Access Denied

By Cory Cowan
Approximately 40% of drug spending growth in Canada is attributed to the substitution of high-cost for low-cost drugs used to treat the same medical condition. In addition, there is an increasing demand from plan members for new and more costly medications. In response, drug plans have developed a number of cost containment strategies targeting high-cost drugs.

One such strategy is prior authorization(PA). PA functions on the premise that plan members must submit medical information from their physicians before reimbursement can be granted to justify the medical need for expensive drugs when less costly alternatives are available. PA is also commonly applied to drugs that have potential for off-label or experimental use, and to restrict payment for drugs administered exclusively in a hospital setting.

PA has become an increasingly popular cost containment strategy used by private and public drug plans in Canada over the last decade. Data from ESI Canada’s database indicates that only 52% of employer groups used PA on one or more drugs in 1999. By 2005this had increased to almost 70%. Currently, all provincial plans use similar forms of PA, although the classification differs by province(“Limited-Use” and “Section 8” in Ontario; “médicaments d’exception” in Quebec; and “special authorization” in Atlantic Canada, for example).

Despite PA’s widespread use as a benefit management tool, it has not been studied extensively.
The literature, while limited, suggests that PA programs have had success in reducing drug costs without increasing the use or costs of other medical services. Several studies report that PA programs have reduced expenditures by as much as 50% in some drug categories. Yet, despite these positive results, two main criticisms remain: the high PA approval rate(usually in excess of 90%), and the increased administrative burden associated with PA.

THE STUDY
To investigate these issues further, ESI Canada conducted a study to examine the drug plan cost savings achieved by PA programs. Specifically, the primary objective was to track the claim histories of drug plan beneficiaries who were denied a point of sale(POS)claim for a variety of drugs targeted by PA, to determine what kind of drug, if any, they eventually did get reimbursed

In order to evaluate the impact of the POS rejection on claimant reimbursement rates for drugs commonly requiring PA, a retrospective claims analysis was completed. Claims data for all claimants with a POS rejection for any one of eight drugs were collected over a one-year period(October 1, 2003 through September 30, 2004). Drugs were grouped by therapy class:

• rheumatoid arthritis(RA)biologics(Remicade® and Enbrel®);
• osteoarthritis drugs(Celebrex® and Vioxx®);
• erectile dysfunction(ED)drugs(Viagra™ and Cialis®); and
• anti-obesity agents(Meridia® and Xenical®).

Individual claimants were followed up to 120 days following their POS PA rejection(30 days for RA and osteoarthritis; 120 days for ED and anti-obesity agents)and were categorized into three main groups:

1)reimbursed for the PA drug,
2)reimbursed for a substitute non-PA drug, or
3)non-reimbursed(though still could have paid for the PA drug out of pocket)

THE FINDINGS
Once claims data were collected and analyzed, it was found that the rate of reimbursement for PA drugs varied greatly depending on the therapy class evaluated(see table below). The highest rate of reimbursement was seen for the RA biologics, for which 74.3% of claimants were reimbursed, whereas the rate of reimbursement for the other drugs classes was relatively low.

This high rate of reimbursement for RA biologics may be due to the fact that rheumatoid arthritis is a severe, chronic, debilitating disease and thus patients desperately need treatment. Secondly, biologic drugs are usually reserved as last line therapies because they require injection under the skin or into the vein, and have potentially serious side effects, such as an increased risk of cancer, associated with their use. As a result, patients who have already progressed through the older, oral medications are probably the types of patients pursuing biologic treatment. Lastly, the RA biologics are extremely costly, ranging between $20,000 and $30,000 per patient per year. Perhaps the need to offset some of this financial burden and have the private plan share some of this cost is critical for the patient.

Although a low percentage of patients(20.7%)with osteoarthritis were reimbursed for Celebrex® and Vioxx®, this therapy class had the greatest percentage of patients who pursued alternate non-PA drugs. A total of 32.2% of patients with a POS rejection had a reimbursed claim for a less costly alternative. The most commonly substituted medications are listed in the table below. In this drug class, prior authorization was quite effective as only 20% received the high cost drug, and those who didn’t either resorted to a less costly alternative or had no subsequent claims.

For the so-called “life-style drugs” to treat ED and obesity, similar rates of reimbursement were seen as with the osteoarthritis class. When evaluating claims for anti-obesity agents, it was found that 30.4% of patients had a reimbursed claim for Xenical® or Meridia® within 120 days of the POS rejection. For ED drugs, the percentage of patients reimbursed was even less at 16.5% or about half the rate of the anti-obesity category.

The second part of the research study was to estimate the impact of PA on drug plan costs while accounting for administrative costs(estimated at $50.00 per PA form reviewed)and expenditures for substitute non-PA drugs. This analysis indicated that the percentage of drug cost savings within osteoarthritis, ED, and anti-obesity drug classes was on average 40% over a given year. For RA biologics, only a 7.3% savings in drug costs was calculated. However, because these drugs are so costly, this translated into $918 of savings per patient. Savings per claimant in the other therapy classes ranged from $135/claimant to $230/claimant for osteoarthritis and anti-obesity drugs, respectively.

The findings show that cost containment strategies implemented at the POS can be highly effective. Of the 4,510 patients who initially attempted to have a prescription filled for one or more of the eight PA drugs, only 27.6% actually received reimbursement for one following a POS rejection. Without the presence of PA at the POS, the majority of these claims would have likely been paid. So, it is essential for plan sponsors to measure savings from the POS since a large percentage of patients do not pursue PA approval beyond this point. Although the PA approval rate may be in excess of 90%, savings are being generated from the POS rejections.

Unfortunately however, we don’t know the clinical implications of patients not having a claim for either a PA or a substitute non-PA drug and whether this negatively impacts drug plan costs and, plan members.

Continuous review and evaluation of PA programs is key to determine if the savings outweigh the costs. PAs can continue to save drug plans money provided the program is managed appropriately, drugs for PA are selected wisely, and the criteria for approval are clinically sound and follow established treatment guidelines.

Cory Cowan is a clinical pharmacist with ESI Canada in Mississauga, Ont.

For a PDF version of this article, click here.