of all identified pathogens, respectively. In the AMC ± CLA group, 60% of patients were treated with both co-amoxiclav and clarithromycin.

The intention-to-treat population showed a 93.5% rate of clinical cure for moxifloxacin and 85.2% for AMC ± CLA, an 8.3% (95% confidence interval [CI], 3.1 to 13.6) higher rate of cure in the moxifloxacin group. The secondary end point, clinical cure at the late follow-up assessment, showed a difference in favor of moxifloxacin of 8.8% in the intention-to-treat population (95% CI, 1.6 to 16.0). The differences in favor of moxifloxacin in the per-protocol (PP) population were 8.1% (95% CI, 2.9 to 13.2) and 9.4% (95% CI, 2.6 to 16.3) at TOC and follow-up assessments, respectively.

For the cost-effectiveness analysis reported below the "effectiveness" unit, the rate of clinical cure, was defined in keeping with health economic practice. The population consisted of all patients properly randomized who received at least one dose of study medication. Patients whose clinical outcomes were recorded as indeterminate or missing were categorized as failures. According to these economic criteria, the rate of clinical cure at the TOC visit was 80.7% for moxifloxacin and 75.4% for AMC ± CLA, a 5.3% (95% CI, 1.2 to 11.8%) higher rate of cure in the moxifloxacin group. Treatment response was classified as treatment failure for 5.6% of patients receiving moxifloxacin and 13.1% of patients receiving AMC ± CLA; the remaining patients were classified as having indeterminate response or missing outcome data.

Secondary Clinical Outcomes
Fewer adverse events, serious adverse events, and deaths were reported in the moxifloxacin group. Among adverse events considered drug related, events related to the GI tract were the most common in both treatment groups. The most frequent of these adverse events included abnormal liver function test results (7.3% moxifloxacin; 5.9% AMC), diarrhea (7.0% moxifloxacin; 5.3% AMC), and nausea (3.3% moxifloxacin; 3.7% AMC).

A faster treatment response, as measured by time to apyrexia, was seen in the moxifloxacin group. Median times to apyrexia were 2 days for the moxifloxacin group and 3 days for the AMC ± CLA group (p = 0.008); 58.6% of patients receiving moxifloxacin were apyrexial by day 2 compared to 46.7% of the patients receiving AMC ± CLA, a statistically significant result. The more rapid onset of action noted in patients receiving moxifloxacin resulted in earlier conversion to oral therapy as demonstrated by the shorter duration of IV treatment by 0.8 days (Table 1 ). In addition, approximately half of the patients receiving moxifloxacin received only 3 days or less of the IV formulation compared to less than one fifth of the AMC ± CLA group. Over the 21-day time frame, patients receiving moxifloxacin had 0.8 days shorter hospital stay (95% CI, - 0.01 to 1.63 days).


View this table:
[in this window]
[in a new window]

  		Table 1. CAP-Related Resource Utilization*

 
The bacteriologic results underlying the clinical outcomes were measured in microbiologically valid patients in the PP population. Bacteriologic success, defined as eradication or presumed eradication, was achieved in 93.7% and 81.7% of moxifloxacin- and AMC±CLA-treated patients, respectively, a 12.0% (95% CI, 1.21 to 22.9%) difference.

Resource Utilization and Costs
Use of resources both within hospital and subsequent to discharge within the study period are reported in Table 1 . These units were valued from the German and French perspectives, according to the method previously described, to give the total CAP-related costs shown in Table 2 . Costs were substantially higher from the French perspective than the German, due to the higher standard unit cost of hospital stays that were reported for France. Moreover, hospital costs accounted for > 98% of the total. This, together with the shorter duration of hospital stays for the moxifloxacin group, produced the result that costs were lower for the moxifloxacin group. Treatment with moxifloxacin was less costly than the AMC ± CLA regimen by 266 (11%) and 381 (9%) in Germany and France. respectively. These differences just failed to reach statistical significance at the 95% confidence level. Resources utilized for treatment of AEs were not collected in the clinical trial and were not estimated for the economic analysis, since based on the incidence of AEs the magnitude of AE-related costs would be minor.


View this table:
[in this window]
[in a new window]

  		Table 2. CAP-Related Costs to the German and French Health-Care Systems*

 
Cost-effectiveness and Sensitivity Analysis
Over the 21-day time frame, treatment with moxifloxacin was dominant over the AMC ± CLA regimen, with 5.3% higher rates of clinical cure and lower total CAP-related cost per patient to both the German and French health-care systems. The base-case results are shown in Table 3 for the German analysis and in Table 4 for the French analysis. The results of several one-way sensitivity analyses are also shown.


View this table:
[in this window]
[in a new window]

  		Table 3. Base-Case and Sensitivity Analysis Results: Germany

 

View this table:
[in this window]
[in a new window]

  		Table 4. Base-Case and Sensitivity Analysis Results: France

 
Since the cost of hospitalization represented such a large component of the overall cost of CAP treatment, one-way sensitivity analyses were conducted using alternative sources of hospitalization cost data while the remaining variables were held constant. For Germany, hospital per diems from the University Hospital in Cologne were substituted, which are representative of a more expensive, teaching hospital environment. For the French analysis, the variance in the hospital per diem costs used in the base-case analysis was explored by substituting the minimum and maximum per diem values. In spite of the effect on the total cost of CAP treatment, moxifloxacin remained dominant in these sensitivity analyses.

The above-mentioned analyses were based on the mean differences in costs and effects between moxifloxacin and AMC ± CLA observed in the TARGET trial,7 but take no account of their sampling distributions. It is possible that for some combinations of costs and effects within the distribution, moxifloxacin would no longer be dominant. To explore the variability inherent in the cost-effectiveness results, a bootstrapped confidence region was constructed.

Figure 1 presents the results of the bootstrap for the French analysis; the German analysis follows a similar pattern so is omitted for brevity. Individual pairs of cost and effectiveness differences between moxifloxacin and the AMC ± CLA treatment are shown on the cost-effectiveness plane as a scatter plot. The origin represents the point of zero difference in costs and effectiveness. The lower right quadrant represents the area in which moxifloxacin dominates AMC ± CLA. The clustering of points in the lower right quadrant demonstrates the very high probability of moxifloxacin being both less costly and more effective than AMC ± CLA. The mean value of the bootstrapped incremental cost-effectiveness ratios (ICERs) of - 10,019 and - 13,732 per additional patient cured for the German and French analysis, respectively, falls in this quadrant, as shown by the solid points lying in the center of the "cloud" of bootstrapped estimates. Only the first 1,000 samples have been displayed on each graph. To assist in the interpretation of the plots, a series of lines have been constructed passing through the origin, whose slopes represent possible acceptability thresholds of the maximum cost that a decision maker may be prepared to pay per additional patient cured. The proportion of points that fall below the line of the chosen acceptability criterion represents the probability that moxifloxacin is cost-effective. If the decision maker wishes to adopt the stricter criterion for cost-effectiveness that all worse clinical outcomes are to be excluded regardless of the magnitude of cost saving, then only those points falling below the threshold line that are to the right of the vertical axis would count.



View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Bootstrap results plotted on the cost-effectiveness plane: France. CAT = cost-effectiveness acceptability threshold (per additional patient cured, in Euros). Note that only 1,000 of the ICERs have been included on the plot.

 
Figure 2 shows the bootstrap results presented as a cost effectiveness acceptability curve for each country. These curves indicate the probability that moxifloxacin is cost-effective vs AMC ± CLA over a continuous range of acceptability thresholds. The horizontal axis shows a range of maximum amounts that a decision maker may be willing to pay to cure an additional patient, while the vertical axis gives the proportion of the estimated cost-effectiveness ratios that meet each cost-effectiveness criterion.



View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Results expressed as incremental cost-effectiveness acceptability curves.

 
The interpretation of the point at which the curves intersect the vertical axis (ie, 0.97 for Germany and 0.95 for France) is that in > 95% of the bootstrapped samples, moxifloxacin is cost saving relative to the AMC ± CLA comparator treatment. The probability of moxifloxacin being cost-effective relative to AMC ± CLA increases as the threshold values increase. For example, if a decision maker is willing to pay up to 2,000 per additional patient cured, the probability of moxifloxacin being cost-effective is 98% and 97% in Germany and France, respectively.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Our study has presented the results of a cost-effectiveness analysis in which the difference in costs and outcomes of moxifloxacin vs a standard antimicrobial therapy, AMC ± CLA, has been assessed in the treatment of patients with CAP from the perspectives of health-care payers in Germany and France. We believe that the conduct of the economic study as an integral part of the TARGET clinical trial7 adds to the validity of the economic conclusions. Germany and France represent two important European countries, but the study could be extended to consider other countries. The "background" duration of hospital stay, a major element of the cost of CAP treatment, did vary significantly across the wide range of countries included in the TARGET study.7 This was to be expected due to differences between countries in the organization and financing of health care and in clinical practice23 ; however, the reduction in the length of stay induced by treatment did not differ by country, as demonstrated by the lack of a country-by-treatment interaction effect. As there was no differential effect on resource use across countries, the use of the pooled trial data for analysis from the perspective of single countries was justified. Cost-effectiveness analyses from the perspectives of other countries participating in the trial would be equally feasible.

The TARGET study7 found statistically significantly higher clinical success rates for moxifloxacin at both primary and secondary end points, the TOC and long-term follow-up assessments, respectively, in the valid-PP population. To assess the economics of moxifloxacin, costs and outcomes were compared in the broader population of all randomized patients who received at least one dose of study medication. The primary effectiveness measure, the rate of clinical success at the TOC visit, showed a difference between groups of 5.3% that was clinically important and favorable for moxifloxacin, but not quite statistically significant (95% CI, 1.2 to 11.8%).

The evaluation time frame selected for economic analysis was 21 days, corresponding to the maximum time to clinical assessment at the TOC visit, the most important clinical end point. The total study period for a patient completing to long-term follow-up at 21 to 28 days after treatment could vary from 28 to 42 days, so that evaluation of cost-effectiveness over the maximum 42-day time period would have required extensive imputation of cost data. The evaluation of costs and consequences within 21 days is a conservative estimate of the CAP-related cost savings, since the difference in mean duration of hospital stay between treatment groups was larger at final follow-up (0.9 days) than at 21 days (0.8 days).

Although the economic evaluation was based on the primary outcome measure of clinical cure rate, an important finding of the TARGET trial7 was the consistency of the results across the various outcome parameters: clinical cure and bacteriologic eradication rates, time to resolution for fever, and duration of hospital stay. The significantly shorter time to resolution of fever in patients treated with moxifloxacin is relevant to both the clinical and economic outcomes of the study. Clinically, this result is consistent with a more rapid onset of action, which is supported by the observation that patients receiving moxifloxacin were converted to oral therapy earlier than patients in the AMC ± CLA group. The duration of IV treatment was shorter for moxifloxacin patients by 0.8 days, and approximately half (50.2%) of patients receiving moxifloxacin received only 3 days or less of the IV formulation compared to less than one fifth (17.8%) of the AMC ± CLA group. Although conversion to oral therapy was entirely at the discretion of the investigator after 3 days of treatment, all published conversion guidelines for the treatment of CAP include resolution of fever as a criterion for switch therapy.24 Consistent with previous studies25 26 27 28 that have demonstrated that early switch strategies have resulted in reduced lengths of stay, patients receiving moxifloxacin were discharged 0.8 days earlier, on average, than patients in the AMC ± CLA group.

There are a number of potential reasons to explain the finding of clinical and bacteriologic superiority of therapy with moxifloxacin compared to AMC ± CLA. Differently from ß-lactams or macrolides, moxifloxacin exhibits concentration-dependent killing of bacteria and is rapidly bactericidal.29 Furthermore, concentrations of moxifloxacin obtained in different pulmonary compartments exceed serum concentrations multiple-fold, thus providing very high local concentrations at the actual site of infection.30

The TARGET trial7 used an unblinded, randomized design. Blinding was considered not feasible for formulation, ethical (multiple placebo infusions required), and methodologic reasons (sequential therapy with flexible regimen). While recognizing that blinding is desirable to minimize the potential for bias, we believe this threat was minimized by the use of standardized assessment methodology. The sound bacteriologic explanation for the observed clinical outcomes suggests that the study design succeeded in its aim of maximizing internal validity subject to the constraints of practical feasibility and ethical acceptability.

The difference in the length of hospital stays between treatment groups was pivotal to the results of the economic evaluation, which found treatment with moxifloxacin to be less costly to the German and French health-care systems by 266 and 381 per patient, respectively, representing resources that can be redeployed in those systems. More than 98% of the total cost of CAP treatment in each analysis was due to the costs incurred in hospital. Although our analysis used an average per diem (which represents the mean of earlier more expensive hospital days and later less expensive days of care) to estimate the cost of hospitalization, it is likely that our estimate of cost savings due to early discharge is conservative given that the days of hospitalization averted were relatively expensive days when IV antibiotic therapy would have been administered. Our results are consistent with those of several other studies of patients with pneumonia requiring hospitalization, which have been conducted in Germany31 and the United States.4 26 32 Each of these studies demonstrated that treatments resulting in shorter hospitalization durations (ranging from 1 to 2.4 fewer days) translated into cost savings relative to the comparator treatments.

Although we used the public prices of pharmaceuticals to estimate costs of study medications received in hospital, actual purchase prices may be lower. These discounts vary by institution and supplier, and information regarding such agreements is not publicly available. It would be difficult, therefore, to estimate the true cost of study medications prescribed in hospital for use in this analysis. However, even if the entire cost of study medications received by the AMC ± CLA group in hospital (ie, Germany, 248; France, 107) were subtracted from the total CAP-related cost, the total cost of treatment with moxifloxacin would remain less than the AMC ± CLA group.

Since the broadest possible patient population was used, consistent with economic guidelines, effectiveness (clinical cure rate) although in favor of moxifloxacin was not statistically significantly different between groups. Moxifloxacin showed cost savings, which also did not reach traditional significance levels due to the high variability commonly seen in cost data; however, bootstrap analysis samples the joint distribution of costs and effects, so does not require separate significance testing for each variable. Even under the most stringent criterion for cost-effectiveness, namely that treatment with moxifloxacin should be cost saving, there is a 97% and 95% chance in the German and French analysis, respectively, of moxifloxacin being cost saving. If decision makers are willing to pay up to 2,000 to treat an additional patient successfully, the probability of moxifloxacin being cost-effective vs AMC ± CLA treatment increased to 98% and 97%, respectively. Decision makers may debate whether cost savings to a hospital created by a shortened length of stay will be realized, since a freed hospital bed is likely to be occupied by a new patient; however, shortening hospital stay may be quite valuable at a time when constraints on hospital beds may be reached due to the seasonal nature of CAP.

We conclude that empirical treatment with sequential IV/po moxifloxacin is cost-effective relative to sequential IV/po AMC ± CLA in the treatment of hospitalized patients with CAP. Treatment with moxifloxacin is likely to result in cost savings to German and French payers, freeing resources for alternative health-care uses.


   Acknowledgements
 
The authors thank the following people for contributing to this study: Jacqueline Gough and Dan Pericak of Innovus Research Inc., Burlington, ON, for statistical analysis; Professor Gérard de Pouvourville, INSERM U 537, Le Kremlin Bicêtre, France, for advice on costing methods for France; Karine Séjean, of the Faculté de Médecine, Saint-Antoine, Paris, for costing of resource items for that country; and Felicitas Kühne, a consultant to Innovus, for costing of resource items for Germany.


   Footnotes
 
Abbreviations: AE = adverse event; AMC ± CLA = co-amoxiclav with or without clarithromycin; CAP = community-acquired pneumonia; CEA = cost-effectiveness analysis; CI = confidence interval; ICER = incremental cost-effectiveness ratio; PP = per protocol; TOC = test of cure

The current study was fully supported financially by Bayer and conducted by Innovus Research Inc., an independent contract research organization under the direction of Michael Drummond. Clinical outcomes were based on the report of the TARGET study designed and conducted by Bayer.

Dr. Rolf Kubin, a clinician employed by Bayer and experienced in the treatment of CAP in Europe, provided input as required throughout the study. Analyses of cost and cost-effectiveness were conducted by Innovus Research Inc., and a project team including individuals from Innovus and Bayer reviewed the analysis plan, study report and manuscript, and agreed by consensus on final manuscript content.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions{at}chestnet.org).

Received for publication May 30, 2002. Accepted for publication February 4, 2003.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Marrie, TJ (1998) Community-acquired pneumonia: epidemiology, etiology, treatment. Infect Dis Clin North Am 12,723-740[ISI][Medline]
  2. Bartlett, JG, Mundy, LM Community-acquired pneumonia. N Engl J Med 1995;333,1618-1624[Free Full Text]
  3. Guest, JF, Morris, A Community-acquired pneumonia: the annual cost to the National Health Service in the UK. Eur Respir J 1997;10,1530-1534[Abstract]
  4. Dresser, LD, Niederman, MS, Paladino, JA Cost-effectiveness of gatifloxacin vs ceftriaxone with a macrolide for the treatment of community-acquired pneumonia. Chest 2001;119,1439-1448[Abstract/Free Full Text]
  5. European Study on Community Acquired Pneumonia (ESOCAP) Committee. Guidelines for management of adult community-acquired lower respiratory tract infections. Eur Respir J 1998;11,986-991[CrossRef][ISI][Medline]
  6. American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163,1730-1754[Free Full Text]
  7. Finch, R, Schürmann, D, Collins, O, et al A randomized controlled trial of sequential intravenous (i.v.) and oral moxifloxacin compared with sequential i.v. and oral co-amoxiclav with or without clarithromycin in patients with community acquired pneumonia requiring initial parenteral treatment. Antimicrob Agents Chemother 2002;46,1746-1754[Abstract/Free Full Text]
  8. Freemantle, N, Drummond, M Should clinical trials with concurrent economic analyses be blinded? JAMA 1997;277,63-64[CrossRef][ISI][Medline]
  9. Drummond, M, Pang, F Transferability of economic evaluation results. Drummond, MF McGuire, A eds. Economic evaluation in health care: merging theory with practice 2001,256-276 Oxford University Press. New York, NY:
  10. Willke, RJ, Glick, HA, Polsky, D, et al Estimating country-specific cost-effectiveness from multinational clinical trials. Health Econ 1998;7,481-493[CrossRef][ISI][Medline]
  11. Rote Liste Service GmbH. Rote Liste 2001. Arzneimittelverzeichnis für Deutschland [medication listing for Germany]. 2001 Editio Cantor Verlag. Aulendorf, Germany:
  12. Thesaurus Vidal-Semp 2001. Paris, France: Vidal, 2001
  13. Verband der Private Krankenversicherung (PKV). Die private Krankenversicherung. Zahlenbericht 1999/2000. Available at: http://www.pkv.de. Accessed July 2, 2003
  14. Ministère de l’Emploi de la Solidarité, Paris. Programme de Médication des Systèmes d’Information (PMSI) 2000. Available at: http://www.le-pmsi.fr. Accessed July 2, 2003
  15. Büro für informations Systeme. Gebührenordnung für Ärzte (GOÄ). Available at: http://www.e-bis.de/goae/defaultFrame.htm. Accessed July 2, 2003
  16. Nomenclature Générale des Actes professionels (NGAP). Paris, France: L’Union des Caisses Nationales de Sécurité (UCANSS), 2001
  17. Tarif Interministériel des Prestations Sanitaires (TIPS). Paris, France: L’Union des Caisses Nationales de Sécurité Sociale (UCANSS), 2001
  18. Gray, A, Raikou, M, McGuire, A, et al Cost-effectiveness of an intensive blood glucose control policy in patients with type 2 diabetes: economic analysis alongside randomised controlled trial (UKPDS 41). BMJ 2000;320,1373-1378[Abstract/Free Full Text]
  19. LePen, C, Lilliu, H, Keller, T, et al The economics of TRACE: a cost-effectiveness analysis of trandolapril in postinfarction patients with left ventricular dysfunction. Pharmacoeconomics 1998;14,49-58[Medline]
  20. Collet, JP, Ducruet, T, Haider, S, et al Economic impact of using an immunostimulating agent to prevent severe acute exacerbations in patients with chronic obstructive pulmonary disease. Can Respir J 2001;8,27-33[Medline]
  21. Briggs, A Handling uncertainty in economic evaluation and presenting the results. Drummond, MF McGuire, A eds. Economic evaluation in health care: merging theory with practice 2001,172-214 Oxford University Press. New York, NY:
  22. Niederman, MS, Bass, JB, Campbell, GC, et al Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. Am Rev Respir Dis 1993;148,1418-1426[ISI][Medline]
  23. Drummond, M, Pang, F Transferability of economic evaluation results. Drummond, MF McGuire, A eds. Economic evaluation in health care: merging theory with practice 2001,256-276 Oxford University Press. New York, NY:
  24. Rhew, DC, Tu, GS, Ofman, J, et al Early switch and early discharge strategies in patients with community-acquired pneumonia. Arch Intern Med 2001;161,722-727[Abstract/Free Full Text]
  25. Marrie, TJ, Lau, CY, Wheeler, SL, et al A controlled trial of a critical pathway for treatment of community-acquired pneumonia: CAPITAL study investigators; Community-Acquired Pneumonia Intervention Trial Assessing Levofloxacin. JAMA 2000;283,749-755[Abstract/Free Full Text]
  26. Omidvari, K, de Boisblanc, BP, Karam, G, et al Early transition to oral antibiotic therapy for community-acquired pneumonia: duration of therapy, clinical outcomes, and cost analysis. Respir Med 1998;92,1032-1039[CrossRef][ISI][Medline]
  27. Hendrickson, JR, North, DS Pharmacoeconomic benefit of antibiotic step-down therapy: converting patients from intravenous ceftriaxone to oral cefpodoxime proxetil. Ann Pharmacother 1995;29,561-565[Abstract]
  28. Siegal, RE, Halpern, NA, Almonoff, PL, et al A prospective randomized study of inpatient IV antibiotics for community-acquired pneumonia: the optimal duration of therapy. Chest 1996;110,965-971[Abstract/Free Full Text]
  29. Dalhoff, A, Petersen, U, Endermann, R In vitro activity of BAY 12–8039, a new 8-methoxyquinolone. Chemotherapy 1996;42,410-425[ISI][Medline]
  30. Soman, A, Honeybourne, D, Andrews, J, et al Concentrations of moxifloxacinin serum and pulmonary compartments following a single 400 mg oral dose in patients undergoing fibre-optic bronchoscopy. J Antimicrob Chemother 1999;44,835-838[Abstract/Free Full Text]
  31. Dietrich, ES, Joseph, U, Vogel, F, et al Cost-effectiveness of ceftriaxone 1g vs second-generation cephalosporins in the treatment of pneumonia in general medical wards in Germany. Infection 1999;27,148-154[Medline]
  32. Rittenhouse, BE, Stinnett, AA, Dulisse, B, et al An economic evaluation of levofloxacin versus cefuroxime axetil in the outpatient treatment of adults with community-acquired pneumonia. Am J Manag Care 2000;6,381-389[Medline]



This article has been cited by other articles:


Home page
 ChestHome page
T. T. Bauer, T. Welte, C. Ernen, B. M. Schlosser, I. Thate-Waschke, J. de Zeeuw, and G. Schultze-Werninghaus
Cost Analyses of Community-Acquired Pneumonia From the Hospital Perspective
Chest, October 1, 2005; 128(4): 2238 - 2246.
[Abstract] [Full Text] [PDF]

Home page
 Eur Respir JHome page
J-B. Wasserfallen, V. Erard, A. Cometta, T. Calandra, and O. Lamy
Cost-effectiveness of full-course oral levofloxacin in severe community-acquired pneumonia
Eur. Respir. J., October 1, 2004; 24(4): 644 - 648.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Drummond, M. F.
Right arrow Articles by Sagnier, P.-P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Drummond, M. F.
Right arrow Articles by Sagnier, P.-P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS