Pharmacoeconomic Model of Serotonin Reuptake Inhibitors: INCLUSION OF GENERIC PRODUCTS
The effect of ADRs on overall treatment costs among SRIs can be substantial. In the model presented during the round-tables, a branded SRI, escitalopram (Lexapro canadian, Forest) incurred the lowest overall expected cost over the course of six months, ahead of another branded SRI, canadian citalopram (Celexa drug, Forest), and a generic SRI, fluoxetine. (Table 1 presents a list of drugs included in the model.) Even if the cost of generic fluoxetine or generic paroxetine were reduced to $0, the model still showed canadian escitalopram therapy to be the most cost-effective strategy, because it had the lowest rate of ADRs within the class.
The promotion within a benefit plan of a branded product over generic drugs can be difficult. Several participants said that even if a model shows that the overall cost of any branded product is lower than that for any other product, including generics, that drug could be grouped with the generics; the generics, however, would not be placed at a disadvantage if the branded drug received a more favorable position on the formulary. Once physicians have adopted the habit of prescribing generics, the participants said, it is undesirable to attempt to induce them to abandon that habit by structuring a formulary so as to favor a branded product over generic ones.
Table 1 SRIs Included in the Sullivan Model
| Generic Name | Trade Name | Manufacturer |
| Generic Citalopram | Celexa medication | Forest |
| Escitalopram generic | Lexapro drug | Forest |
| Fluoxetine medication | Prozac canadian, Sarafem drug | Lilly |
| Paroxetine tablet | Paxil canadian | GlaxoSmithKline |
| Paroxetine controlled release | Paxil® CR | GlaxoSmithKline |
| Sertraline canadian | Zoloft drug | Pfizer |
| Venlafaxine | Effexor generic | Wyeth-Ayerst |
| Venlafaxine canadian extended release | Effexor® XR | Wyeth-Ayerst |
| * The generic product was used in the model. | ||
The participants said that since the time generic SRIs became available, one way in which their utilization has been encouraged is by requiring physicians to obtain prior authorization for continued use of branded products. The use of generic SRIs also has been driven by cost-containment strategies that cap the annual amount of drug reimbursement. Under this system, physicians may prescribe as they see fit, but patients—instead of demanding a branded drug that necessitated prior authorization per the rules of the formulary—now worry about exceeding their cap. As a result, patients may resist the prescription of a branded product instead of a generic one. A drug on a higher tier of such a formulary might not be prescribed with any regularity, because its co-payment may be too high relative to a drug benefit cap or because it exhausts the benefit limit for a patient using multiple drugs.
Participants said that the Sullivan model might be useful for helping P&T committees decide on the tier to which an SRI is assigned. As matters stand, generics will be favored as first-line therapy.
USEFULNESS OF THE MODEL
Participants said that, from the perspective of a health plan, the problem with many pharmacoeconomic models is that they deal with costs occurring outside the health plan, such as quality of life and absenteeism. These concerns may be important to employers, but a health plan is interested in events that have an impact on the overall budget of the health plan (e.g., rates and costs of hospitalizations and office and emergency room visits). This approach may differ for health plans in which the major client base consists of employers.
canada pharmacy mall
The participants said that the only kind of pharmaco-economic model that a P&T committee should trust is one that is transparent (i.e., where all assumptions, rates, and costs are shown to readers and, ideally, can be altered by readers to fit their own circumstances) and one that a committee could adapt to suit the demographic characteristics of its own plan. It is extremely important to know the assumptions behind the sensitivity analyses in a model, they said, because those assumptions may not correspond with the experience of a given health plan. In addition, a good model must be able to evaluate all products in a class and must be flexible enough to accommodate market changes.
The ability to customize the model depends on the availability of specific health plan data. In one organization, there would be access to pharmaceutical data, but the medical data might be less complete or less detailed. Merging pharmacy and medical data can be extremely difficult.
