|
Decision analysis of antibiotic and diagnostic strategies in ventilator-associated pneumonia |
Ost D E, Hall C S, Joseph G, Ginocchio C, Condon S, Kao E, Larusso M, Itzla R, Fein A M |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology This study considered the treatment of ventilator-associated pneumonia (VAP). Four diagnostic strategies were considered in conjunction with four antibiotic regimens. The diagnostic strategies were no diagnostic test (empirical treatment only), quantitative nonprotected endotracheal cultures, bronchoscopy, and nonbronchoscopic mini-bronchoalveolar lavage (mini-BAL). The antibiotic regimens consisted of administering either no, one, two or three antibiotics. The types of antibiotic were not reported. Following a positive diagnostic test, the antibiotics would be adjusted to treat the organism identified. If the diagnostic test was negative, antibiotic treatment would continue if the patient was unstable and cease if the patient was stable. Sixteen possible combinations of diagnostic strategy and antibiotic use were assessed.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised immunocompetent patients in intensive care, intubated for 7 days, with symptoms of VAP. Symptoms of VAP were based on the Centers of Disease Control criteria of fever, purulent secretions, leukocytosis, and radiographic infiltrates.
Setting The setting was tertiary care. The economic study was carried out in the USA.
Dates to which data relate The effectiveness and resource use evidence was derived from studies published between 1990 and 2003. The price year was 2002.
Source of effectiveness data The effectiveness data were derived from a review or synthesis of completed studies.
Modelling A decision analysis model was constructed (using DATA PRO) to determine the clinical effectiveness, the resource use and the costs of each of the 16 strategies.
Outcomes assessed in the review The following model input parameters were identified from the review:
the baseline mortality;
the probability that the initial antibiotics cover pathogen;
the probability that one, two and three antibiotics cover pathogen;
the probability of positive bronchoscopic cultures;
the probability of positive nonbronchoscopic cultures;
the probability of positive tracheal secretion cultures;
the mortality attributed to VAP if the antibiotics were correct;
the efficacy of adequate initial antibiotic therapy;
the efficacy of adequate late antibiotic therapy;
the increase in hospital stay due to VAP with correct antibiotics;
the increase in intensive care unit stay due to VAP with correct antibiotics;
the increase in length of stay with inadequate antibiotics in comparison with adequate antibiotics; and
the probability of stopping antibiotics if the diagnostic test is negative.
Study designs and other criteria for inclusion in the review The types of publication included in the review were clinical trials, congresses, controlled clinical trials, guidelines, journal articles, meta-analysis, multi-centre studies, randomised controlled trials and reviews of the literature. The selection criteria for the studies included immunocompetent human adults, VAP diagnosed on the basis of protected specimen culture or histopathology, and peer-reviewed journals only.
Sources searched to identify primary studies MEDLINE was searched from 1986 to 2002 for primary studies. The bibliographies of the included papers were examined for additional papers.
Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included Twenty-nine primary studies were included in the review.
Methods of combining primary studies Investigation of differences between primary studies Results of the review The following model input parameters were identified from the review:
the baseline mortality was 17.5%;
the probability that the initial antibiotics cover pathogen was 60%;
the probability that one antibiotic covers pathogen was 40%;
the probability that two antibiotics cover pathogen was 60%;
the probability that three antibiotics cover pathogen was 80%;
the probability of positive bronchoscopic cultures was 85%;
the probability of positive nonbronchoscopic cultures was 80%;
the probability of positive tracheal secretion cultures was 70%;
the mortality attributed to VAP if the antibiotics were correct was 15%;
the efficacy of adequate initial antibiotic therapy was 0.55;
the efficacy of adequate late antibiotic therapy was 0.05;
the increase in hospital stay due to VAP with correct antibiotics was 11.8 days;
the increase in intensive care unit stay due to VAP with correct antibiotics was 9.9 days;
the increase in length of stay with inadequate antibiotics, compared with adequate antibiotics, was 3 days; and
the probability of stopping antibiotics if the diagnostic test was negative was 50%.
Measure of benefits used in the economic analysis The measure of health benefit used was hospital survival.
Direct costs The direct costs of the hospital were included in this study. The resource use data were taken from the model that provided the clinical effectiveness data. The costs of providing hospital care, including the costs of diagnostic tests, antibiotic treatment and intensive care unit stay, were identified. These costs were taken from hospital charges. The authors reported that charges were adjusted for departmental cost-to-charge ratios. However, they did not report the value of that ratio. A breakdown of the unit costs was provided. The price year was 2002. The costs were appropriately not discounted as they were incurred during less than one year.
Statistical analysis of costs No statistical analysis of the costs was undertaken.
Indirect Costs No indirect costs were included in the study.
Sensitivity analysis One-, two- and three-way sensitivity analyses were undertaken to assess variability in the data. A Monte Carlo simulation was also performed so that all variables were simultaneously varied. The majority of data ranges were taken from values given in the literature. The ranges of the cost data were plus or minus 25% of the baseline value.
Estimated benefits used in the economic analysis The survival probabilities were as follows:
no initial antibiotics, empirical treatment, 0.536;
no initial antibiotics, empirical treatment plus unprotected endotracheal cultures, 0.544;
no initial antibiotics, empirical treatment plus mini-BAL, 0.545;
no initial antibiotics, empirical treatment plus bronchoscopy, 0.546;
one initial antibiotic, empirical treatment, 0.600;
one initial antibiotic, empirical treatment plus unprotected endotracheal cultures, 0.605;
one initial antibiotic, empirical treatment plus mini-BAL, 0.605;
one initial antibiotic, empirical treatment plus bronchoscopy, 0.606;
two initial antibiotics, empirical treatment, 0.632;
two initial antibiotics, empirical treatment plus unprotected endotracheal cultures, 0.635;
two initial antibiotics, empirical treatment plus mini-BAL, 0.635;
two initial antibiotics, empirical treatment plus bronchoscopy, 0.635;
three initial antibiotics, empirical treatment, 0.663;
three initial antibiotics, empirical treatment plus unprotected endotracheal cultures, 0.665;
three initial antibiotics, empirical treatment plus mini-BAL, 0.665; and
three initial antibiotics, empirical treatment plus bronchoscopy, 0.666.
Cost results The total costs were as follows:
no initial antibiotics, empirical treatment, $29,733;
no initial antibiotics, empirical treatment plus unprotected endotracheal cultures, $30,327;
no initial antibiotics, empirical treatment plus mini-BAL, $30,430;
no initial antibiotics, empirical treatment plus bronchoscopy, $30,633;
one initial antibiotic, empirical treatment, $28,190;
one initial antibiotic, empirical treatment plus unprotected endotracheal cultures, $28,292;
one initial antibiotic, empirical treatment plus mini-BAL, $28,344;
one initial antibiotic, empirical treatment plus bronchoscopy, $28,523;
two initial antibiotics, empirical treatment, $27,853;
two initial antibiotics, empirical treatment plus unprotected endotracheal cultures, $27,462
two initial antibiotics, empirical treatment plus mini-BAL, $27,464;
two initial antibiotics, empirical treatment plus bronchoscopy, $27,618;
three initial antibiotics, empirical treatment, $27,516;
three initial antibiotics, empirical treatment plus unprotected endotracheal cultures, $26,632;
three initial antibiotics, empirical treatment plus mini-BAL, $26,583; and
three initial antibiotics, empirical treatment plus bronchoscopy, $26,712.
Synthesis of costs and benefits The authors calculated and reported the average cost per survival for each strategy. They also calculated and reported the incremental cost-effectiveness ratios for strategies using the same number of initial antibiotics.
The incremental cost per additional survival was as follows:
no initial antibiotic, empirical treatment, $55,447;
no initial antibiotics, empirical treatment plus unprotected endotracheal cultures, $72,847;
no initial antibiotics, empirical treatment plus mini-BAL, $101,479;
no initial antibiotics, empirical treatment plus bronchoscopy, $433,261;
one initial antibiotic, empirical treatment, $47,002;
one initial antibiotic, empirical treatment plus unprotected endotracheal cultures, $20,734;
one initial antibiotic, empirical treatment plus mini-BAL, $86,184;
one initial antibiotic, empirical treatment plus bronchoscopy, $634,288;
two initial antibiotics, empirical treatment plus unprotected endotracheal cultures, $43,260;
two initial antibiotics, empirical treatment plus mini-BAL, $819,710;
three initial antibiotics, empirical treatment plus mini-BAL, $39,967;
three initial antibiotics, empirical treatment plus bronchoscopy, $1,375,978.
A separate analysis was performed that considered antibiotic use as another form of "cost" (in terms of promoting antibiotic resistance). When comparing all strategies, the strategy of three antibiotics combined with mini-BAL was the most cost-effective. The incremental cost was 200.8 antibiotic days per additional survivor in comparison with the no antibiotic strategy.
The sensitivity analysis indicated that the cost results were sensitive to the average cost of antibiotics per day, the length of treatment, the sensitivity of diagnostic testing, and the cost of the mini-BAL test.
Authors' conclusions The most cost-effective strategy for the assessment and treatment of ventilator-associated pneumonia (VAP) was initial treatment with three antibiotics and testing with mini-bronchoalveolar lavage (mini-BAL).
CRD COMMENTARY - Selection of comparators This study did not use an explicit comparator in its assessment of diagnosis and treatment options. The authors reported only that, based on a review of the literature and American Thoracic Society guidelines, two antibiotics were used for the baseline analysis. You should consider how the various options examined relate to current practice in your setting before applying the results of this study.
Validity of estimate of measure of effectiveness The estimate of effectiveness used in this study was taken from a model. The input parameters were taken from primary studies that were identified through what appears to have been a systematic review of the literature. Although the search criteria and the sources searched were reported, the criteria used to assess the primary studies were not described. In addition, the authors did not indicate how the results of the primary studies were combined to produce a single value.
Validity of estimate of measure of benefit The measure of health benefit used in the economic analysis was taken from the model. The authors did not explicitly justify their choice of health benefit.
Validity of estimate of costs The authors reported that a hospital perspective was adopted. All the costs appropriate to that perspective appear to have been included in the analysis. The unit costs and resource use were reported, which assists the generalisability of the findings of the study. A comprehensive sensitivity analysis of resource use and unit costs was undertaken to assess the impact of variability in the data. This also enhances the generalisability of the study. Discounting was unnecessary since all the costs were incurred during less than one year. The price year was reported, which will facilitate future reflation exercises.
Other issues The authors presented their results in a comprehensive manner and their conclusions reflected the scope of their analysis. They did not compare their findings with other similar studies, nor did they discuss how their results could be generalised to other settings. The authors acknowledged that their model was limited to widely available diagnostic tests, and by uncertainty in the relationship between the number of antibiotics used and the effectiveness of coverage.
Implications of the study The authors did not make any direct recommendations for further research or changes to practice.
Bibliographic details Ost D E, Hall C S, Joseph G, Ginocchio C, Condon S, Kao E, Larusso M, Itzla R, Fein A M. Decision analysis of antibiotic and diagnostic strategies in ventilator-associated pneumonia. American Journal of Respiratory and Critical Care Medicine 2003; 168(9): 1060-1067 Indexing Status Subject indexing assigned by NLM MeSH Anti-Bacterial Agents /economics /therapeutic use; Bronchoalveolar Lavage /economics /methods; Bronchoalveolar Lavage Fluid /microbiology; Bronchoscopy /economics; Costs and Cost Analysis; Decision Support Techniques; Drug Therapy, Combination; Humans; Intubation, Intratracheal /economics; Pneumonia, Bacterial /diagnosis /economics /etiology /therapy; Respiration, Artificial /adverse effects; Software; Survival Analysis AccessionNumber 22003001449 Date bibliographic record published 31/05/2005 Date abstract record published 31/05/2005 |
|
|
|