Formulary And Economic Considerations

The high cost of new antibacterial therapies has forced institutions to develop strategies for limiting inappropriate prescribing. For institutions that have well-controlled formularies, the pharmacy and therapeutics (P&T) committee is often the gatekeeper when it comes to evaluating new products for formulary admission and creating restrictions for product use. When making such decisions, it is important that the P&T committee take into account the consequences and costs of untreated or inappropriately treated infections in addition to the safety profiles and costs of competing antibacterial agents. The consequences of infection, whether treated appropriately or not, are enormous, considering that 5% to 10% of patients admitted to acute care facilities develop at least one nosocomial infection and 90,000 of these patients die each year as a result of infectious complications (6). The cost to treat these infections is approximately $5 billion per year (6), and this figure is likely an underestimate, considering the indirect costs associated with nosocomial infections such as surgical site infections that frequently occur after discharge and are treated in outpatient settings (39). Recently, the Joint Commission on Accreditation of Hospitals included death due to nosocomial infection as a sentinel event requiring reporting and root cause investigation.

A basic understanding of outcomes research and particularly pharmacoeconomics is necessary when evaluating new antibacterials for formulary admission. This evaluation should go beyond simple cost comparisons. The need for this understanding has been recognized by P&T committee members as evidenced by a survey of pharmacy directors in acute care facilities in which more than 90% of respondents indicated that pharmacoeconomic information was used in formulary development (40). Despite the increased use of outcomes research in the decision-making process, some clinicians are skeptical of the value of health economics in evaluating new therapies. The skepticism will likely diminish as increased standardization occurs relative to the methodology and report-ing of outcomes research. The U.S. Public Health Service has published a series of articles pertaining to this issue (41-43), and other specialty organizations, such as the American Thoracic Society, have developed recommendations for their respective areas (44).

There are four general types of pharmacoeconomic analyses that are used to help determine the possible value of any given therapy. The first type, cost minimization analysis, is the most common type reported in the medical literature. This analysis is used to compare the costs associated with competing therapies and does not consider outcomes or benefits. In fact, it usually presumes that outcomes are equivalent. For example, antibacterial X costs $1,000 for a course of therapy compared with $500 for therapy Y, so therapy Y is the preferred approach because it is less costly. What if therapy X decreased the length of hospital stay or time on mechanical ventilation compared with therapy Y in patients with ventilator-associated pneumonia? In this case, the savings associated with therapy X could be large, considering that the development of ventilator-associated pneumonia has been shown to increase mean hospital charges per patient by more than $40,000 (45). What if therapy Y causes thrombocytopenia in 10% of patients receiving a course of therapy? These benefits and risks have associated costs that could easily offset the extra $500 cost of therapy X. The required presumption of therapeutic equivalence limits the usefulness of cost-minimization data.

Cost-benefit analysis is another type of pharmacoeconomic assessment that can be used to compare competing therapeutic strategies. In cost-benefit analysis, both the numerator (i.e., the benefit) and the denominator (i.e., the cost) are expressed in dollars. The advantage of this form of analysis is that it allows for comparisons of totally different treatment modalities. For example, the cost of infectious complications associated with the use of antibacterial-impregnated catheters could be compared to the cost of treating such complications with systemic antibacterial agents once infection occurs. In this example, the hypothesized benefit might be reduced cost of infectious complications associated with the catheters. If the benefit was clearly in favor of the catheters, the use of the catheters might be justified even if the relative costs associated with product acquisition (the denominator) were the same with both approaches. The major limitation of cost-benefit analysis is that it requires conversion of all benefits into a dollar figure, which may be very difficult with certain study end points (e.g., cost of death). Therefore, cost-benefit analysis tends to be used more for programmatic analyses than for therapeutic comparisons.

The last two types of pharmacoeconomic analysis are similar and are generally viewed as the preferred approaches when outcomes are not, or cannot be assumed to be, similar. In cost-effectiveness analysis the numerator (i.e., cost) is expressed as dollars but the denominator is expressed as an effectiveness measure, such as the number of infections cured. As in the cost-benefit calculation, the cost-effectiveness ratio must take into account all substantial cost differences associated with competing therapies, including the costs of adverse drug effects and hospitalization. Assigning these costs may be difficult, and there is debate among experts as to which form of medical resources evaluation (e.g., costs vs. charges) is preferred (46). Once calculated, the cost-effectiveness ratios of competing strategies are easy to compare. The situation becomes more complicated when there is no directly competing therapy, as occurred when drotrecogin alfa was marketed. Because drotrecogin alfa decreased mortality in a large, well-controlled trial, a cost-effectiveness calculation would require that a value judgment be made concerning the appropriate cost per year of life gained (47).

A variant of cost-effectiveness analysis is cost utility analysis, in which the denominator is expressed as a utility, most commonly, quality-adjusted life years (QALYs). The QALY is a unit of measurement on a numeric scale of 0 (usually representing death) to 1 (perfect health) and is determined using a variety of techniques, most notably, surveys. Although some authorities consider cost utility analysis to be the optimal economic analysis, utility data are rarely available for recently released medications. On the other hand, cost-effectiveness analyses are being increasingly performed with new medica-tions. In the absence of published cost-effectiveness analyses, it is often possible for P&T committee members to perform rudimentary analyses based on published clinical studies. As with cost-benefit analysis, both forms of cost-effectiveness analysis measure consequences as well as costs.

Figure 1 is a decision tree that models a hypothetical scenario in which three antibacterial agents are being compared for the treatment of pneumonia. A cost-effectiveness analysis is being used in which the numerator (i.e., cost) of each of the antibacterial pathways is the total cost of therapy associated with hospitalization, including costs for the antibacterial agent and its ADEs. It was assumed that the baseline cost of successful treatment was $69,000, whereas the cost associated with unsuccessful treatment was $105,000 (45). The cost of death was assumed to cost the hospital $105,000; note that this assumption demonstrates the importance of the perspective of the economic analysis, which in this case, is the institution. In addition to the antibacterial cost that is shown in the tree, the cost of a serious ADE was assumed to cost an additional $4,000 (2). The denominator (i.e., effectiveness) is based on hypothetical QALYs. Figure 1 provides the probabilities that specific events will occur. As an example, cost utility calculations are shown in Table 2 for drug X. The cure rates were based on two recent clinical trials in which antibacterial agents were compared for the treatment of community-acquired pneumonia (48,49).

ifikx'jijw iia.nl ^y*

Ujtw

  • waSlOdftlv)(
  • InftvtxraiKrt i Í
  • quot;JkdU
  • InftvtxraiKrt
  • quot;JkdU

LLWJ

IfWK'U K'ABO/IJJ

Figure 1 Decision tree for cost utility comparison of three hypothetical antibacterial drugs for the treatment of community-acquired pneumonia. Numbers below each branch represent event probabilities. The event cost and quality-of-life adjustment ($/QALY) is provided for each terminal branch. ADE, adverse drug event.

The decision tree demonstrates the problem with simply looking at medication cost to determine the best treatment scenario. In this example, the most cost-effective option (i.e., antibiotic Y) exhibited the lowest cost per treatment, considering overall cost of treatment. However, antibiotic Y was ten times more expensive than antibiotic Z. A simple cost assessment resulted in the following order of preference: antibiotic Y=$72,419/patient>antibiotic X= $73,270/patient>antibiotic Z=$74,454/patient. After ad-

Table 2 Example Cost Utility Calculation for Treatment X Based on the Hypothetical Decision Tree (Fig. 1)*

Cost per typical patient

Quality-of-life adjustment

0.92x0.1x$74,000=$6,808

0.92x0.1x0.3=0.0276

0.92x0.9x$70,000=$57,960

0.92x0.9x1=0.828

0.08x0.3x$106,000=$2,544

0.08x0.3x0=0

0.08x0.7x0.1x$110,000=$616

0.08x0.7x0.1x0.3=0.00168

0.08x0.7x0.9x$106,000=$5,342

0.08x0.7x0.9x1=0.0504

Sum of event costs=$73,270

QALYs (sum) 50.90768

  • The cost utility value, defined as the cost divided by the quality-adjusted life year (QALY) adjustment, is $80,723/QALY.
  • The cost utility value, defined as the cost divided by the quality-adjusted life year (QALY) adjustment, is $80,723/QALY.

justment for quality of life, the order of preference for antibiotics X and Z was different. The resulting preference order was antibiotic Y=$74,723/QALY>antibiotic Z=$78,017/QALY >antibiotic X/QALY=$80,723. When looking at the decision tree and calculating results, it becomes obvious that pharmacoeconomic calculations may become complex depending on the assumptions of the modeling process.

A number of approaches have been used to control the high costs of antibacterial agents and their associated adverse events, beginning with formulary evaluation and continuing with a variety of medication and disease management programs. Examples include therapeutic interchange strategies for antibacterial agents deemed to have equivalent efficacy and adverse effect profiles, intravenous to oral substitution, dosing regimen optimization programs, clinical pathways or protocols, and therapeutic guidelines (50). Some of these strategies, such as initiation of antibacterial agents within 4 to 8 hours of hospital admission for patients with pneumonia, are relatively easy to implement and have been shown to reduce length of stay and mortality (51-53). Other strategies that focus on alterations in prescribing (e.g., broad-spectrum antibacterials initially, with rapid streamlining based on microbiologic results) are usually more difficult to implement but may yield rewards in terms of reduced antibacterial consumption, decreased resistance, and possibly decreased mortality (54,55).

Guidelines for many infectious diseases have been published, and many of the evidence-based versions can be found on the National Guideline Clearinghouse website (56). Although these guidelines are based on evidence from clinical trials, opinions of the contributors enter into the guidelines when published studies do not allow evidence-based conclusions. Despite any shortcomings, consensus guidelines provide an important framework for clinical decisions. However, they should not be applied without careful evaluation of individual situations.

Some of the more controversial approaches used primarily to prevent resistance (but that have economic implications) involve infection control measures such as antibacterial agent cycling or other attempts to restrict prescribing (57,58). These types of mandated control measures are the focus of much debate, given the disagreement among experts regarding treatment options, as exemplified by an international conference that attempted to derive consensus on the diagnosis and treatment of ventilator-associated pneumonia (59). The participants reached consensus on issues such as the importance of local surveillance programs, but there was little agreement on the appropriate choice of antibacterial agents other than the need to customize empiric therapy.

Most of the approaches used to control costs or prevent resistance have not been evaluated by formal economic analyses, and the evidence of their resource savings is usually based on cost minimization. As noted in one review of these strategies, no savings are likely to occur if a strategy is developed but never directly applied to patient care (60). Ways to insure that the strategies are utilized include their development and implementation using a multidisciplinary, systems-based approach that takes into account institution-specific issues (61).

Was this article helpful?

0 0

Post a comment