AJIC: American Journal of Infection Control
Volume 31, Issue 1 , Pages 1-8, February 2003

Which antimicrobial impregnated central venous catheter should we use? Modeling the costs and outcomes of antimicrobial catheter use☆☆☆

Seattle, Washington, and Ann Arbor, Michigan

From the Pharmaceutical Outcomes Research and Policy Program, Department of Pharmacy, University of Washington, Seattlea; Veterans Administration Puget Sound Health Care System, Seattleb; Ann Arbor Veterans Administration Medical Centerc; Department of Internal Medicined and Patient Safety Enhancement Program,e University of Michigan, Ann Arbor

Article Outline

Abstract 

Background: Catheter-related bloodstream infections are costly and associated with substantial morbidity and mortality. Trials suggest that central venous catheters impregnated with minocycline/rifampin, although more expensive, are clinically superior to chlorhexidine/silver sulfadiazine impregnated catheters. It remains unclear whether minocycline/rifampin catheters are cost-effective for all high-risk patients or only those requiring longer-term catheterization. Methods: We developed a series of decision models with patient-level clinical trial data to determine whether minocycline/rifampin catheters are cost-effective for patients requiring various durations of catheterization. We calculated incremental cost-effectiveness ratios for patients catheterized for durations ranging from 1 to 25 days. Results: The data were too sparse to estimate cost-effectiveness for patients catheterized less than 8 days. The probability that minocycline/rifampin catheters were cost-effective compared with chlorhexidine/silver sulfadiazine catheters in patients catheterized for 8 days was 91%. The probability that the minocycline/rifampin catheters in patients catheterized 13 days or longer resulted in cost savings was more than 95%. Conclusions: Our analysis suggests that central venous catheters coated with minocycline/rifampin are cost-effective for patients catheterized for at least 1 week and lead to overall cost savings when patients are catheterized for 2 weeks or longer. Policies for the use of antimicrobial catheters in high-risk patients should reflect patients' expected duration of catheterization. (Am J Infect Control 2003;31:1-8.)

 

Catheter-related bloodstream infections (CR-BSI) are common, costly, and associated with morbidity and mortality.1, 2, 3, 4, 5, 6 Central venous catheters coated with either the antiseptics chlorhexidine and silver sulfadiazine or the antibiotics minocycline and rifampin reduce the risk of CR-BSI compared with standard, uncoated catheters.7, 8, 9 Infection control experts recommend the use of catheters coated with chlorhexidine/silver sulfadiazine or minocycline/rifampin when the risk of CR-BSI is high; however, there is little guidance on choosing between the 2 types of catheters.10, 11, 12

The results of a recent clinical trial suggest that minocycline/rifampin catheters reduce the risk of CR-BSI compared with chlorhexidine/silver sulfadiazine catheters.13 The trial, which randomized patients to catheterization with either minocycline/rifampin catheters or chlorhexidine/silver sulfadiazine catheters, tracked the incidence of catheter colonization and CR-BSI for patients catheterized up to 55 days. Kaplan-Meier plots of time to CR-BSI show that the risk of CR-BSI was similar between the groups during the first 10 days of catheterization, although more CR-BSIs occurred in the chlorhexidine/silver sulfadiazine catheter group. After 10 days, the Kaplan-Meier curves separate as CR-BSIs occur in the chlorhexidine/silver sulfadiazine arm of the study. During the course of the study, the minocycline/rifampin catheters reduced the risk of CR-BSI by 92%.

Although the efficacy of the minocycline/rifampin catheters observed during the course of the trial was impressive, chlorhexidine/silver sulfadiazine catheters may be preferable for patients expected to be catheterized for a short duration (eg, less than 10 days) because of their similar efficacy and lower cost.10, 13, 14, 15, 16 To clarify clinical recommendations, we developed a series of decision models with data from the trial comparing the catheters13 to determine whether minocycline/rifampin catheters are cost-effective for all patients at high-risk of CR-BSI or just those expected to require longer-term catheterization.

Back to Article Outline

Methods 

Model 

Like the authors of a prior catheter cost-effectiveness study,9 we used decision models to compare the outcomes of patients catheterized with minocycline/rifampin catheters with those of patients catheterized with chlorhexidine/silver sulfadiazine catheters and the associated costs for each group.17, 18 We built 25 models representing patients catheterized for different durations: 1 to 25 days (Fig 1).

We choose externally coated chlorhexidine/silver sulfadiazine catheters instead of uncoated catheters as the comparator because infection control experts recommend the use of a coated catheter in high-risk patients.10, 11, 12 Chlorhexidine/silver sulfadiazine catheters coated on both the internal and external surfaces have recently received Federal Drug Administration approval, but clinical trial data evaluating the new catheter's efficacy and safety are not yet available. The study perspective was that of a health care payer, and the time horizon was the lifetime of the patient. Definitions for catheter colonization and CR-BSI reflected those used in the trial comparing the 2 coated catheters.13

Patient population 

We obtained patient-level data from the trial investigators who compared the central venous catheters coated with minocycline/rifampin and chlorhexidine/silver sulfadiazine.13 Thus, the characteristics of our modeled patients were the same as those of the 689 evaluable patients enrolled in the trial. Trial subjects were hospitalized adults at high risk of developing a CR-BSI who were likely to require a triple-lumen, noncuffed central venous catheter for at least 3 days. High-risk patients included those in intensive care units and those who were immunocompromised. The average age of the population was 56 years.

Baseline risk of events 

In the decision models, a catheterized patient was at risk for 1 of the 5 following events: remain catheterized with no sequelae; have the catheter removed with no colonization, local infection, or CR-BSI; have a colonized catheter removed with no evidence of local infection or CR-BSI; have a colonized catheter removed with evidence of a local infection but no CR-BSI; or have a catheter removed with CR-BSI (see Fig 1). Because patients catheterized with chlorhexidine/silver sulfadiazine catheters were the reference group, we used their data to represent the baseline risk of an event. For each duration of catheterization (1-25 days), we used the smoothed, cumulative number of events in the chlorhexidine/silver sulfadiazine arm of the trial to calculate the probability that the patients with chlorhexidine/silver sulfadiazine catheters had an event.19 For each duration of catheterization, the probability of a patient remaining catheterized was 1 minus the sum of the probabilities of the other events, since the other events involved catheter removal. The risk of death after a CR-BSI reported in the trial was 14%.13 We assumed that 5% of colonized catheters (in the absence of a CR-BSI) would be associated with signs of local infection (Table 1).20

Table 1. Parameters and costs used in the decision models
Reference caseMR-favoredCSS-favoredReference
Parameters*
Local infection given colonization0.050.400.0020
Death due to CR-BSI0.140.240.043, 4, 5, 6, 13
Life expectancy (y)1520109, 21, 22, 23
Utility0.800.900.7024, 25, 26
Costs, 2000 ($)
Additional cost of MR catheter9.664.8319.3213
CR-BSI10,45220,90452266, 9
Cost of managing a local infection2253381139
*Estimates of the risk and relative risk of CR-BSI, catheter colonization, and catheter removal varied over time, whereas other parameters were constant over time. Estimates that varied over time are available on request from the author.

MR , Minocycline and rifampin impregnated central venous catheters; CSS , chlorhexidine and silver sulfadiazine impregnated central venous catheters; CR-BSI , catheter-related bloodstream infection.

Risk of events in the patients with minocycline/rifampin catheters 

We calculated the probability of an event for patients with minocycline/rifampin catheters by multiplying the risk of events in the patients with chlorhexidine/silver sulfadiazine catheters (baseline risk) by the relative risk of events (risk of an event in the minocycline/rifampin group relative to the risk of an event in the chlorhexidine/silver sulfadiazine group). We calculated the relative risk of events and the corresponding confidence interval at each duration of catheterization with use of the cumulative number of events in the trial.13 When the number of events in either catheter group was 0 (zero), we had insufficient information to calculate relative risks or confidence intervals.

Quality adjusted life years (QALYs) 

We estimated the life expectancy of study patients not dying of CR-BSI with use of survival data from a study of intensive care patients with demographic and clinical features similar to those of our study population.21 We used the survival data to calculate mortality rates for the first 2 years after catheterization by subtracting CR-BSI–attributable deaths from the total number of deaths, assuming 60%22 of the patients would have a central venous catheter, 5% would develop a CR-BSI, and 15%9 would die as a result of the CR-BSI. We then applied these rates to our models, assuming that the mortality rates would return to population levels after 2 years.21, 23 Thus, the average life expectancy of patients requiring catheterization and not dying of CR-BSI was 15 years, assuming 27% of patients would die within 1 year of catheterization, 5% would die within 2 years, and 67% would live an additional 22 years.21, 23 We ranged the average life expectancy from 10 to 20 years in sensitivity analyses.

Because many patients requiring central venous access have underlying diseases that impair their functioning and well-being, we adjusted our life expectancy estimate to account for reduced quality of life. Unfortunately, quality-of-life estimates specific to patients requiring central venous access were not available. However, because much of our study population suffered from cardiovascular disease or cancer,13 we used the health states of lymphoma, leukemia, and congestive heart failure, for which values were available, to obtain quality-of-life estimates. The estimates ranged from 0.35 to 0.98, but the majority fell in the range of 0.70 to 0.90.24, 25, 26 We choose 0.80 to represent the average quality of life of patients after catheterization and varied the value from 0.70 to 0.90 in sensitivity analyses. We attenuated average life expectancy with the quality estimate and discounted the result at a rate of 3% annually, as recommended by the US Panel on Cost-Effectiveness Analysis.27

Costs 

We limited costs to those incurred during the period of hospitalization (see Table 1). We obtained the incremental cost of the minocycline/rifampin catheter from published data.13 Estimates of the cost of CR-BSI and local infection have been described previously.9 We updated health care payer costs to 2000 US dollars with use of the Consumer Price Index specific to medical care.28

Outcome assessment and sensitivity analyses 

We calculated incremental cost-effectiveness ratios (difference in costs/difference in QALYs) for patients catheterized from 1 to 25 days to determine the length of catheterization required for minocycline/rifampin catheters to be cost-effective. For the “reference case” analysis, we estimated costs and outcomes with the point estimates that were the most probable. In the reference case, we defined minocycline/rifampin catheters as “cost-effective” when the incremental cost-effectiveness ratio was below $50,000/QALY.29 We defined minocycline/rifampin catheters as “cost-saving” when they simultaneously reduced costs and increased QALYs.

We used a variety of sensitivity analyses to evaluate the effect of our assumptions on the results. To assess the overall uncertainty in the results, we used Monte Carlo simulations to perform a multivariate sensitivity analysis with the ranges of values shown in Table 1. This technique is described in detail elsewhere.30, 31 We assumed that probabilities followed a logistic-normal distribution. Relative risks and costs were generally assumed to follow a gamma distribution because of the skewed nature of their confidence intervals.31 Because there was no evidence to suggest otherwise, we assumed that the cost of the minocycline/rifampin catheters and life expectancy followed a normal distribution. The confidence intervals around the CR-BSI relative risk estimates were severely right-skewed, so we log transformed and modeled them with use of a normal distribution.

In the multivariate sensitivity analysis, we simulated the use of 10,000 catheters for each duration of catheterization and recorded the proportion of results that were cost-effective or cost-saving. When interpreting the multivariate sensitivity analysis results, we called the minocycline/rifampin catheters “cost-effective” when the incremental cost-effectiveness ratios were either less than $50,000/QALY or cost-saving.29 We defined minocycline/rifampin catheters as “cost-saving” when they both reduced costs and increased QALYs.

We also performed 1-way sensitivity analyses and scenario sensitivity analyses to check the robustness of our findings. In the scenario sensitivity analyses, we set all model estimates to favor the minocycline/rifampin catheters (minocycline/rifampin–favored) and then all estimates to favor the chlorhexidine/silver sulfadiazine catheters (chlorhexidine/silver sulfadiazine–favored).

Main assumptions in the analysis 

Several assumptions underlie the reference case analysis. First, the risk of death after CR-BSI is the same for the 2 catheters. Second, the cost of CR-BSI is the same regardless of the survival outcome. Third, there are no costs or adverse outcomes associated with catheter colonization (without local infection) or catheter removal in the absence of CR-BSI. Fourth, the survival experience of high-risk patients requiring central venous catheterization is similar to the survival of patients hospitalized in an intensive care unit. Finally, the primary benefit of minocycline/rifampin catheters is the reduced risk of local infection and CR-BSI; any other complications (eg, mechanical, hypersensitivity reactions, or antimicrobial resistance) are equally likely for the minocycline/rifampin and chlorhexidine/silver sulfadiazine catheters.

Back to Article Outline

Results 

Because no cases of CR-BSI occurred among patients randomized to the minocycline/rifampin catheters for the first 7 days of catheterization,13 we were unable to estimate the relative risk of CR-BSI and the cost-effectiveness of the minocycline/rifampin catheters for patients expected to be catheterized less than 8 days.

Among patients expected to be catheterized more than 1 week, the clinical and economic benefits of the minocycline/rifampin catheters increased with the duration of catheterization. For patients catheterized 8 days, minocycline/rifampin catheters were clinically and economically superior to chlorhexidine/silver sulfadiazine catheters, reducing CR-BSI by 0.7%, increasing QALYs by 0.009, and reducing costs by $67 in the reference case (Table 2).

Table 2. Reference case results: the difference in cumulative costs and quality adjusted life years (QALYs) for varying lengths of catheterization*
Expected duration of catheterizationCosts ($) (reference case, 95% confidence range)QALYs (reference case, 95% confidence range)Cost/QALY (reference case)
1 week or less†———
8 days−67 (88, −165)0.009 (−0.009, 0.016)Cost-saving‡
15 days−260 (−29, −580)0.03 (0.007, 0.053)Cost-saving‡
22 days−294 (−40, −654)0.034 (0.009, 0.059)Cost-saving‡
*Negative costs indicate that the minocycline/rifampin (MR) catheters were cost-saving, and positive QALYs indicate that the MR catheters were more effective. †The data were insufficient to assess cost-effectiveness or cost-savings of MR catheters in patients expected to be catheterized 1 week or less. ‡Ratios are not meaningful when an intervention is cost-saving (eg, the intervention costs less, yet is more effective).
Accounting for uncertainty in the multivariate sensitivity analysis, the probability that minocycline/rifampin catheters were cost-effective in patients catheterized for 8 days was 91% (Fig 2).
  • View full-size image.
  • Fig. 2. 

    Probability that the minocycline and rifampin catheters are cost-effective (ie, less than $50,000/quality-adjusted life-year threshold or cost-saving) for patients expected to be catheterized for varying lengths of time.(Data were too sparse to estimate the probability of cost-effectiveness for patients catheterized less than 8 days.)

For patients catheterized 13 days, minocycline/rifampin catheters reduced CR-BSI incidence by 2.0%, increased QALYs by 0.024, and reduced costs by $203 in the reference case. Accounting for uncertainty in the multivariate sensitivity analysis, the probability that the minocycline/rifampin catheters were cost-saving for patients catheterized 13 or more days was 97.4%.

The effect of varying individual model estimates within their specified ranges while holding all others constant was small. The only parameter having a strong individual effect on the results was the relative risk of CR-BSI. Replacing the point estimates of CR-BSI relative risks with the upper confidence limits increased the duration of catheterization at which minocycline/rifampin catheters became cost-saving from 8 to 13 days. When individually varying the remaining risks and costs within their likely ranges, minocycline/rifampin catheters remained cost-saving for patients catheterized longer than 1 week.

The use of minocycline/rifampin–favored estimates did not change the duration of catheterization at which the minocycline/rifampin catheters became cost-saving (8 days), but the cost savings increased from the reference case estimate of $67 to $222. When the assumptions most favorable to chlorhexidine/silver sulfadiazine catheters were used, the minocycline/rifampin catheters were not cost-effective nor cost-saving until used in patients catheterized for more than 2 weeks (15 days).

Back to Article Outline

Discussion 

We compared the clinical and economic consequences of the use of central venous catheters coated with minocycline/rifampin versus chlorhexidine/silver sulfadiazine in patients at high risk for CR-BSI by decision analytic techniques. Our analysis demonstrated that the probability of cost-effectiveness and cost-savings depended on the duration patients were expected to be catheterized. Because of limited data, we were unable to evaluate whether the minocycline/rifampin catheters were cost-effective for patients requiring short-term (eg, 7 days or less) catheterization. Accounting for uncertainty, minocycline/rifampin catheters were cost-effective at least 91% of the time for patients catheterized more than 7 days and were cost-saving 97.4% of the time for patients catheterized 13 days or more.

Elements of our analysis differ from previous cost analyses in several ways.10, 32 First, our analysis directly compares minocycline/rifampin versus chlorhexidine/silver sulfadiazine catheters. Second, we estimated the risks of CR-BSI and death with use of patient-level data from a randomized clinical trial.13 Third, we estimated cost-effectiveness for patients expected to require catheterization for different lengths of time. Fourth, we attributed a small cost to managing local catheter site infections in the absence of CR-BSI. Finally, we evaluated the uncertainty of our findings with a wide range of parameter estimates in the sensitivity analyses.

The trial data that forms the basis of our analysis did not enroll enough patients to compare the efficacy of the antimicrobial catheters in patients catheterized less than 8 days.13 Thus, we were unable to compare the cost-effectiveness of the catheters for patients requiring short-term catheterization. However, published time-to-event data for both types of antimicrobial catheters suggest that they reduce CR-BSI compared with standard, uncoated catheters in patients catheterized 1 week or less.7, 33 Including only the cost of the catheters and the cost and the risk (estimated from Kaplan-Meier plots) of CR-BSI, both antimicrobial catheters save costs relative to uncoated catheters for short-term catheterization. These results are uncertain, however, because CR-BSI before day 8 is relatively rare.

We assumed that patients with CR-BSI would suffer an attributable mortality of 14% on the basis of the deaths observed in the trial.13 Prior cost-effectiveness studies have assumed a similar risk,9 but recent analyses adjusting for severity of illness before infection suggest that mortality attributable to CR-BSI may be less.34, 35, 36 To reflect our uncertainty about the attributable mortality of CR-BSI, we used a wide range of estimates in sensitivity analyses; within the range of attributable mortality risks examined (4%-24%), our findings were robust.

The use of clinical estimates obtained from trial efficacy data (in contrast to effectiveness data, which would more closely approximate the risks in general practice) limits the generalizability of our findings. As expected in a clinical trial designed to assess efficacy, an intense, strict level of care was dictated by the study protocol. The trial prohibited recatheterization with use of a guidewire, thereby limiting insertion of study catheters to new venipunctures.13 The protocol dictated daily patient evaluation, with thrice-weekly dressing changes and disinfection with 10% povidone-iodine. Deviations from these procedures, which may be common in the generalized patient care setting, may affect the findings.

Our findings are also dependent on the validity of the outcomes measured in the trial. The clinical and economic benefits of the minocycline/rifampin catheters would be overstated if, as suggested by some clinicians, the efficacy estimates from the trial were inflated by residual antimicrobial activity in the minocycline/rifampin catheters in culture.37, 38

Our models do not account for potential differences in risk of catheter complications. Shortly after the introduction of the chlorhexidine/silver sulfadiazine catheters in Japan, 13 hypersensitivity reactions were noted among the 117,000 chlorhexidine/silver sulfadiazine catheters sold before their use was halted.39, 40, 41 To our knowledge, no cases of hypersensitivity reactions have been reported in the United States for either chlorhexidine/silver sulfadiazine or minocycline/rifampin catheters. Including hypersensitivity reactions as a side effect of the chlorhexidine/silver sulfadiazine catheters in our models has a negligible effect on the results. Assuming the risk and cost of hypersensitivity reactions are 0.011% (13/117,000) and $1192,9 respectively, the incremental cost increases by only $0.13.

We also did not account for potential mechanical complications that may be associated with minocycline/rifampin catheters. Anecdotally, the guidewires provided with the catheter tray are stiff, occasionally preventing successful insertion. This complication contributed to the decision of one university medical center to discontinue use of the minocycline/rifampin catheters (personal communication, TM Hooton, MD, September 11, 2000). Accounting for insertion complications would reduce the cost-effectiveness of the minocycline/rifampin catheter. The impact on the results would depend on the frequency with which the catheter insertion failed and required replacement.

Antibiotic resistance is an issue of concern for the minocycline/rifampin catheters because minocycline and rifampin are both used for systemic antibiotic therapy. Although a clinical trial failed to find evidence of antimicrobial resistance or detectable systemic levels of minocycline or rifampin after use of the minocycline/rifampin catheters,42 colonization and infection with resistant organisms would likely have poor clinical and economic outcomes.

We analyzed the cost-effectiveness of the antimicrobial catheters from the perspective of a health care payer, so our models only included the direct costs of catheter-related outcomes; we excluded the indirect costs that would be part of a cost-effectiveness analysis conducted from the societal perspective. Nevertheless, our findings are conservative because including indirect costs would increase the cost-savings of the minocycline/rifampin catheters.

The chlorhexidine/silver sulfadiazine catheters serving as the comparator in our cost-effectiveness analysis were only coated on the external surface. Chlorhexidine/silver sulfadiazine catheters coated on both the internal and external surfaces have recently received Federal Drug Administration approval, but clinical trial data evaluating the new catheter's efficacy and safety are not yet available. Publication of studies comparing double-coated chlorhexidine/silver sulfadiazine versus minocycline/rifampin catheters may take several years. In the meantime, our analysis can aid clinicians and hospital decision-makers to formulate clinical policies for the externally coated chlorhexidine/silver sulfadiazine and minocycline/rifampin catheters.

A final limitation of our analysis is one shared by most decision-modeling approaches—the expectation that the synthesis of disparate data elements will produce logical and consistent results. Even though much of our analysis derived model estimates directly from the randomized trial comparing the 2 catheters,13 we relied on additional sources to estimate cost parameters,6, 9 utility estimates,24, 25, 26 life expectancy,9, 21, 22, 23 and the risks of mortality and local infection.3, 4, 5, 6, 20 Because combining different sources of information increases the uncertainty of the findings, we evaluated a wide range of estimates for each parameter in the sensitivity analyses. The robustness of our findings implies that the conclusions are credible under widely varied assumptions and thus useful to inform decision-making.

The probability that minocycline/rifampin catheters are cost-effective or cost-saving increases with the expected duration of use. Our analysis suggests that central venous catheters coated with minocycline/rifampin are cost-effective for patients expected to be catheterized for at least 1 week and lead to overall cost savings in patients expected to be catheterized for 2 weeks or longer. Policies for the use of antimicrobial catheters in high-risk patients should reflect patients' expected duration of catheterization.

Back to Article Outline

Acknowledgements 

We thank Rabih Darouiche, MD, for sharing his data; Leonard Mermel, DO, and Thomas M. Hooton, MD, for sharing their clinical expertise; and the Research Foundation for the Prevention of Complications Associated with Health Care for funding the study.

Back to Article Outline

References 

  1. Raad II , Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect Dis. 1992;15:197–208
  2. Maki DG. Infections caused by intravascular devices used for infusion therapy. In: 2nd ed.  Bistro AI,  Waldvogel FA editor. Infections associated with indwelling medical devices. Washington (DC): ASM Press; 1994;p. 155–205
  3. Martin MA, Pfaller MA, Wenzel RP. Coagulase-negative staphylococcal bacteremia. Mortality and hospital stay. Ann Intern Med. 1989;110:9–16
  4. Arnow PM, Quimosing EM, Beach M. Consequences of intravascular catheter sepsis. Clin Infect Dis. 1993;16:778–784
  5. Smith RL, Meixler SM, Simberkoff MS. Excess mortality in critically ill patients with nosocomial bloodstream infections. Chest. 1991;100:164–167
  6. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1994;271:1598–1601
  7. Raad I, Darouiche R, Dupuis J, Abi-Said D, Gabrielli A, Hachem R, et al.  Central venous catheters coated with minocycline and rifampin for the prevention of catheter-related colonization and bloodstream infections. A randomized, double-blind trial. The Texas Medical Center Catheter Study Group Ann Intern Med. 1997;127:267–274
  8. Veenstra DL, Saint S, Saha S, Lumley T, Sullivan SD. Efficacy of antiseptic-impregnated central venous catheters in preventing catheter-related bloodstream infection: a meta-analysis. JAMA. 1999;281:261–267
  9. Veenstra DL, Saint S, Sullivan SD. Cost-effectiveness of antiseptic-impregnated central venous catheters for the prevention of catheter-related bloodstream infection. JAMA. 1999;282:554–560
  10. Wenzel RP, Edmond MB. The evolving technology of venous access. N Engl J Med. 1999;340:48–50
  11. Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med. 2000;132:391–402
  12. Guideline for prevention of intravascular device-related infections . Part II. Recommendations for the prevention of nosocomial intravascular device-related infections. Hospital Infection Control Practices Advisory Committee. Am J Infect Control. 1996;24:277–293
  13. Darouiche RO, Raad II , Heard SO, Thornby JI, Wenker OC, Gabrielli A, et al.  A comparison of two antimicrobial-impregnated central venous catheters. N Engl J Med. 1999;340:1–8
  14. dos Santos CC, Stewart TE, McGeer A. Anti-microbial or antiseptic impregnated central venous catheters: update on a “central” debate. Intensive Care Med. 2000;26:246–249
  15. Saint S, Veenstra DL, Lipsky BA. The clinical and economic consequences of nosocomial central venous catheter-related infection: are antimicrobial catheters useful?. Infect Control Hosp Epidemiol. 2000;21:375–380
  16. Spencer RC. Novel methods for the prevention of infection of intravascular devices. J Hosp Infect. 1999;43(suppl):S127–S135
  17. Detsky AS, Naglie G, Krahn MD, Naimark D, Redelmeier DA. Primer on medical decision analysis: part 1—getting started. Med Decis Making. 1997;17:123–125
  18. Detsky AS, Naglie G, Krahn MD, Redelmeier DA, Naimark D. Primer on medical decision analysis: part 2—building a tree. Med Decis Making. 1997;17:126–135
  19. Greenland S. Dose-response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology. 1995;6:356–365
  20. Tacconelli E, Tumbarello M, Pittiruti M, Leone F, Lucia MB, Cauda R, et al.  Central venous catheter-related sepsis in a cohort of 366 hospitalised patients. Eur J Clin Microbiol Infect Dis. 1997;16:203–209
  21. Niskanen M, Kari A, Halonen P, the Finnish ICU Study Group . Five-year survival after intensive care—comparison of 12,180 patients with the general population. Crit Care Med. 1996;24:1962–1967
  22. NNIS System . National Nosocomial Infections Surveillance (NNIS) System Report, data summary from October 1986-April 1998. Atlanta (Ga): Hospital Infections Program, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services; June 1998;
  23. Centers for Disease Control and Prevention . National vital statistics report. Available at http://www.cdc.gov/nchs/data/nvsr/nvsr47/nvs47_28.pdf1999;
  24. Norum J, Angelsen V, Wist E, Olsen JA. Treatment costs in Hodgkin's disease: a cost-utility analysis. Eur J Cancer. 1996;32A:1510–1517
  25. Kavanagh T, Myers MG, Baigrie RS, Mertens DJ, Sawyer P, Shephard RJ. Quality of life and cardiorespiratory function in chronic heart failure: effects of 12 months' aerobic training. Heart. 1996;76:42–49
  26. Tengs TO, Wallace A. One thousand health-related quality-of-life estimates. Med Care. 2000;38:583–637
  27. Gold MR, Siegal JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996;
  28. Bureau of Labor Statistics. Consumer price index 2000. Available at: http://146.142.4.24/cgi-bin/surveymost
  29. Garber AM, Phelps CE. Economic foundations of cost-effectiveness analysis. J Health Econ. 1997;16:1–31
  30. Krahn MD, Naglie G, Naimark D, Redelmeier DA, Detsky AS. Primer on medical decision analysis: part 4—analyzing the model and interpreting the results. Med Decis Making. 1997;17:142–151
  31. Doubilet P, Begg CB, Weinstein MC, Braun P, McNeil BJ. Probabilistic sensitivity analysis using Monte Carlo simulation. A practical approach. Med Decis Making. 1985;5:157–177
  32. Cook Critical Care . Catheter related septicemia: magnitude of the problem. Available at http://www.grii.cpm/cook_critical_care/product_news/spectrum.html2000;
  33. Maki DG, Stoltz SM, Wheeler S, Mermel LA. Prevention of central venous catheter-related bloodstream infection by use of an antiseptic-impregnated catheter. Ann Intern Med. 1997;127:257–266
  34. Renaud B, Brun-Buisson C. Outcomes of primary and catheter-related bacteremia. A cohort and case-control study in critically ill patients. Am J Respir Crit Care Med. 2001;163:1584–1590
  35. Rello J, Ochagavia A, Sabanes E, Roque M, Mariscal D, Reynaga E, et al.  Evaluation of outcome of intravenous catheter-related infections in critically ill patients. Am J Respir Crit Care Med. 2000;162:1027–1030
  36. Soufir L, Timsit JF, Mahe C, Carlet J, Regnier B, Chevret S. Attributable morbidity and mortality of catheter-related septicemia in critically ill patients: a matched, risk-adjusted, cohort study. Infect Control Hosp Epidemiol. 1999;20:396–401
  37. Paterson DL. Antimicrobial-impregnated central venous catheters. N Engl J Med. 1999;340:1761
  38. Bach A. Antimicrobial-impregnated central venous catheters. N Engl J Med. 1999;340:1761
  39. Potential hypersensitivity reactions to chlorhexidine-impregnated medical devices. Center for Devices and Radiological Health Available at http://www.fda.gov/cdrh/chlorhex.html2000;
  40. World Health Organization . Central venous catheters (Arrowgard) recalled: anaphylactic shock. Information exchange system alert No. 62. Geneva: Switzerland: WHO; 1997;
  41. Yasukawa T, Fujita Y, Sari A. Antimicrobial-impregnated central venous catheters. N Engl J Med. 1999;340:1762
  42. Raad I, Darouiche R, Hachem R, Mansouri M, Bodey GP. The broad-spectrum activity and efficacy of catheters coated with minocycline and rifampin. J Infect Dis. 1996;173:418–424

☆ Dr Saint was supported by a Career Development Award from the VA Health Services Research and Development Service during the time much of this work was completed.

☆☆ Reprint requests: David L. Veenstra, PharmD, PhD, Pharmaceutical Outcomes Research and Policy Program, Department of Pharmacy, University of Washington, Box 357630, Seattle, WA 98195.

PII: S0196-6553(02)48235-1

doi:10.1067/mic.2003.35

AJIC: American Journal of Infection Control
Volume 31, Issue 1 , Pages 1-8, February 2003

Article Tools

* Image Usage & Resolution