Utilization and Cost of Sildenafil in a Large Managed Care Organization With a Quantity Limit on Sildenafil

OBJECTIVES: Erectile dysfunction (ED) affects approximately 30 million men in the United States. The objectives of this study were to (1) assess the cost and utilization of sildenafil citrate (Viagra), an oral therapeutic agent for ED, in a large managed care organization (MCO) with a quantity limit of 6 units per 30-day supply and (2) describe the incidence of comorbid conditions and the severity of cardiovascular disease in adult male users of sildenafil. METHODS: Pharmacy claims for sildenafil were identified from an administrative database of claims with dates of service in calendar year 2001 for male members aged 18 years or older. Medical claims for MCO members who had sildenafil claims were used to identify comorbid diseases and categorize patients by degree of cardiovascular risk. High risk was defined as having at least 1 medical claim with a diagnosis of diabetes mellitus, ischemic heart disease, abdominal aortic aneurysm, or peripheral arterial disease, and medium risk was defined as not having any diagnosis in the high-risk category but at least 1 cardiovascular risk factor that included smoking, hypertension, hypercholesterolemia, family history of premature coronary heart disease, or being aged 45 years or older. RESULTS: There were 67,914 pharmacy claims for sildenafil during 2001 for 20,281 MCO members, an average of 3.3 pharmacy claims per patient. The prevalence of sildenafil use was 54.1 per 1,000 male MCO members aged 18 years or older. The total allowed charges for sildenafil pharmacy claims in 2001 were $3.56 million, of which patients paid 26.6% in average cost-share, and the net MCO cost per member per month (PMPM) was $0.18. A total of 1,681 patients (8.3%) exceeded their quantity restrictions for sildenafil tablets in 2001, of which 1,362 (81.0%) paid cash and 319 (19.0%, or 1.6% of all sildenafil users) appealed and received approval from the MCO for additional sildenafil tablets beyond the restriction of 6 tablets per month. Medical claims were available for 15,644 sildenafil patients (77.1%), and 12,720 sildenafil users (81.3% of those with medical claims) were judged to be at high or medium cardiovascular risk. CONCLUSIONS: A quantity limit of 6 tablets of sildenafil per 30-day period was associated with a drug cost to users and the MCO of $0.25 PMPM. Sildenafil users paid an average cost-share of 26.6%, resulting in a net drug cost of $0.18 PMPM to the MCO.

muscle throughout the body but in much higher concentrations in the vascular smooth muscle of the penis. 9,10 Sildenafil has been shown in clinical trials to be an effective and well-tolerated treatment in patients with ED, including patients with comorbid medical conditions. [11][12][13][14] In a meta-analysis of 27 trials involving 6,659 men, Fink et al. found that sildenafil improves erectile function and is well tolerated by men of varying health status. 15 In a double-blind, placebo-controlled study of sildenafil in men with ED and clinically stable coronary artery disease, DeBusk et al. reported that more patients taking sildenafil had improved erections (64% versus 21%) and improved intercourse (65% versus 19%) than those taking placebo, with no serious drugrelated cardiovascular consequences. 16 In a retrospective analysis of data from multiple efficacy and safety studies of sildenafil in patients with ED and ischemic heart disease who were not taking nitrates, Conti et al. reported improved erections in patients taking sildenafil compared with those taking placebo (70% versus 20%). 6

Costs and Economics of Sildenafil
The introduction of sildenafil for ED and the growing public awareness of ED resulted in an 84% increase in the number of U.S. men seeking and using treatments for ED from 1998 through 2002. 17 Within 2 years of the introduction of sildenafil, the number of patient visits for the chief complaint of ED increased in Mexico (279%), the United States (250%), the United Kingdom (103%), Spain (90%), and Germany (55%). 18 Wilson et al. attributed the rising cost of managing ED in the United Kingdom to a 3-fold increase in the number of men visiting their general practitioners for ED, many of whom are referred to specialists. 19 Health plans have been concerned about the impact sildenafil will have on their pharmacy budgets because of the nature of the disease it treats, namely, that ED is a self-reported condition, and, for this reason, potentially large numbers of men could seek to obtain sildenafil for self-reported ED. 20,21 Several studies, however, have shown that sildenafil has not had the adverse economic effect initially anticipated. For example, Smith et al. found the cost-effectiveness of sildenafil treatment to be comparable to accepted therapies for other medical conditions, such as cholesterol-lowering medications, coronary artery bypass grafting, and renal dialysis. 22 Smith et al. suggested that sildenafil is in the same category as other treatments for non-life-threatening illnesses that affect only quality of life and are covered by insurance, such as migraine headaches.
Although there are no articles published on evaluation of the costs associated with adding sildenafil coverage in a managed care drug benefit plan, 3 studies, in abstract form, found only lower than expected pharmacy benefit costs. [23][24][25] Lehman and Duttagupta discovered that the drug costs per member per month (PMPM) of adding sildenafil coverage to 4 health plans with 93,000 to 15 million members ranged from $0.04 to $0.21, much less than the predicted estimate of $1.00. 23 In a study evaluating PMPM drug costs of sildenafil in managed care organizations (MCOs) that did not restrict the quantity of tablets dispensed, Cherayil and Duttagupta found that actual PMPM costs ($0.03 to $0.24) were also significantly lower than the projections. 24 When MCOs did impose restrictions on the number of sildenafil tablets allowed per prescription, but without requirement for prior authorization for sildenafil, drug costs were also lower than expected, at $0.07 to $0.18 PMPM. 25

Managed Care Interventions
Among the methods employed by MCOs to control pharmacy costs is exclusion from coverage, imposition of quantity limits, or higher copayments. In a study examining which factors MCOs use to make drug coverage decisions, Titlow et al. concluded that value judgments, rather than cost, seemed to play a central, though largely unspoken, role in drug coverage decisions. 21 For sildenafil in particular, Titlow et al. discovered, among 53 organizations surveyed, that the most common method of controlling sildenafil cost was by limiting the quantity of medication covered or the duration of its use. Sixty-four percent of MCOs placed limits on sildenafil coverage, 23% did not cover treatment for sexual dysfunction at all, 21% required prior authorization, and only 2% of MCOs covered sildenafil without any restrictions. 21 Other research has found the quantity limit for sildenafil to range from 4 to 12 tablets per month. 18,26,27 The present study analyzes sildenafil utilization and cost associated with a quantity limit of 6 units per 30-day supply and describes the incidence of comorbid conditions and the severity of cardiovascular disease in sildenafil users in a large MCO with 1.2 million pharmacy lives.

ss Methods
Pharmacy claims data for sildenafil for calendar year 2001 were obtained from the pharmacy benefit management company of a large MCO with 1.2 million pharmacy lives located in the mid-Atlantic states. Medicare beneficiaries, who comprised 2.5% of the MCO population, were eligible for a senior pharmacy benefit drug program, which had an annual benefit maximum of $1,000. At the time of this study, sildenafil was on the second tier of the drug formulary and was restricted to a maximum of 6 tablets per month (or up to 18 tablets for a 3-month supply). There was a gender edit that permitted only male members to receive sildenafil. The refill-too-soon edit was 75%, meaning that a covered member could not obtain a refill of sildenafil until at least 75% of the days in the period had transpired, 23 days for a 30-day pharmacy claim or 68 days for a 90-day pharmacy claim for sildenafil. Physicians could appeal sildenafil claim denials, and decisions were made on a case-by-case basis since there were no specific criteria for medical exception. The study cohort was composed of men aged 18 years or older with continuous pharmacy benefits for calendar year 2001 and with at least 1 pharmacy claim for sildenafil during the calendar year.
In the pharmacy claims database, there were both positive and negative (reversed) claims. Pharmacists generate a positive claim when tablets are dispensed by the pharmacy, and a negative claim occurs when the filled prescription is reversed. Reasons for negative claims include an error in the submission of the electronic claim or if the member does not pick up the prescription. All positive and negative matched claim pairs were deleted to ensure that only prescriptions received by members were included in the data used in this study.
The data set contained the following fields: unique generator member identification, which was different from the member' s actual identification to protect member confidentiality; age as of January 1, 2001; amount paid by the MCO; amount of member copayment; drug name; dispense date; dose; number of tablets dispensed; and the days supply (number of drug therapy days submitted by the pharmacist on the pharmacy claim).
Members who received sildenafil were grouped into high-, medium-, and low-cardiovascular-risk categories based on an adaptation of classification criteria in the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) guideline. 28 The high-risk category consisted of patients with at least 1 of the following diagnoses: diabetes mellitus, ischemic heart disease, abdominal aortic aneurysm, or peripheral arterial disease. Patients in the medium-risk category were those who did not have any diagnoses from the high-risk category but had at least 1 of the following cardiovascular risk factors: smoking, hypertension, hypercholesterolemia, family history of premature coronary heart disease, or were aged 45 years or older. The remaining patients were placed into the low-risk category.

ss Results
A total of 67,914 prescription claims for sildenafil occurring during calendar year 2001 for 20,281 patients were available for analysis. The prevalence of sildenafil use was 54.1 per 1,000 male MCO members aged 18 years or older. The mean ± standard deviation (SD) age of patients in the cohort was 53.1 ± 10.4 (median 53) years. Most of the patients in the cohort were between age 50 and 59 years (38.5%) ( Table 1). There were 2,559 sildenafil patients (12.6%) enrolled in a senior pharmacy benefit program for Medicare beneficiaries with a $1,000 annual benefit maximum.
The number of sildenafil prescriptions filled per utilizing member for this MCO cohort during 2001 was 3.3 ± 2.7 (mean ± SD; median = 2), with a range of 1 to 29 prescriptions (Table 2). Most prescriptions (85%) were for 6 tablets at a time (6.3 ± 2.4; median = 6), with a range of 1 to 100 tablets. Some (9.3%) prescriptions were for fewer than 6 tablets, and some were for 7 or more tablets (3.4% for 7 to 11 tablets and 2.5% for 12 or more tablets).
Median annual sildenafil utilization for calendar year 2001, extrapolated from partial-year use, was 29.4 tablets per utilizing member per year, or 2.5 tablets per month, which includes both cash and MCO-paid pharmacy claims. There were 933 pharmacy claims (1.4%) for 6,127 tablets of the 25 mg dose of sildenafil, whereas the remainder of the prescriptions was evenly split between 33,208 pharmacy claims (207,994 tablets) for the 50 mg dose (48.9%) and 33  More than half of the 18,899 members had their first sildenafil prescription for the year filled in the first quarter of 2001 (9,722 members, 51.4%). Successively lower numbers of members filled their first prescription in the second (3,620 members, 19.1%), third (2,925 members, 15.5%), and fourth quarters (2,632 members, 13.9%), respectively.

Medical Claims Data and Member Comorbidity
Among the 20,281 patients who had sildenafil claims in 2001, there were 15,644 patients (77.1%) who had at least 1 medical claim for a comorbid disorder. The mean ± SD age of members in this group was 53.4 ± 9.9 (median 54) years. A total of 135,298 medical claims (average 8.65 per patient) were included in the evaluation. About 81% of these sildenafil users were classified as either high risk or medium risk for cardiovascular events based on an adaptation of the risk factor classification outlined by the NCEP ATP III guideline. 28 Hypertension (37%), dyslipidemia (36%), and diabetes mellitus (18%) were the most common comorbid conditions. (Table 3) ss Discussion Our study found that most (85%) members from the MCO filling sildenafil prescriptions were between the ages of 40 and 69 years, with 10% of members younger than 40 years and 5% of members 70 years or older. These results are consistent with other published research describing the characteristics of sildenafil users. 29 In a study evaluating sildenafil prescribing practices immediately after its approval, Harrold et al. found that patients prescribed sildenafil tended to remain on the initial dose they were prescribed, with little reported dose titration. 29 The majority (59%) of patients were initially prescribed the 50 mg dose, with 35% prescribed an initial dose of 25 mg and 7% prescribed an initial dose of 100 mg. In the present study, only 1.4% of pharmacy claims were for the 25 mg dose of sildenafil, and about one half (48.9%) of all sildenafil pharmacy claims were for the 50 mg dose and one half (49.7%) were for the 100 mg dose.
Our median annual sildenafil utilization for calendar year 2001 was 29.4 tablets per year, or 2.5 tablets per month, which corresponds to results from a study by Delate et al., 17    tablets dispensed over 9 months were 23.3 ± 20.5, which corresponds to approximately 2.6 tablets per month. Sildenafil accounted for about 0.5% of the annual pharmacy budget, with an MCO drug cost of $0.18 PMPM in 2001. This $0.18 PMPM drug cost for sildenafil falls in the range of the 3 poster abstracts reported previously ($0.03-$0.24 PMPM). [23][24][25] A total of 8.3% of members exceeded their quantity restrictions for sildenafil tablets in 2001; 6.7% paid cash for the additional tablets, whereas 1.6% appealed and received approval for additional tablets from the MCO beyond the restriction of 6 tablets per month. Quantity restrictions did not appear to negatively impact most members because they did not maximize their benefit of 6 tablets per month. Previous research found that just over 50% of patients who use sildenafil for the first time respond on the first attempt at sexual intercourse, but at least 6 attempts are required before an 80% success rate is achieved. 30,31 Educating patients regarding expectations and the possibility of increasing the dose is important to treatment success. Atiemo et al. found that 42% of prior sildenafil nonresponders achieved success after reeducation, with 81% of the failures being attributed to incorrect administration. 32 About 94% of the patients had a sustained response at 26 months.

Frequency Distribution of Sildenafil Claims Among Users in 2001
We were able to determine comorbidity occurrence for members using sildenafil who had medical coverage and at least 1 medical claim. We found hypertension, dyslipidemia, and diabetes mellitus to be the top 3 comorbid diagnoses in men prescribed sildenafil. The results from the present study are consistent with other published research describing the characteristics of sildenafil users. Harrold et al. found that 49% of plan members from a Massachusetts health maintenance organization who received sildenafil also had hypertension, 42% had hyperlipidemia, 33% were receiving a medication associated with ED as a possible side effect (such as beta-blockers, diuretics, digoxin, antipsychotic agents, and others), 25% had diabetes mellitus, 16% had ischemic heart disease, and 5% had a history of radical prostatectomy. 29 In the present study, 81% of members in whom comorbid health conditions could be determined, were classified as either high risk or medium risk for cardiovascular conditions. Previous studies have shown a strong link between ED and many of the conditions for which members in our cohort accessed medical care. 3,4,6,16 In fact, Johannes et al. observed that the age-adjusted risk for ED was higher for men with diabetes, heart disease, and hypertension, 4 which supports our data obtained from the medical claims.
Scope of drug coverage and the amount of member costshare are important factors in member satisfaction with the pharmacy benefit plan. Motheral and Heinle concluded that about 75% of respondents felt that their out-of-pocket copayment was the most important feature of their prescription benefit plan, with an additional 20% of respondents listing copayment as the second most important factor. 33 Other factors considered included the list of drugs covered by the plan, having a mail-order option, which pharmacies accepted the plan, and getting help with questions or problems with drug coverage. In a survey evaluating member satisfaction with their phamacy benefit plans, Desselle found that total out-of-pocket cost was the second most important factor out of 9 total factors rated by respondents. 34 The only factor that ranked higher for plan satisfaction was the list of drugs on the formulary. The amount of medication (days supply) the plan allows for each pharmacy visit was ranked the third most important factor.

Limitations
The data from the pharmacy benefit manager did not provide an indication or diagnosis for the use of sildenafil, which could affect interpretation of utilization patterns, such as the off-label use of sildenafil for primary pulmonary hypertension, 35 and we did not assess the medical claims to determine diagnosis information for this purpose. Second, as with any analysis of pharmacy claims data, there was no way to determine whether the patient used the medication, only that the prescription was filled and the tablets were received by the patient.
Third, we could not determine from the database if cash payment for sildenafil prescriptions was due to exhaustion of benefits for those Medicare beneficiaries with an annual benefit maximum or by sildenafil users in excess of the quantity limit of 6 tablets per month. Of the 1,362 patients who paid cash for sildenafil, 136 (10.0%) were patients aged 65 years or older. We estimated but could not verify that 53 of these seniors exceeded their $1,000 annual benefit maximum, accounting for 3.9% of the patients who paid cash for sildenafil prescriptions. Conversely, approximately 96% of the cash prescriptions for sildenafil were assumed to be from patients who had exceeded the limit of 6 tablets per 30-day period.
We also could not determine how often members simply purchased sildenafil prescriptions outside of the pharmacy claims system due to the use of sildenafil in excess of the quantity limit. Our data may further underreport the actual use of sildenafil since we could not account for samples provided by physicians to patients during medical encounters.
Because the data collection period was limited to 2001, it was not possible to determine duration of sildenafil use or to categorize patients as new or established sildenafil users. Drug use patterns may differ between new and established sildenafil users, and sildenafil had been available for about 3 years at the time of the study. The population for this MCO probably included both members initiating and continuing therapy with sildenafil, as might be expected for other MCO populations.
Our attempt to determine comorbid conditions in the study population was limited since medical claims were not found in 2001 for all sildenafil users; for example, a sildenafil user could have had a medical encounter prior to the 2001 study period. The medical record information was also limited by the maximum of 3 diagnosis codes per medical claim and may therefore be incomplete. Other reasons for inaccurate or incomplete coding were possible, including accidental or intentional miscoding.
Lastly, at the time of our study, sildenafil was the only phosphodiesterase type 5 inhibitor on the market, and it may not be reasonable to extrapolate our findings to the current environment in which other ED therapies are now available. A recent clinical monograph on ED therapies was published in the Journal of Managed Care Pharmacy. 36

ss Conclusions
Our study found that the majority of members who used sildenafil were between the ages of 40 and 69 years and had a medium to high risk for a cardiovascular event based on an adaptation of the classification system in the NCEP ATP III guideline for treatment of dyslipidemia. A quantity limit of 6 tablets of sildenafil per 30-day period was associated with a drug cost to users and the MCO of $0.25 PMPM. Sildenafil users paid an average cost-share of 26.6%, resulting in a net drug cost of $0.18 PMPM to the MCO. The impact of sildenafil on the MCO' s pharmacy budget was 0.5%, and 91.7% of members did not exceed their sildenafil quantity restriction.