It's the law of unintended consequences: A restrictive drug formulary maintained by a health plan may influence how physicians treat patients unaffiliated with that plan. This phenomenon is known as the spillover effect, says Y. Richard Wang, MD, a researcher in the public policy department of AstraZeneca Pharmaceuticals.
His research implies the limited value of restrictive formularies in controlling overall utilization, and the research points to how restrictive plans affect each other across geographical and commercial lines.
In the last three years, Wang conducted two studies of the effect of restrictive formularies on regional utilization of proton pump inhibitors in California and Maine. "At issue is whether a highly restrictive formulary causes overall utilization to rise for a specific drug," he says.
Using two independent databases — InfoScan Formulary Database by MediMedia USA (which owns this publication) and Formulary Focus by IMS Health — Wang and colleagues examined the effect that formulary changes by PacifiCare in California had on utilization of PPIs by non-PacifiCare physicians. The two PPIs are rabeprazole and pantoprazole.
Wang's work, the results of which were published in the American Journal of Managed Care in January 2005 in an article titled "Spillover Effects of Restrictive Drug Formularies: A Case Study of PacifiCare in California," demonstrate that restrictive formularies are "blunt tools for managing the pharmacy benefit" that can inadvertently effect overall utilization patterns, says Dana Goldman, PhD, of the Rand Corp.
"Formulary decisions are often made on the basis of ingredient costs and manufacturer rebates rather than clinical outcomes," says Goldman. "Even in more enlightened pharmacy and therapeutics committees, the decisions can still be based on aggregate measures without [allowing any] tailoring to each patient's circumstances."
Also, employers have incentives to discourage use beyond what may be best for patients since many of the benefits of pharmacotherapy accrue long after people have left their jobs, says Goldman. "The result is that plans are designing formularies that are socially wasteful, and that patients perceive as overly intrusive," he says.
Wang found that the number of rabeprazole and pantoprazole prescriptions increased simultaneously for non-PacifiCare patients and that the increase was positively associated with PacifiCare's share of practice. For non-PacifiCare prescriptions, a 10 percent increase in PacifiCare's share of practice led to a 3.3 percent share increase for rabeprazole and a 1.6 percent share increase for pantoprazole. He concluded that "PacifiCare's PPI formulary changes generated significant spillover effects onto non-PacifiCare patients in California."
The PacifiCare research reinforced a finding that Wang made two years earlier when he looked at a change in Maine's Medicaid drug formulary related to PPIs.
In that study, Wang conducted a three-state comparison after Maine's Medicaid program implemented a restrictive drug formulary for the proton pump inhibitor class, with pantoprazole as the only preferred drug. He and his colleagues studied both the Medicaid and the non-Medicaid market shares of pantoprazole in Maine for three months after the change, and compared those findings to utilization of the drug in New Hampshire and Vermont in the same period.
"We studied the effects of a restrictive drug formulary of one third-party payer on a physician's prescribing patterns, not only for patients of that specific payer but also for other patients," says Wang.
"The size of the spillover effects may be related to the market share of the third-party payer and its drug formulary. And the spillover [affects] drug treatment choices for patients covered by other insurers and patients without drug insurance."
Wang and his colleagues found that after three months, the market share of pantoprazole in Maine increased 79 percent for Medicaid prescriptions, compared to a 1-percent increase in New Hampshire and a 2-percent increase in Vermont during the same period.
It also increased 10 percent for cash prescriptions, but only 3 percent in each of the two neighboring states, and it increased 7 percent for other third-party payer prescriptions, compared to 1 percent each in the other two states. The researchers concluded that "Maine's Medicaid drug formulary generated spillover effects in cash and other third-party payer markets, with somewhat stronger effects in the cash market."
In discussing the Maine research, Wang points to the fact that private insurers and state Medicaid programs generally have followed different routes in addressing the rising costs of prescription drugs.
"Private insurers have increasingly turned away from restrictive formularies and have favored patient-side financial incentives, such as three-tiered copayments, to control drug costs," he says. "Medicaid programs have been unwilling and often unable to use high levels of patient cost sharing and so have increasingly abandoned open formularies and have adopted more restrictive ones. Such a change in market dynamics makes Medicaid, often a dominant player by market share, a potential source of spillover effects."
Goldman says that HMOs as well as government policymakers should heed the implications of Wang's research and consider comparatively open formulary designs based primarily on outcomes and patient preferences.
"A promising trend is the move toward benefit-based copayments," says Goldman. "This approach makes the patient copayment depend on the expected therapeutic benefit, as determined using available evidence. The idea is to ensure that effective drugs get into the hands of patients who truly need them, and that the patients continue to take them."
Henry Grabowski, PhD, a health care economist at Duke University, says that Wang's research has additional implications. "The research leads to the need to know more about how restrictive formularies affect the costs of other medical services, and about their effects on overall patient health status and quality of life," says Grabowski.
"Maine is clearly representative of what is happening in many other states in terms of a movement to closed formularies with prior authorization," says Grabowski. "Some states have initiated preferred drug programs that focus on only a few major therapeutic classes. Others have adopted comprehensive programs spanning virtually all drugs except for a few exempt categories. And many other states have pending legislation or programs that will restrict access to drugs."
The net result, says Grabowski, is that restrictive policies in the public sector could lead to restrictive formularies in the private sector, and a subsequent managed care backlash.
Will federal policy officials then be more inclined to respond with very restrictive formularies if this has become the cost-containment instrument of choice in state plans?"
The adverse effect of restrictive formularies may be to reduce the choice of drugs in other plans and populations whose formularies are based on a more enlightened design related to therapeutic outcomes, says Goldman.
"Economic benefits based on therapeutic value and quality reporting represent a good start," he says. "Cost-effectiveness analysis is not sufficient."