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

An economic evaluation of activated protein C treatment for severe sepsis.

Manns BJ, Lee H, Doig CJ, Johnson D, Donaldson C.

N Engl J Med 2002;347:993-1000. [abstract]

Reviewed by Ayse A Arikan MD, Pediatric Critical Care Section Baylor College of Medicine

Review posted March 30, 2004

I. What is being studied?:

The study objective:

To estimate the cost-effectiveness of recombinant activated protein C (rhAPC) treatment for patients admitted to the ICU with severe sepsis by comparing conventional treatment to rhAPC use.

The study design:

Cost-effectiveness analysis. Data from a retrospective cohort study was collected to estimate the subsequent mortality after discharge and direct health related costs in sepsis survivors treated with conventional care.

The patients included:

For the retrospective cohort study, 819 patients with suspected or known infection from three tertiary care hospitals in The Calgary Health Region in Canada over a period of three years were eligible for the study. Of these, 787 patients who had a valid health identification number enabling follow-up were included. To ensure that their patients fulfilled eligibility criteria for the PROWESS study (and would therefore been eligible to receive rhAPC)(3), a manual review of 40 randomly selected charts was also done.

The patients excluded:

22 patients who did not have valid ID numbers were excluded from the retrospective cohort study.

The outcomes assessed:

Incremental cost per life year gained, and incremental cost per quality-adjusted life-year gained with rhAPC compared to conventional treatment.

II. Are the results valid?

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

A. Costs and outcomes for each strategy?

Yes. The average weekly cost of conventional care was estimated from the database. The authors did state that they calculated the cost for conventional treatment for each patient; however, this figure is not reported in the manuscript.

For indirect cost estimation, the authors have used a study that demonstrated decreased employment after ICU discharge (17% and 0% for ages younger and older than 61 years, respectively) and have used this figure as a surrogate measure to account for the lost production in sepsis survivors (1). (Indirect costs estimated as 0.17 x average gross annual salary) Outcomes for each strategy were determined individually; baseline risk of in-hospital and post-discharge mortality was derived from the database, the survival benefit of rhAPC treatment was taken from the PROWESS study. Additional costs for the adverse outcomes resulting from major complications (e.g. treatment of serious GI bleed in the ICU) associated with instituting the treatment were also considered in the total cost of treatment.

B. From whose viewpoint?

For the baseline analysis, the perspective of the purchaser of health-care services was used. For the sensitivity analysis, the perspective was broadened to society.

This choice was very appropriate since increased health care consumption by sepsis survivors after discharge from the hospital can be an important factor for the community.

The individual patient's viewpoint was also incorporated into the study by performing a cost-utility analysis where a composite index was used to adjust for quality of life .

C. Cost-effectiveness/benefit/utility study?

The baseline analysis is a cost-effectiveness study. One of the sensitivity analyses is a cost utility analysis that takes a composite index utility score of 0.6 (derived from the estimate of the post hospital discharge quality of life of ARDS patients (2)) incorporated into outcome.

2. Were the costs and outcomes properly measured and valued?

A. Was clinical effectiveness established?

Yes. The intervention in question - rhAPC - was shown to decrease relative risk for mortality in adults with sepsis by 19.4% and absolute risk of death by 6% in the PROWESS study and is FDA approved for use in adults with severe sepsis. There is controversy over whether rhAPC is effective in the group of septic patients with APACHE scores < 24. The FDA has mandated a large study of these patients as a post-hoc subgroup analysis showed no benefit in this population.

B. Were costs measured accurately?

To estimate the costs for the conventional care cohort, the mean daily hospital cost and physicians charges derived from the Calgary Health Region database was used. For the cost of health care after hospital discharge, the sum of the annual cost of hospitalizations, emergency room visits, and day-surgery within the above health region in addition to physician claims was used. Then, the cost of rhAPC ($6,800 per therapeutic course) was added to the cost of conventional care calculated as above. Since there was a 1.5% increased risk of serious bleeding associated with activated protein C use reported in the PROWESS study, the authors used the cost of treating gastrointestinal bleeding in the ICU ($8,306/bleeding episode) as a surrogate for estimating the cost of managing clinically important bleeding in the ICU and have thus added $122 to the cost per patient treated with rhAPC.

C. Were data on costs and outcomes appropriately integrated?

Yes. The intervention in question was compared to current care for treatment of sepsis. Since withholding treatment from patients with presumed severe sepsis would be unethical, inclusion of a no intervention or placebo arm in the study would not be appropriate.

3. Was appropriate allowance made for uncertainties in the analysis?

Several sensitivity analyses were performed including a cost-utility analysis that takes the quality of life subsequent to ICU discharge into account. The relative risk of death was varied within the 95% confidence interval reported in the original PROWESS trial, the estimates of in-hospital and subsequent death rates were varied by 25%. To account for the higher hospital costs expected in the US, the authors also carried out another sensitivity analysis where estimates of cost of health care were increased by 50%.

Monte Carlo simulation was also performed which allows simultaneous sensitivity analysis of all variables by converting all variables into probabilities and calculating the cost-effectiveness acceptability curve of rhAPC. According to the results, there was an "86% probability that rhAPC use in sepsis would be cost-effective if one were willing to pay $50,000 per quality-adjusted life-year gained."

4. Are estimates of costs and outcomes related to the baseline risk in the treatment population?

Yes for outcome. It is unclear, however, if those who have lower APACHE scores also have lower ICU costs and complications. Although overall costs were varied through their range across the population, no interaction between baseline risk and costs was performed.

It is unclear how accurately the risks and treatment effects extrapolated from the PROWESS study - which had numerous exclusion criteria - apply to this study's cohort.

Subgroup analysis based on stratification for age and severity of illness by admission APACHE II scores was carried out.

III. What are the results?

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

The incremental cost per life year gained with rhAPC was $27,936.

2. Do incremental costs and outcomes differ between subgroups?

Yes, cost effectiveness was sensitive to the age of the patients and the severity of illness. When the figures from the PROWESS study was used, which assumed that rhAPC is equally effective in the group with low APACHE scores, the cost for treating patients with APACHE II ≤ 24 was $35,632. After incorporating the estimated relative risk of death by APACHE score category based upon the results of the post-hoc analysis carried out by the FDA, however, the cost per life-year gained was $19,723 for patients with APACHE II scores ≥ 25; and $575,054 for those with APACHE II ≤ 24.

Even though the cost per life-year gained seemed to be higher for treating patients who were younger than 40, this difference was mainly arose from the less severe disease in this group. Among the patients with APACHE II ≥ 25, the cost-effectiveness decreased with increasing age due to decrease in expected survival.

Analysis was not sensitive to indirect costs.

3. How much does allowance for uncertainty change the results?

The analysis was not sensitive to various analyses except for the variation in the estimate of effectiveness of treatment in the APACHE subgroups.

IV. Will the results help me in caring for my patients?

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

It is clear that treatment with rhAPC is cost-effective for adult patients with severe sepsis and APACHE scores above 24. It is unclear whether rhAPC should be used in adult patients with APACHE scores at or below 24. This is because the post-hoc APACHE subgroup analysis by the FDA called into question the efficacy of rhAPC in the low APACHE subgroup. Usually, a cost per quality adjusted life year of US $100,000 or less is considered acceptable by healthcare decision makers. All of the costs per quality adjusted life year gained were well below this cutoff throughout the broad range of the sensitivity analysis with the exception of the low-APACHE subgroup when using the FDA estimates of efficacy in this subgroup.

2. Could my patients expect similar health outcomes? Are the study patients similar to my own? Is the study clinical management similar to my local practice?

No. The mortality of severe sepsis in the pediatric population is much lower (<10% compared to 30-50%) than the adults. The benefit of rhAPC use in decreasing morbidity and length of stay in the pediatric sepsis patients is currently being studied. We will need to know more about the usefulness of rhAPC in our patient population before deciding on cost-effectiveness. Since the mortality of pediatric sepsis is lower, a different outcome measure will need to be chosen for determining benefits in the pediatric population.

Yes. Even though this study was conducted in Canada, Canadian practice is quite similar to US practice and the sensitivity analysis performed by altering the cost estimates should incorporate the difference between Canadian and US costs.

3. Could I expect similar costs?

It is reasonable to speculate that the ICU and hospital costs for treating children are not very different than treating adults. In addition, although children are not part of the work-force, there still might be some indirect costs to society from parental absence from work.

Provided activated protein C were to be effective in children, it might be expected to be more cost-effective in the population of children at high risk of mortality secondary to longer survival expected in this age group.

References

1. Munn J, Willatts SM, Tooley MA. Health and activity after intensive care. Anaesthesia. 1995 Dec;50(12):1017-21 [abstract]
2. Angus DC, Musthafa AA, Clermont G, et al. Quality-adjusted survival in the first year after the acute respiratory distress syndrome. Am J Respir Crit Care Med. 2001 May;163(6):1389-94. [abstract]
3. Bernard GR, Vincent JL, Laterre PF, et al. Recombinant human protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study group. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med. 2001 Mar 8;344(10):699-709 [abstract]

 

Comments

Back to the EB Journal Club Index