Steven R. Krantz, Robert S. Rase and Robert W. Piepho
Objective:
Evaluate the efficacy, safety, and cost-effectiveness of a formulary transition from nifedipine GITS (Procardia XL, Pfizer) to felodipine ER (Plendil, Astra Merck).
Design:
Retrospective analysis of patient records for three months before and three months after formulary transition was initiated.
Setting:
A mixed-model metropolitan health-maintenance organization with 110,000 members.
Patients:
Health records of 246 patients with stable hypertension were reviewed.
Intervention:
The formulary long-acting calcium-channel blocker was switched from nifedipine GITS to felodipine ER.
Main Outcome Measures:
Patient blood pressures, frequency of adverse drug events, and cost of medications.
Result:
Average preconversion and postconversion blood pressure values were 144/85 mm Hg and 142/84 mm Hg, respectively. Of 380 subjects, 28% were identified as hypertensive (systolic blood pressure ³ 140 mm Hg and diastolic blood pressure ³ 90 mm Hg) before conversion, while 21% were identified as hypertensive after the formulary switch. Differences between treatments were not statistically significant. Chest pain-the most commonly noted adverse effect during nifedipine GITS therapy (7.7%)-decreased after conversion to felodipine ER (4.1%). The most common adverse effect associated with felodipine ER was edema (8.9%). Felodipine ER costs represented approximately 85% of nifedipine GITS costs, resulting in a 15% savings to the health care system.
Conclusion:
The therapeutic equivalence and comparable safety profile of the two medications combined with the 15% cost savings realized with felodipine ER validated the successful formulary switch from nifedipine GITS to felodipine ER at this health-maintenance organization.
Key Words:
Formularies, Therapeutic substitution, Calcium-channel blockers, Costs, Adverse drug events, Drug efficacy, Nifedipine, Felodipine.
J Managed Care Pharm 1996; 2: 642-646.
When similar drug therapies exhibit comparable safety and efficacy, decisions made by health care providers as to which drug to include in their formularies may depend on factors such as cost, patient response and tolerance, drug safety with concomitant disorders, and patient and physician preference.1 In 1992, Humana Health Care Plans, a mixed-model HMO with approximately 110,000 members, changed its formulary long-acting dihydropyridine calcium-channel blocker from nifedipine GITS (Procardia XL, Pfizer) to felodipine ER (Plendil, Astra Merck). Felodipine and nifedipine are both members of the dihydropyridine class of calcium-channel blockers, with demonstrated safety and efficacy for the treatment of essential hypertension.2-4
Although members of the dihydropyridine class have different pharmacodynamic effects, three comparative clinical trials conducted with hypertensive patients have shown these two agents to have similar efficacy and adverse event profiles.1,5,6 Felodipine is more selective for vascular smooth muscle over myocardial tissue than nifedipine1; at doses producing equivalent vasodilation, felodipine lacks nifedipine's negative inotropic effect.7-9 The probability of similar patient health outcomes combined with a potential cost savings instigated this large metropolitan HMO to undertake this formulary switch. Humana Health Care Plans' formulary transition from nifedipine GITS to felodipine ER was initiated in September 1992 and has continued to the present.
This study, a retrospective record review of 380 patients with stable hypertension who switched from nifedipine GITS to felodipine ER between January 1993 and December 1994, was conducted to determine whether health outcomes were affected by the formulary switch. This managed care organization (MCO) also sought to determine whether felodipine ER was less expensive than nifedipine GITS, and a more cost-effective treatment of essential hypertension.
Methods
The Humana Health Care Plans formulary was modified to convert patients on nifedipine GITS to felodipine ER in September 1992. To facilitate the formulary change, information was provided to all prescribers in the Humana system, and product interchange occurred upon renewal of prescriptions at all pharmacy sites in the Humana system. Patients without angina receiving nifedipine GITS to treat hypertension were switched to felodipine ER at a dose ratio of 1 mg felodipine ER per 6 mg nifedipine GITS.
Data Collection
Records of 380 patients who were switched to felodipine between January 1993 and December 1994 were examined. A six-month period of each record was evaluated (three months before and three months after the transition from nifedipine GITS to felodipine ER) for historical, efficacy/response, and safety data. Historical data included patient characteristics (age, gender, weight, height, and race) and patient history (use of other antihypertensive drugs, use of potentially interacting drugs, and duration of therapy). Records with the following variances were excluded from consideration:
Efficacy/response data included preconversion and postconversion systolic and diastolic blood pressures, calcium-channel blocker doses, dosage adjustments, number of office visits, changes in therapy, treatment failures (i.e., lack of blood pressure control or discontinuation of felodipine ER treatment for medical reasons), and any new diagnoses during the three months following the change in therapy.
Safety data included all reported adverse effects.
The relative effectiveness of nifedipine GITS and felodipine ER was determined by comparing steady-state preconversion and postconversion blood pressure readings. Steady state was defined as the blood pressure reading on the conversion date for the preconversion measure and the last blood pressure value recorded during the study period for the postconversion measure. The interval between these measurements varied from one to three months, depending upon scheduling of visits. Hypertension was defined as a systolic blood pressure reading of >= 140 mm Hg and/or a diastolic reading of >= 90 mm Hg, consistent with the recommendation of the Fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure.10
Compliance and costs were also monitored. Compliance was based on patient prescription refills. Patients were considered to be compliant if their prescription refill included sufficient doses to cover the number of days in each patient's preconversion and postconversion period. Costs to the system were estimated based upon costs per unit dose and the frequency with which the medication was administered.
Statistical Analysis
Data were coded and entered in standard format on a VAX-1 mainframe computer at the University of Missouri at Kansas City. The database was validated using standard procedures, and an error rate of less than 1% was determined. A random selection of every fifth record was verified manually.
The effects of nominal data (gender, race, and use of other antihypertensive drugs) on the outcome of the formulary transition were assessed using chi-square analysis. Interval and ratio data were examined using analysis of variance (ANOVA) and multiple regression procedures. Multiple regression was used to identify the potential impact of factors or clusters of characteristics on clinical outcomes and adverse drug events during the transition process. The accumulated data on safety were analyzed as net data, since it is possible that some adverse effects may be ascribed to other medications initiated following the discontinuance of nifedipine GITS.
Conclusions based upon these analyses employed traditional levels of significance (p < 0.05). The study was powered on the variable of blood pressure and the analysis indicated the ability to detect very small effects (0.2 for difference analyses and r = 0.19 for correlational analyses), with 90-95% confidence intervals.
Results
Of 380 patients converted from nifedipine to felodipine during the course of the study, 246 efficacy-evaluable patient records remained after removal of excluded patients or incomplete files. The majority of patients were Caucasian (53.3%), while 37.0% were African-American. Average patient age was 59 years, and average weight and height were 87.3 kg (192 pounds) and 167.6cm (66 inches), respectively (Table l).
Table 1. Patient Demographic Characteristics and Concomitant Drug Use
| Characteristic | Value |
| Mean (± S.D.) age (years) | 59.4 ± 14.0 |
| Age range (years) | 29.5-90.1 |
| Mean (± S.D.) weight (pounds) | 191.6 ± 51.5 |
| Weight range (pounds) | 95-395 |
| Mean (± S.D.) height (inches) | 66.3 ± 4.2 |
| Height range (inches) | 52-77 |
| Gender | |
| No. men (%) | 107 (43.5) |
| No. women (%) | 139 (56.5) |
| Ethnicity | |
| No. Caucasian (%) | 131(53.3) |
| No. African-American (%) | 91 (37.0) |
| No. Asian (%) | 3 (1.2) |
| No. Hispanic (%) | 2 (0.8) |
| No. not indicated (%) | 19 (7.7) |
| Adjunct Medications | |
| No. of patients on ACE inhibitors (%) | 83 (33.7) |
| No. of patients on diuretics (%) | 83 (33.7) |
| No. of patients on beta blockers(%) | 38 (15.4) |
| No. of patients on calcium-channel blockers (%) | 24 (9.8) |
| No. of patients on alpha blockers (%) | 5 (2.0) |
| Interacting Medications | |
| No. of patients on cimetidine (%) | 12 (4.9) |
| No. of patients on yohimbine (%) | 1 (0) |
| No. of patients on phenytoin (%) | 1 (0) |
| No. of patients on loxapine (%) | 1 (0) |
| No. of patients on carbamazepine (%) | 1 (0) |
Of the 246 patients, 175 (71%) received at least one adjunct medication for hypertension during the study period, with the most frequently prescribed medications being lisinopril (27.6%) and hydrochlorothiazide (17.9%). Approximately one third of these patients also concurrently used either an angiotensin-converting enzyme (ACE) inhibitor or diuretic ( or, in some cases, both) at some point during the assessment period. Approximately 10% of patients received concurrent therapy with other calcium-channel blockers. A total of 16 patients (6%) received medications that would potentially interact with calcium-channel blockers, the most common being cimetidine (n = 12).
Statistical analyses, including ANOVA, chi-square analysis, and multiple regression, yielded no statistically significant correlations or differences in preconversion and postconversion measures. Thus, no statistically significant differences were found for any safety or efficacy parameter measured.
Record review indicated 98.5% of patients were compliant with their medication regimens. Mean preconversion blood pressure values were 144/85 mm Hg, while mean postconversion values were 142/84 mm Hg. Some 28% of subjects were identified as hypertensive (systolic blood pressure >= 140 mm Hg and diastolic blood pressure >= 90 mm Hg) during nifedipine GITS treatment, while 21% were identified as hypertensive using this definition during felodipine ER therapy. The differences in these values were not statistically significant. Mean blood pressure values and hypertension prevalence before and after the formulary transition are shown in Table 2.
Table 2. Drug Effectiveness: Blood Pressures and Hypertension Prevalence before and after Formulary Transition (n = 246)
| Stable preconversion blood pressures | |
| Mean (± SD) systolic value (range, mm Hg) | 144.3 ± 20.2 (102-200) |
| Mean (± SD) diastolic (range, mm Hg) | 84.9 ± 13.5 (50-140) |
| Stable postconversion blood pressures | |
| Mean (± SD) systolic value (range, mm Hg) | 141.9 ± 21.1 (90-210) |
| Mean (± SD) diastolic (range, mm Hg) | 83.5 ± 12.0 (56-125) |
| Prevalence of hypertension preconversion | |
| No. patients with systole > 140 and diastole > 90 mm Hg (%) | 68 (27.6) |
| No. patients with diastole >90 mm Hg (%) | 69 (28.0) |
| Prevalence of hypertension postconversion | |
| No. patients with systole > 140 and diastole > 90 mm Hg (%) | 52 (21.1) |
| No. patients with diastole >90 mm Hg (%) | 69 (28.0) |
Adverse effects were identified from patient-reported complaints in the medical record. Chest pain was the most common preconversion adverse effect (7.7%); prevalence decreased after the formulary conversion (4.1%). The most common adverse effect after conversion was edema (8.9%). Complete data regarding adverse effects before and after transition are in Table 3. A total of 14 new diagnoses, representing a range of medical conditions and distributed across nine patients, were recorded after transition to felodipine ER (Table 4). Only two of these (vertigo and supraventicular tachycardia) were considered to be potentially related to the new regimen.
Table 3. Prevalence of Adverse Effects before and after Formulary Transition
| Preconversion | Postconversion | ||||
|---|---|---|---|---|---|
| Adverse Effect | N | % | N | % | X2 |
| Chest pain | 19 | 7.7 | 12 | 4.9 | 2.531 |
| Edema | 13 | 5.3 | 22 | 8.9 | 2.492 |
| Dizziness | 10 | 4.1 | 12 | 4.9 | 0.190 |
| Headache | 8 | 3.1 | 15 | 6.1 | 2.234 |
| Gastrointestinal reactions | 5 | 2.0 | 7 | 2.8 | 0.342 |
| Heart rate changes | 5 | 2.0 | 7 | 2.8 | 0.342 |
| Nocturia | 3 | 1.2 | 3 | 1.2 | 0a |
| Shortness of breath | 3 | 1.2 | 4 | 1.6 | 0a |
| Rashes | 2 | 0.8 | 5 | 2.0 | 0.580 |
| Asthenia | 2 | 0.8 | 1 | 0.4 | 0a |
| Otherb | 0 | 0 | 5 | 2.0 | 0a |
| Flushing | 1 | 0.4 | 1 | 0.4 | 0a |
| Rhinitis | 1 | 0.4 | 2 | 0.8 | 0a |
| Sleep disturbances | 1 | 0.4 | 3 | 1.2 | 0.252 |
| Fatigue | 1 | 0.4 | 3 | 1.2 | 0.252 |
| Impotence | 1 | 0.4 | 1 | 0.4 | 0a |
Table 4. Additional Diagnoses First Detected in the Three Months after Transition to Felodipine ER
| Diagnosis | No. Patients |
| Glucose intolerance | 2 |
| Degenerative changes/knee | 2 |
| Seborrhea of scalp | 1 |
| Sternal osteomyelitis | 1 |
| Possible deep-vein thrombosis | 1 |
| Myocardial infarction | 1 |
| Adenocarcinoma | 1 |
| Vertigo | 1 |
| Noninsulin-dependent diabetes mellitus | 1 |
| Supraventricular tachycardia | 1 |
| Chronic obstructive pulmonary disorder | 1 |
| Left ventricular hypertrophy | 1 |
During the study period, 27 patients (11%) stopped taking felodipine ER. Some 11 patients had no reason for the discontinuation listed in the medical record. The remainder cited these reasons for drug discontinuation: headache (n = 4), dizziness (n = 2), fixed drug eruption (n = 2), suspected congestive heart failure (n = 2), peripheral edema (n = 2), drug discontinued upon hospitalization (n = 1), "too expensive" (n = 1), hives (n = 1), and tachycardia (n = 1).
A one-time administrative "switching cost" was estimated to be approximately 2-4% of the medication cost. The mean dosage of nifedipine GITS across subjects was 50.0 mg/day, while the mean dosage of felodipine ER was 8.6 mg/day; both doses were within their recommended dosage ranges. The mean cost of nifedipine GITS was $1.26/day, or $37.77/month. The mean cost of felodipine ER was $1.04/day, or $31.14/month. Felodipine ER charges represented approximately 85% of nifedipine GITS charges, resulting in a 15% savings to the health care system.
Discussion
A recent prospective study, involving a similar formulary switch at a VA hospital showed felodipine ER treatment to be therapeutically equivalent to nifedipine GITS therapy in a group of 87 elderly patients, most of whom were Caucasian men.5 Likewise, this retrospective analysis of a formulary transition affecting members of a large metropolitan HMO demonstrated no significant changes in blood pressures or increases in adverse effects when nifedipine GITS was replaced with felodipine ER. The 246 patients in this analysis are typical of patients receiving treatment for hypertension. African-American individuals are disproportionately represented in both this research sample (37%) and the population of patients suffering from hypertension. The height and weight data suggest that our group was more overweight than the population in general, which is also typical of the hypertensive population. The gender distribution is fairly balanced for a cross section of hypertensive patients in this age group, since the male-to-female ratio was 44:56 (107 men; 139 women).
Data on before and after blood pressure values and frequencies with which hypertension continues to be a primary diagnosis suggest that felodipine ER is at least as effective in controlling blood pressure as nifedipine GITS. Indeed, of the 28% of all patients identified as hypertensive preconversion, approximately one fourth of these patients experienced better blood pressure control with felodipine ER.
One of the primary justifications for choosing felodipine ER as a formulary alternative to nifedipine GITS was its strong safety profile. For the large, heterogenous patient population of a metropolitan HMO, medications with the least probability of exacerbating concomitant conditions are preferred. Clinical trials have established the safety of felodipine in patients with hypercholesterolemia, diabetes mellitus, arthritis, impaired renal function, asthma, and chronic obstructive pulmonary disorders.11-17 Moreover, its lack of a negative inotropic effect on myocardial tissue7-9 may make felodipine ER a better choice than nifedipine GITS for patients with impaired cardiac function.
As seen in previous clinical trials,1,18 the most common adverse effects likely to be associated with felodipine in this trial were edema and headache. Overall, adverse effect data from this trial confirm that the formulary transition introduced no additional safety risks to patients. The frequency of adverse effects potentially related to medication use were similar after formulary transition to those reported before the formulary transition, as confirmed by chi-square analysis (Table 3). No statistically significant differences were found between before and after values.
Finally, the data indicate that the formulary transition provided cost-minimization support for the formulary decision. The cost to the system, per patient per month, for this class of medication was reduced by approximately 15% as a result of the conversion to felodipine ER. Although an administrative "switching cost" was estimated to be 2-4% of medication cost, as a one-time expense, it was offset by the prospect of substantial long-term savings. Thus, this one formulary change, with no additional safety risks or reduction in effectiveness of treatment, resulted in a medication cost savings of more than $150,000 per year to Humana Health Care Plans of Kansas City.
Conclusion
The therapeutic equivalence and comparable safety profile of the two medications combined with a 15% cost savings realized with felodipine ER validated the successful switch from nifedipine GITS to felodipine ER at this health-maintenance organization.
Steven R. Krantz, Ph.D., is Associate Professor, School of Nursing, University of Missouri at Kansas City; Robert S. Rase, Pharm.D., is Clinical Coordinator, Humana Health Systems; and Robert W. Piepho, Ph.D., is Dean and Professor, School of Pharmacy, University of Missouri at Kansas City, Kansas City, Missouri.
Correspondence: Robert W. Piepho, Dean, School of Pharmacy, University of Missouri at Kansas City, 5005 Rockhill Road, Kansas City, MO 64110.
Acknowledgments: To Sheila Owens for assistance with the development of this manuscript.
Copyright © 1996, Academy of Managed Care Pharmacy, Inc. All rights reserved.