Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Cost-effectiveness of antiseptic-impregnated central venous catheters for the prevention of catheter-related bloodstream infections

Veenstra DL, Saint S, Sullivan SD.

JAMA. 1999;282:554-60. [abstract] [full-text for a limited time]

Review posted September 8, 2000

1. Were all important strategies and outcomes included?

STRATEGIES
Clinical strategies compared were the use of chlorhexidine-silver sulfadiazine-impregnated catheters (antiseptic-impregnated catheters; AIC) with standard catheters.

OUTCOMES CONSIDERED

1. Catheter-related blood stream infection (CR-BSI) defined as the growth of the same organism from a peripheral culture and from a catheter tip.
2. Catheter colonization (CC) +/- evidence of local infection
3. No infectious complications from catheter insertion
4. Hypersensitivity reaction in the setting of AIC
5. Death

Outcomes were determined from the perspective of the health care payer.

2. Was an explicit and sensible process used to identify, select and combine the evidence into probabilities?

Decision analysis was largely based on the data accumulated in the meta-analysis conducted by the same authors entitled: "Efficacy of Antiseptic-Impregnated Central Venous Catheters in Preventing Catheter-Related Bloodstream Infection" (1). The meta-analysis reviews studies identified through MEDLINE database from January 1966 to January 1998 which were randomized trials comparing chlorhexidine-silver sulfadiazine-impregnated central venous catheters with standard central venous catheters. The outcomes assessed were CC and CR-BSI confirmed by catheter culture. Twelve studies were used in the analysis of catheter colonization and eleven were used in the analysis of CR-BSI.

PROBABILITY DETERMINATION

1. Probability of CR-BSI and CC were based on the meta-analysis of 13 randomized controlled trials. Risk ratios for both CR-BSI [0.58; 95%CI (0.4-0.85)] and CC [0.61; 95%CI (0.51-0.73)]were determined using Mantel-Haenszel methods.
2. Probabilities of CR-BSI and CC with standard catheters were determined by pooling the proportion of standard catheters associated with CR-BSI and CC respectively.
3. Probability of CR-BSI with AIC was determined by multiplying the risk ratio for CR-BSI with AIC times the probability of CR-BSI with standard catheters.
4. Probability of CC with AIC was determined by multiplying the risk ratio for CC with AIC times the probability of CC with standard catheters.
5. Probability of death attributable to CR-BSI was based on reports in the literature. [15% (range 5-25%)
6. Probability of hypersensitivity reaction was based on incidence reported in Japan of 11.1 cases per 100,000 AIC. [0.0111% (range 0.0056-0.0222)]
7. Probability of death from hypersensitivity based on 1 death in 13 cases of hypersensitivity reactions in Japan. [7.7% (range 3.9- 15.4)]

3. Were the utilities obtained in an explicit and sensible way from credible sources?

The analysis did not include calculation of utilities. Rather their outcome assessment was based on clinical outcome (incidence of CR-BSI, incidence of death attributable to CR-BSI, and or hypersensitivity reaction, and direct medical costs). The incremental value for each of these measures was determined by subtracting the result from the standard catheter by the result for the AIC. Values were calculated as follows:

1. Incidence of CR-BSI decrease: 5.2% to 3%; a relative decrease of 42%.
2. Expected incidence of death attributable to CR-BSI +/- hypersensitivity reaction decreased 0.78% to 0.45%; a relative decrease of 42%
3. Incidence of local infections decreased 12.4 to 7.5%.

The medical cost assessment warrants an evaluation based on the "Economic Analysis" outline found in the Users' Guides to the Medical Literature. Cost is one of the main outcomes determined in the clinical decision analysis of AIC's. In order to assess the validity of the report, one must answer the following questions.

3A. Are the results valid? 

a) Did the analysis provide a full economic comparison of the health care strategies?

Yes. The analysis compared both efficacy and costs of AIC's vs. standard catheters.

i. Was a broad enough viewpoint adopted?

The cost and outcomes were assessed from the viewpoint of the health care payer. When one considers costs for a given patient's hospital stay, the health care payer could actually be the patient, the insurance company, the hospital or the medicaid/medicare system. There was no consideration of the cost of productive days lost due to a prolonged hospital stay.

ii. Were all the relevant clinical strategies compared?

No consideration was given to antibiotic-impregnated catheters compared to antiseptic-impregnated catheters in this analysis. Preliminary data suggests the reduction of CR-BSI is even greater with certain antibiotic-impregnated catheters. Thus, the catheter type that may produce the greatest cost efficiency and outcome is that impregnated with antibiotics. More data is needed regarding efficacy of clinical outcome of antibiotic-impregnated catheters.

b) Were the costs and outcomes properly measured and valued?

i. Was clinical effectiveness established?

Yes. A significant decrease in the incidence of CR-BSI, incidence of death attributable to CR-BSI, and or hypersensitivity reaction was determined.

ii. Were costs measured accurately?

A simple assessment of charge to patient was employed based on prolonged hospitalization, cost of catheter and cost of potential outcomes of either strategy. No procedural or professional charges were considered. Reimbursement rates were not considered.

iii. Were data on costs and outcomes appropriately integrated?

The calculation of an incremental cost-effectiveness ratio was not conducted because the intervention is dominant with greater efficacy and lower costs.

c) Was appropriate allowance made for uncertainties in the analysis?

Sensitivity analysis determined that for the range of central values determined, AIC's still decrease overall medical costs.

 For example, sensitivity analysis determined that AIC's would be cost-effective as long as the following conditions existed:

1. attributable cost of CR-BSI was at least $687
2. relative risk of CR-BSI was at most 0.97
3. incidence of CR-BSI was at least 0.4%
4. additional catheter cost was at most $221

d) Are estimates of costs and outcomes related to the baseline risk in the treatment population?

Yes they are related. The parameters most influential on the cost- effectiveness of AIC's are the cost of CR-BSI, relative risk for CR-BSI and incidence of CR-BSI. The authors estimate that for the base-case population, the cost of CR-BSI is $9738 (range $4869- 19,476), the relative risk for CR-BSI 0.582 (range 0.398-0.851) and the incidence of CR-BSI is 5.2 % (range 3.9-6.5). These values are well outside the threshold parameters determined by the sensitivity analyses that would result in a negative cost-effectiveness ratio.

a. What were the incremental costs and outcomes of each strategy?

1. Additional cost of CR-BSI in previous report in 1990 was $33,268.
2. Additional cost of CR-BSI in current evaluation was $9738 (range $4869-19,476)
3. ICU stay is extended by 6 days and general ward by 6.5 days.
4. U Washington MC costs $1152 and $375 per ICU and ward day respectively.
5. No procedural, professional charges considered
6. Additional cost of AIC: $25
7. Cost of hypersensitivity reaction: $1192 (range $596-2384)
8. Cost of managing local infection: $210 (range $105-315)
9. NET COST SAVINGS WITH AIC VS STANDARD CATHETER: $196 per catheter

b. Do incremental costs and outcomes differ between subgroups?

Subgroups were not considered in the analysis.

c. How much does allowance for uncertainty change the results? Please see Question # II.3 below in decision analysis section.

a. Are the treatment benefits worth the harms and the costs?

Yes. The base-case analysis suggests that for every 300 AIC used, approximately $59,000 will be saved, 7 cases of CR-BSI avoided and one death prevented.

b. Could my patients expect similar outcomes?

While no adequate studies have been performed in children regarding reduction of CR-BSI with AIC's, the data reported seems to be applicable to my patients.

c. Could I expect similar costs?

While cost of care is probably similar between pediatric and adult patients, there are probably some differences. While there is no direct loss of productivity when children are hospitalized, parental duties may necessitate days of work lost. There is thus likely an indirect cost to society.

Returning now to the decision analysis questions....

4. Was the potential impact of any uncertainty in the evidence determined?

A multivariate sensitivity analysis using Monte Carlo simulation was performed to evaluate the impact of uncertainty in all model parameters. Threshold values were established for each parameter. Additionally, to test the robustness of the results, they set all parameters in the model to favor standard catheters in a worse-case scenario.

AIC's were associated with decreased costs and increased efficacy over the central range of values calculated in the multivariate sensitivity analysis. These findings held for the worse-case scenario. Sensitivity analysis also revealed the most influential individual parameters on the incremental cost of the catheter to be the cost of the CR-BSI [range $4869.00-19,476.00 with parameter threshold $687.00] the RR for CR-BSI [range 0.398 to 0.851 parameter threshold 0.970] and the incidence of CR-BSI [range 3.9- 6.5% with parameter threshold of 0.4%]. Finally, sensitivity analysis revealed those parameters most influential on the incremental incidence of death to be the probability of death attributable to AIC [range -0.54% to -0.11% with parameter threshold of 0.2%], the RR for CR-BSI [range -0.47 to -0.25% with parameter threshold of 0.99%] and the baseline risk of CR-BSI [range -0.41 to - 0.25% with parameter threshold of 0.07%].

1. In the baseline analysis, does one strategy result in a clinically important gain for patients? If not, is the result a "toss-up"?

The baseline analysis reveals AIC achieves both increased efficacy and decreased costs. AIC's decrease the incidence of CR-BSI and the incidence of death due to CR-BSI or hypersensitivity as well as decrease the overall medical costs associated with catheter use. This strategy clearly has clinical benefit.

2. How strong is the evidence used in the analysis?

The probabilities used in the analysis were primarily based on the meta-analysis previously mentioned. Eleven studies demonstrated an AIC related reduction in CR-BSI but only one demonstrated statistical significance. However, the summary effect found in the main analysis and sensitivity analyses in the meta-analysis demonstrated a 40% reduction in the risk of CR-BSI. The lack of statistical significance was thought to be due to lack of adequate power. Twelve studies examining CC decrease with AIC demonstrated statistically significant reduction on meta-analysis.

3. Could the uncertainty in the evidence change the result?

Sensitivity analysis demonstrates a decrease in medical care costs and decrease in catheter-related deaths and hypersensitivity reactions over a wide range of parameter values. Parameter thresholds were determined and are far outside the ranges considered viable for each parameter.

1. Do the probability estimates fit my patients' clinical features?

The patient population were patients at high risk for catheter related infections. One third were in the ICU, and in 2 studies the patients were those receiving TPN. The remaining patients were from a variety of hospital settings including some immunocompromised patients. The mean duration of treatment ranged 5.1 to 11.2 days. Our patient population in the ICU is similarly at high risk for catheter related infections. The biggest difference is the age of the patient population. Pediatric patients may be less likely to develop catheter related infection due to differences in bacterial colonization patterns and sites of catheter placement. A lower probability of infection may lower the degree of reduction in CR-BSI associated with AIC.

2. Do the utilities reflect how my patients would value the outcomes of the decision?

It is not possible to speculate how my patients would value the outcomes of the decision.

References:

1. Veenstra et al. Efficacy of antiseptic-impregnated central venous catheters in preventing catheter-related bloodstream infection: a meta-analysis. JAMA. 1999; 281:261-267. [abstract] [full-text for a limited time]
   • PedsCCM EB Journal Club Review, by M. K. Moynihan Hardart

Comments

Back to the EB Journal Club Index