Gregory Hess, MD, MBA; Kafi N. Sanders, MPH; Jerrold Hill, PhD; and Larry Z. Liu, MD, PhD
Patient switching of prescription drug brands within a therapeutic class has become increasingly prevalent with tiered drug plan formularies. Although switching from more expensive brand name drugs to less expensive brands or generic equivalents may reduce aggregate prescription costs, therapeutic benefit may be compromised if the patient is not switched to a drug with an equivalent therapeutic profile. This could have especially important implications for the treatment of asymptomatic conditions, in which the loss of benefit might not be immediately obvious—for example, cardiovascular risk reduction.
Coronary heart disease (CHD) affects an estimated 13.2 million persons in the United States, and is the single largest cause of death among both men and women, accounting for 1 of 5 US deaths in 2003.1
Low-density lipoprotein cholesterol (LDL-C) is an important risk factor for CHD; randomized controlled trials have demonstrated that lowering LDL-C reduces the risk for major coronary events. The National Cholesterol Education Program Adult Treatment Panel III guidelines, published in 2002, identified LDL-C as the primary target of lipid-lowering therapy.2,3
Statins are among the most commonly prescribed and potent LDL-C–lowering agents; however, they differ in their therapeutic profiles. For example, although the recommended dosage ranges for simvastatin and atorvastatin are similar (5-80 mg/day and 10-80 mg/day, respectively4,5
), the 2 drugs differ in potency, as discussed below.
A number of randomized studies have compared the effects on LDL-C of atorvastatin versus simvastatin at various dose levels.6-19
These studies differ in design (open-label or double-blind; parallel-group or crossover), numbers of patients randomized to atorvastatin versus simvastatin (range, <100 to >1700), treatment duration (range, 6-54 weeks), patient characteristics (mean baseline LDL-C levels ranging from approximately 172.5-346 mg/dL), and dosing regimens (fixed-dose, dose escalation, or titration to target). Despite these differences, it has been a consistent finding that atorvastatin is significantly more potent than simvastatin at milligram-equivalent doses—including 10 mg,6-8,10,11,14,16
and 80 mg.13,15,18
of 5 of the above studies7,9,12,13,16
found that both drugs reduce LDL-C in a dose-dependent manner, and that the percent decrease of LDL-C from baseline is slightly greater with atorvastatin 10, 20, or 40 mg than with simvastatin 20, 40, or 80 mg, respectively. Similar findings were reported in a post-hoc analysis21
of 3 studies.9,12,13
We examined whether patients switching from atorvastatin to either branded or generic simvastatin received therapeutically equivalent (or higher) simvastatin doses. For purposes of this study, simvastatin doses of 20, 40, and 80 mg were considered therapeutically equivalent to atorvastatin doses of 10, 20, and 40 mg. Because the maximum recommended dosage of simvastatin is 80 mg/day, there is no simvastatin dose equivalent to atorvastatin 80 mg.Methods
We conducted a retrospective, patient-level analysis of pharmacy claims for patients with an atorvastatin prescription dispensed in September 2005, the index month.
The claims were obtained from Surveillance Data Inc's data warehouse (Table 1
) in National Council for Prescription Drug Programs 5.2 format. The claims are collected electronically on a daily basis from approximately 40 000 retail pharmacies covering all 50 states. The volume of claims exceeds 1.4 billion annually dispensed prescriptions. Each patient in the database is assigned a unique synthetic identification number allowing for the longitudinal tracking of the claims per patient over time.
All patients with a prescription drug claim for atorvastatin in the index month, September 2005, were selected for inclusion. The resulting sample contained 453 409 unique patients. These patients were subsequently followed for 12 months to determine
the percent switching, including the subset that switched to simvastatin and their relative therapeutic doses after switching. A number of studies comparing milligram-equivalent doses of atorvastatin and simvastatin have demonstrated that atorvastatin is more potent than simvastatin.6-8,10,11,13-19
Thus, patients switching to the same or lower milligram dose of simvastatin were classified as receiving a lower therapeutic dose compared with their atorvastatin dosing.
The following categories were used to characterize patients switching from atorvastatin to any other statin, the subset switching to any simvastatin, and the subsets switching to branded or generic simvastatin:
1. Patients switching to any other statin
: The total number of unique patients by month 12 who did not have the same statin dispensed that they had in the index month.
2. Patients switching to any simvastatin, branded or generic
: The count of unique patients who switched from atorvastatin to any milligram dose level of simvastatin within 12 months of the index period. If the patient switched to the branded
drug and then generic or vice versa, the patient was counted uniquely only once in this calculation.3. Patients switching to branded simvastatin
: The count of unique patients who switched from atorvastatin to any milligram dose level of branded simvastatin within 12 months of the index period.4. Patient switching to generic simvastatin
: The count of unique patients who switched from atorvastatin to any milligram dose level of nonbranded simvastatin within 12 months of the index period. If the patient switched to the branded drug and then generic or vice versa, the patient was counted in this category and category 3.
Patients switching to simvastatin were also categorized by their dose level for atorvastatin in the index month and the dose level of the simvastatin they switched to, using the following criteria: 5. Patients switching to a specific dose of either branded or generic simvastatin from a specific dose of atorvastatin
: The count of unique patients who switched from a specific dose of atorvastatin at index to a specific dose of either branded
or generic simvastatin within 12 months of the index period. If the patient switched to both the branded drug and generic drug at the specific simvastatin dose level during the study period, the patient was counted uniquely only once at the simvastatin dose level. For example, a patient switching from 10-mg atorvastatin to branded 20-mg simvastatin
and then to generic 20-mg simvastatin counts as 1 switch from 10-mg atorvastatin to 20 mg of either branded or generic simvastatin. Note that if a patient had multiple changes in dose levels of simvastatin during the study period, the patient was counted in each of the dose levels of simvastatin prescribed. 6. Patients switching to a specific dose of branded simvastatin from a specific dose of atorvastatin
: The count of unique patients who switched from a specific dose of atorvastatin at index to a specific dose of branded simvastatin within 12 months of the index period. Note that if a patient switched to more than 1 dose level of branded simvastatin during the study period, the patient was counted in each of the dose levels of branded simvastatin prescribed.7. Patients switching to a specific dose of generic simvastatin from a specific dose of atorvastatin
: The count of unique patients who switched from a specific dose of atorvastatin at index to a specific dose of simvastatin within 12 months of the index period. Note that if a patient switched to more than 1 dose level of a generic simvastatin during the study period, the patient was counted in each of the dose levels of generic simvastatin prescribed.
The last 3 categories were used to classify patients as switching to either an equivalent-or-higher or a lower therapeutic dose.Results
Among the 453 409 patients using atorvastatin at the beginning of the study, 56 919 (13%) switched to another statin (Table 2
, switching category 1). The likelihood of switching was similar for patients taking 10 mg to 40 mg of atorvastatin at the start of the study (12%-14%), but somewhat lower for patients at the 80-mg dose (7%). A total of 13 530 patients (3% of all patients) switched to either branded or generic simvastatin by the end of the study period. Of these, 8537 patients switched to branded simvastatin and 9766 switched to generic simvastatin sometime during the year, indicating multiple changes of medication during the year for many patients (Table 2, switching categories 2-4).
PDF is available on the last page.