Community and Mail Service Pharmacy

RESEARCH Comparison of Costs of Community and Mail Service Pharmacy Norman V. Carroll, Ilia Brusilovsky, Bryan York, and Robert Oscar ABSTRACT Objective: To compare the costs of prescriptions dispensed through mail service and community pharmacies to quantify the comparative costs of the two types of pharmacies. Design: Modeling study based on 1-year of claims data from a cohort of patients. Setting: A health plan in the northeastern United States. Patients: Approximately 100,000 members of the health plan. Intervention: The plan used a small pharmacy benefits manager (PBM) and a mail service pharmacy that was not owned by a major PBM, a three-tier benefit design, and specified that patients could get a 90-day supply through mail service for the equivalent of two 30-day community pharmacy copayments. Main Outcome Measures: Actual total, ingredient, plan, and patient costs of prescriptions dispensed through a mail service pharmacy and the estimated costs of those same prescriptions at community pharmacies. Results: Total costs for the 44,847 prescriptions dispensed through mail service were $6,401,624. Had these prescriptions been dispensed at community pharmacies, costs would have been $6,902,252. Ingredient costs were $6,401,624 through mail versus $6,633,170 at community pharmacies. Total costs to the health plan were $4,726,637 through mail versus $4,417,733 at community pharmacies. Member costs were $1,674,987 through mail versus $2,484,519 at community pharmacies. Conclusion: Compared with community pharmacies, the mail service pharmacy was less expensive overall, less expensive for patients, but more expensive to the health plan. From the health plan s perspective, the loss of copayments in the mail service benefit was greater than the savings on ingredient costs and dispensing fees. Keywords: Economics, cost analysis, cost containment, health insurance, pharmacy benefits management. J Am Pharm Assoc. 2005;45:336 343. Received June 14, 2004, and in revised form August 27, 2004. Accepted for publication October 30, 2004. Norman V. Carroll, PhD, is Professor, Pharmacy Administration, School of Pharmacy, Virginia Commonwealth University, Richmond, Va. Ilia Brusilovsky, BA, is a student in the School of Medicine, Virginia Commonwealth University; at the time of the study, he was Development Analyst, RxEOB, Richmond, Va. Bryan York, BS, is Senior Partner, and Robert Oscar, BPharm, is President, RxEOB, Richmond, Va.. Correspondence: Norman V. Carroll, PhD, School of Pharmacy, Virginia Commonwealth University, P.O. Box 980533, Richmond, VA 23298-0533. Fax: 804-828- 8359. E-mail: nvcarroll@vcu.edu Disclosure: Dr. Carroll serves as a consultant to RxEOB. Apart from this, the authors declare no conflicts of interest or financial interests in any products or services mentioned in this article, including grants, employment, gifts, stock holdings, or honoraria. Previously presented in part at the International Society of Pharmacoeconomics and Outcomes Research Ninth Annual Meeting, Arlington, Va., May 19, 2004. 336 Journal of the American Pharmacists Association www.japha.org May/June 2005 Vol. 45, No. 3

Community and Mail Service Pharmacy Costs RESEARCH Expenditures on prescription drugs in noninstitutionalized settings have grown from $40.3 billion in 1990 to $141 billion in 2001. 1,2 Health plans, which now pay for more than 70% of prescriptions dispensed, 3 have increasingly turned to mail service pharmacies as one means of controlling prescription drug spending. As a result, the percentage of total U.S. prescription sales made through mail service pharmacies grew from 5.1% in 1990 to 14.1% in 2002. 3,4 Mail service pharmacies offer the potential for cost savings for chronic medications on both ingredient costs and dispensing fees. Mail service pharmacies may be able to fill prescriptions at lower costs, and thereby lower dispensing fees, because of efficiencies of operation. 5 9 Most mail service pharmacies are high volume, highly automated operations. Mail service pharmacists may also be more efficient because they are not subject to interruptions from telephone calls and patients questions as are community pharmacists. In addition, mail service plans typically dispense 3-month AT A GLANCE Synopsis: Mail service pharmacy was more expensive for health plan sponsors than community pharmacy would have been had it been used, according to this cost analysis of 44,847 mail service prescription claims processed by a northeastern U.S. health plan. While overall costs were lower for mail service prescriptions, the loss of copayment income for the mail service plan was greater than the savings realized by the plan: the plan saved 7.8% in overall costs but patient copayments were 12.7% lower. A breakeven analysis indicated that these results would not have been altered by differences in ingredient cost reimbursements, generic dispensing rates, or wastage rates typically observed in the market. Analysis: Mail service pharmacies are used by health plans as one means of controlling the rising costs of prescription medications. Although mail service pharmacies can provide lower unit costs for prescriptions, health plans may not realize these lower costs. Health plans provide incentives to get patients to use mail service pharmacies, such as a 3-month supply of medication for a 2-month copayment. When monthly copayments averaged $5 to $10, loss of copayment income had little effect on benefit plan costs. With the larger copayments now typical of benefit plans, loss of copayment income can result in higher realized plan costs, as demonstrated in this study. Reducing or eliminating incentives to use mail service pharmacies may lower plan costs but might not produce savings. Patients have been shown to prefer using community pharmacies if not offered an economic incentive to use mail service pharmacies. supplies of medications. Community pharmacies are usually limited to dispensing 1-month supplies. Because the pharmacy charges the pharmacy benefits manager (PBM) a set dispensing fee each time a prescription is filled, dispensing a greater months supply minimizes the total number of prescriptions dispensed and, consequently, the total amount paid by the sponsor for dispensing fees. Mail service pharmacies may also be able to lower ingredient costs because they receive larger product discounts from manufacturers than do community pharmacies. 5,7 10 Mail service pharmacies can provide lower unit costs for prescriptions, but this does not necessarily mean that health plans actually realize lower costs by using mail service pharmacies. Health plans typically provide patients with economic incentives to use mail service pharmacies. These incentives most commonly consist of allowing patients to receive a 3-month supply of medication from the mail service pharmacy for the price of 2 monthly copayments. To receive an equivalent supply of medication from the community pharmacy, the patient would have to pay 3 monthly copayments. When copayments averaged $5 $10 a month, the loss of one copayment on a 90-day supply of medication was not a major economic concern for health plans. However, with copayments now averaging $15 $20 per 30 day supply, 11 and growing annually, the loss of one copayment per 90-day mail service prescription has become a major cost to health plans. As a result, the economic incentive for health plans to use mail service pharmacies has been reduced. Objective The objective of this study was to quantify and compare the costs of mail service and community pharmacies. We compared total costs, ingredient costs, costs to the health plan, and costs to the patient. Methods We compared the actual cost of prescriptions dispensed through a mail service pharmacy with what those same prescriptions would have cost if they had been dispensed through community pharmacies. This was done by applying community pharmacy pricing to mail service prescriptions while taking into account the copayment structure of the plan, then comparing actual mail service costs with estimated community pharmacy costs for the same prescriptions. We based our analysis on prescription claims submitted to a health plan located in the northeastern United States between July 1, 2002, and June 30, 2003. The plan provided coverage to about 100,000 members. For the year, a total of 73,567 mail service claims and 1,130,768 community pharmacy claims were submitted. The health plan used a small PBM and a mail service pharmacy that was not owned by a major PBM. The plan specified that no more than a 30-day supply could be dispensed through community Vol. 45, No. 3 May/June 2005 www.japha.org Journal of the American Pharmacists Association 337

RESEARCH Community and Mail Service Pharmacy Costs pharmacies and that mail service pharmacies would dispense 90- day supplies. The plan paid a $2.00 dispensing fee for each prescription dispensed through community pharmacies; no dispensing fee was paid for prescriptions dispensed by the mail service pharmacy. For brandname drugs, the plan paid community pharmacies the average wholesale price (AWP) less 15% and paid the mail service pharmacy AWP less 17%. Generic drugs were priced according to a proprietary maximum allowable cost (MAC) schedule. Administrative fees were the same for community pharmacy and mail service prescriptions. We selected only those claims from programs with a three-tier benefit design. We did this to eliminate the variability that might be caused by different benefit designs and because the three-tier design was the most common. After eliminating non three-tier designs, 65,947 mail service and 764,289 community pharmacy claims remained. Next, we identified the 201 products for which the most mail service prescriptions were dispensed and selected all mail service claims for these products (Table 1). The top 201 products accounted for 44,847 mail service claims. A product was defined as a drug Table 1. Products a Included in Cost Analysis Accolate 20 mg tablet Accupril 10, 20, and 40 mg Aciphex 20 mg tablet (EC) Actonel 5, 30, and 35 mg Actos 15, 30, and 45 mg Advair 100/50 and 250/50 diskus Aggrenox capsule (XR) Albuterol 90 mcg inhaler Allegra 60 and180 mg ; Allegra-D tablet (XR) Allopurinol 100 and 300 mg Alprazolam 0.5 mg tablet Altace 2.5, 5, and 10 mg capsules Amaryl 2 and 4 mg Ambien 10 mg tablet Amiodarone HCl 200 mg tablet Amitriptyline HCl 25 mg tab Asacol 400 mg tablet (EC) Atacand 16 mg tablet Atenolol 25, 50, and 100 mg Atenolol/chlorthalidone 50/25 tablet Avandia 4 and 8 mg Avapro 150 mg tablet Bextra 10 mg tablet Bisoprolol/HCTZ 2.5/6.25 and 5/6.25 Cartia XT 120, 180, and 240 mg capsules (XR) Celebrex 200 mg capsule Celexa 20 and 40 mg Clarinex 5 mg tablet Clonazepam 0.5 mg tablet Combivent inhaler Cozaar 50 and 100 mg Detrol LA 4 mg capsule (XR) Digitek 125 and 250 mcg Diovan 80 and 160 mg Doxazosin mesylate 2, 4, and 8 mg Effexor XR 75 and 150 mg capsules (XR) Enalapril maleate 5, 10, and 20 mg Estradiol 1 and 2 mg Estratest HS tablet Evista 60 mg tablet Famotidine 20 mg tablet Femhrt 1/5 tablet Flomax 0.4 mg capsule Flonase 0.05% nasal spray Fluoxetine HCl 20 and 40 mg capsules Folic acid 1 mg tablet Fosamax 70 mg tablet Furosemide 20 and 40 mg Gemfibrozil 600 mg tablet Glucophage XR 500 mg tablet (XR) Glucotrol XL 5 and 10 mg (XR) Glucovance 2.5/500 and 5/500 mg Glyburide 5 mg tablet HCTZ 12.5 mg capsule and 25 and 50 mg Humalog 100 U/mL vial Humulin 70/30 vial Hydroxychloroquine 200 mg tablet Hyzaar 50-12.5 and 100-25 Ibuprofen 800 mg tablet Indapamide 2.5 mg tablet Isosorbide mononitrate 30 and 60 mg (XR) Klor-con 10 meq tablet Levoxyl 25, 50, 75, 88, 100, 112, 125, 150, 175, and 200 mcg Lexapro 10 mg tablet Lipitor 10, 20, 40, and 80 mg Lisinopril 5, 10, 20, and 40 mg Lisinopril/HCTZ 20-12.5 and 20-25 Lorazepam 0.5 and 1 mg Lotrel 5/20 mg capsule Lovastatin 20 mg tablet Medroxyprogesterone 2.5 mg Metformin HCl 500, 850, and 1,000 mg Methotrexate 2.5 mg tablet Metoprolol 50 and 100 mg Miacalcin 200 U nasal spray Naproxen 500 mg tablet Nasonex 50 mcg nasal spray Neurontin 300 mg capsule Nexium 20 and 40 mg capsules Niaspan 500 mg tablet (XR) Nifedical XL 30 and 60 mg (XR) Norvasc 5 and 10 mg Omeprazole 20 mg capsule (DR) Ortho Tri-cyclen 28 tablet Paxil 20 mg tablet Plavix 75 mg tablet Plendil 10 mg tablet (XR) Potassium chloride 10 meq capsule (XR) and 20 meq tablet (XR) Pravachol 20 and 40 mg Prednisone 5 mg tablet Premarin 0.3, 0.625, 0.9, and 1.25 mg Prempro 0.625/2.5 mg tablet Prevacid 30 mg capsule (DR) Prilosec 20 mg capsule (DR) Propoxy-N/APAP 100-650 tab Proscar 5 mg tablet Protonix 40 mg tablet (EC) Ranitidine 150 mg tablet Rhinocort aqua nasal spray Serevent 21 mcg inhaler Singulair 10 mg tablet Spironolactone 25 mg tablet Synthroid 100 mcg tablet Tamoxifen 20 mg tablet Terazosin 2, 5, and 10 mg capsules Timolol 0.5% gel/solution Toprol XL 25, 50, and 100 mg (XR) Trazodone 50 and 100 mg Triamterene/HCTZ 37.5/25 and 75/50 Tricor 54 and 160 mg Vagifem 25 mcg vaginal Verapamil 180 and 240 mg (XR) Vioxx 25 mg tablet Warfarin sodium 5 mg tablet Wellbutrin SR 150 mg tablet (XR) Xalatan 0.005% eye drops Yasmin 28 Zocor 10, 20, 40, and 80 mg Zoloft 50 and 100 mg Zyrtec 10 mg tablet Abbreviations used: APAP, acetaminophen; DR, delayed release; EC, enteric coated; HCl, hydrochloride; HCTZ, hydrochlorothiazide; SA, sustained action; XR, extended release. a A product was defined as a drug with a specific name, strength, and dosage form. All generically dispensed products with the same name, strength, and dosage form were considered the same product. 338 Journal of the American Pharmacists Association www.japha.org May/June 2005 Vol. 45, No. 3

Community and Mail Service Pharmacy Costs RESEARCH with a specific name, strength, and dosage form. All generically dispensed products with the same name, strength, and dosage form were considered the same product. For example, all prescriptions dispensed generically for atenolol 100 mg were considered to be the same product. However, a product dispensed for a brandname drug was considered to be a different product than its generic equivalent. That is, prescriptions dispensed generically as atenolol 100 mg were considered to be different products than those dispensed for AstraZeneca Pharmaceuticals Tenormin 100 mg even though atenolol is the active ingredient in Tenormin. We then matched all mail service claims for the top 201 products with all community pharmacy claims dispensed for the same product on the same date. Claims were matched by date to control for price changes over time. Community pharmacy claims without a matching mail service claim were discarded. Each mail service claim had at least one community pharmacy match. The 44,847 mail service claims had 206,258 matching community pharmacy claims. From this sample of claims, we calculated, for each product, the total number of mail service prescriptions dispensed, the mean mail service ingredient cost per metric unit, mean metric mail service quantity per prescription, and the mean mail service copayment. (It was necessary to calculate the mean copayments because there were a number of different copay structures within the plan. The mail service copayments for 90-day supplies by tier varied from a low of $10/$20/$50 to a high of $24/$60/$100.) From these we then calculated the total mail service cost, total cost to the health plan, total ingredient cost, and the total cost to members of the plan for each of the top 201 products. These were calculated as follows: Total mail service cost = No. mail service prescriptions (mean mail service ingredient cost per metric unit mean mail service quantity per prescription) Total mail service cost to the health plan = No. mail service prescriptions (mean mail service ingredient cost per metric unit x mean mail service quantity per prescription mean mail service copayment) Total mail service ingredient cost = No. mail service prescriptions (mean mail service ingredient cost per metric unit mean mail service quantity per prescription) Total mail service cost to members = No. mail service prescriptions mean mail service copayment From the sample of matched community pharmacy claims, we calculated the mean community pharmacy ingredient cost per metric unit and mean community pharmacy copayment for each prescription. From these, we then calculated what each mail service prescription would have cost had the prescription been dispensed at community pharmacies. These were calculated as follows: Total community pharmacy cost = No. mail service prescriptions (mean community pharmacy ingredient cost per metric unit mean mail service quantity per prescription + 3 community pharmacy dispensing fee) Total community pharmacy cost to the health plan = No. mail service prescriptions (mean community pharmacy ingredient cost per metric unit mean mail service quantity per prescription + 3 community pharmacy dispensing fee 3 mean community pharmacy copayment Total community pharmacy ingredient cost = No. mail service prescriptions mean community pharmacy ingredient cost per metric unit mean mail service quantity per prescription Total community pharmacy cost to members = No. mail service prescriptions 3 mean community pharmacy copayment Mail service quantities dispensed were typically for 3 months, whereas community pharmacy quantities were typically for 1 month. Because of this, we multiplied community pharmacy copayments and dispensing fees by 3 to make them comparable with what would have been charged had 3-month supplies been dispensed at community pharmacies. Both mail and community pharmacy copayments were adjusted for situations in which the total cost of the prescription was less than the copayment. In these situations, the copayment amount was set to the total cost of the prescription. We then totaled the sums for the 201 products to determine the total amounts that were actually spent for mail service and the total amounts that would have been spent had these prescriptions been dispensed through community pharmacies. Calculations were made for the total sample of prescriptions, then separately for all prescriptions dispensed as generic products and for all prescriptions dispensed as brandname products. Statistical significance of differences between mail service and community pharmacy costs was assessed by examination of 95% confidence intervals. We conducted break-even analyses to determine the levels at which differences in reimbursement for ingredient costs for brandname drugs, generic dispensing rates, and waste between mail service and community pharmacy would substantially affect the results. Results A description of the age and gender distribution of plan participants is shown in Table 2. The results of the cost comparison are shown in Table 3. All differences between mail service and community pharmacy costs were statistically significant at P <.05. The total cost of the 44,847 mail service claims, as dispensed through the mail service pharmacy, was $6,401,624. Had these claims been dispensed at community pharmacies, the cost would have been $6,902,252. Thus, the total cost of these claims was $500,628 (or 7.8%) more expensive when dispensed through community pharmacies. This difference occurred because community pharmacies charged $269,082 (4.2%) more for dispensing fees and $231,546 (3.6%) more in ingredient costs than did the mail service pharmacy. While total costs were lower through mail service, costs to the health plan were higher. Total costs to the plan were $4,726,637 Vol. 45, No. 3 May/June 2005 www.japha.org Journal of the American Pharmacists Association 339

RESEARCH Community and Mail Service Pharmacy Costs Table 2. Age and Gender Distribution of Prescription Recipients in Health Plan a All Plan Patients Study Patients Receiving Using Mail Service Prescriptions Characteristic No. (%) No. (%) Gender Men 3,362 (39) 40,567 (40) Women 5,239 (61) 59,874 (60) Total 8,601 (100) 100,441 (100) Age Under 18 years 319 (4) 33,696 (34) 18 24 years 177 (2) 8,600 (9) 25 34 years 343 (4) 12,902 (13) 35 44 years 870 (10) 13,694 (14) 45 54 years 1,766 (21) 12,111 (11) 55 64 years 1,659 (19) 6,939 (7) 65 years or older 3,461 (40) 12,405 (12) Total 8,595 (100) 100,347 (100) a Totals are not equal because of missing data. through the mail service pharmacy and would have been $4,417,733 through community pharmacies. Thus, the plan would have paid $308,904 (4.8%) less had the prescriptions been dispensed through community pharmacies. However, plan members would have paid $809,532 (12.7%) more in copayments if community pharmacies had been used. We also compared mail and community pharmacy costs separately for brandname (Table 4) and generic products (Table 5). Total community pharmacy cost for brandname drugs was $300,482 (5.4%) more than total mail service cost. Community pharmacy was $136,490 (2.4%) more expensive for ingredient costs and $163,992 (2.9%) more for dispensing fees. Total health plan costs for brandname drugs were $350,172 (8.3%) less expensive at community pharmacies than mail service. Higher community pharmacy costs for ingredient costs and dispensing fees were less than the $650,654 (11.9%) difference in patient copayments. Total community pharmacy costs for generic drugs were $200,146 (24.2%) more than total mail cost. Community pharmacy was $95,056 (11.5%) more expensive for ingredient costs and $105,090 (12.7%) more for dispensing fees. Health plan costs for generic drugs were $41,269 (5.0%) more at community pharmacies than mail service. Higher community pharmacy costs for ingredients and dispensing fees exceeded the $158,877 (19.2%) difference in copayments. There was a 2 percentage point difference in the AWP discount on brandname drugs between mail and community pharmacies for the health plan. The average AWP difference for employer-sponsored plans in 2001 was about 5 percentage points. 11 A break-even analysis indicated that the AWP discount difference had to be at least 6.6 percentage points for dispensing through mail service pharmacies to cost plan sponsors the same amount as dispensing through community pharmacies (Table 6). At AWP discount differences less than 6.6%, community pharmacies were less expensive for the plan sponsor. The design of our study assumed the same generic dispensing rate of 39.1% for both mail and community pharmacies. Breakeven analysis indicated that the community pharmacy generic dispensing rate would need to be 46.0% for total cost of dispensing Table 3. Comparison of Mail Service and Community Pharmacy Costs for All Prescriptions (n = 44,847) Mail Service Pharmacy Community Pharmacy Mean a ± SD, $ Total, $ Mean a ± SD, $ Total, $ Total costs 142.74 ± 108.34 6,401,624 153.91 ± 110.01 6,902,252 Ingredient costs 142.74 ± 108.34 6,401,624 147.91 ± 110.01 6,633,170 Dispensing fee 0 0 6.00 ± 0 269,082 Cost to health plan 105.39 ± 95.58 4,726,637 98.51 ± 92.17 4,417,733 Patient copayment 37.35 ± 20.06 1,674,987 55.4 ± 28.37 2,484,519 a All means were significantly different at P <.05. Table 4. Comparison of Mail Service and Community Pharmacy Costs for Prescriptions Filled Using Brandname Drugs (n = 27,332) Mail Service Pharmacy Community Pharmacy Mean a ± SD, $ Total, $ Mean a ± SD, $ Total, $ Total costs 203.94 ± 90.52 5,574,122 214.94 ± 92.38 5,874,604 Ingredient costs 203.94 ± 90.52 5,574,122 208.94 ± 92.38 5,710,612 Dispensing fee 0 0 6.00 ± 0 163,992 Cost to health plan 152.81 ± 89.28 4,176,578 140.00 ± 89.60 3,826,406 Patient copayment 51.13 ± 12.27 1,397,544 74.94 ± 17.50 2,048,198 a All means were significantly different at P <.05. 340 Journal of the American Pharmacists Association www.japha.org May/June 2005 Vol. 45, No. 3

Community and Mail Service Pharmacy Costs RESEARCH Table 5. Comparison of Mail Service and Community Pharmacy Costs for Prescriptions Filled Using Generic Drugs (n = 17,515) Mail Service Pharmacy Community Pharmacy Mean a ± SD, $ Total, $ Mean a ± SD, $ Total, $ Total costs 47.25 ± 48.00 827,502 58.67 ± 52.80 1,027,648 Ingredient costs 47.25 ± 48.00 827,502 52.67 ± 52.80 922,558 Dispensing fee 0 0 6.00 ± 0 105,090 Cost to health plan 31.40 ± 44.41 550,058 33.76 ± 48.41 591,327 Patient copayment 15.84 ± 6.01 277,444 24.91 ± 7.58 436,321 a All means were significantly different at P <.05. through community pharmacies to be the same as total cost of dispensing through mail service pharmacies (Table 7). This suggests that community pharmacy would be a less expensive option for both plan sponsors and in terms of total costs if community pharmacies generic dispensing rates exceeded those of mail service pharmacies by more than 7 percentage points. The available evidence suggests that the wastage rate for mail service pharmacies is about two times greater than that for community pharmacies. 10 Break-even analysis indicated that dispensing through community pharmacies would yield lower total cost, in addition to lower cost to the plan sponsor, if the mail wastage Table 6. Break-even Analysis for Ingredient Cost Reimbursement Difference for Brandname Drugs At what AWP discount difference would dispensing through mail service pharmacies equal the cost to plan sponsors of dispensing through community pharmacies? Data needed for calculation Mean AWP ingredient cost for brandname drugs was $246. A total of 27,332 brandname prescriptions were dispensed. Dispensing generic drugs through mail service pharmacies was $41,269 less expensive to the plan than was dispensing through community pharmacies. Mail service pharmacy s cost advantage was a greater AWP discount. Community pharmacy s cost advantage was the difference in patient copayments ($23.81) less the dispensing fee ($6.00) Formula and calculation of break-even point The break-even point is the AWP discount difference at which the mail service pharmacy s cost advantage on brandname ingredient costs and generic drugs is equal to community pharmacies cost advantage on patient copayments less the dispensing fee. y is the AWP discount difference between mail service and community pharmacies for brandname drugs No. brandname prescriptions dispensed (mean brandname AWP AWP discount difference y) + savings on generic drugs through mail service pharmacies = number brandname prescriptions dispensed (copayment difference dispensing fee) 27,332 (246 y) + 41,269 = 27,332 (23.81 6.00) y = 0.066, or 6.6% Abbreviation used: AWP, average wholesale price. rate exceeded 16.8% and the community pharmacy wastage rate was 8.4% (Table 8). The available evidence suggests that actual wastage rates are lower than this. 10 Discussion Our results indicated that mail service dispensing was more expensive for this health plan than dispensing through community pharmacies would have been. While the mail service pharmacy generated a 7.8% savings on total costs, it collected 12.7% less in patient copayments. The result was 4.8% higher costs to the health plan. Break-even analyses indicated that differences in ingredient cost reimbursement for brandname drugs, generic dispensing rates, and wastage rates of the sizes typically seen in the market would not substantially alter these results. We could find only one empirical study that compared the cost Table 7. Break-even Analysis for Generic Dispensing Rates At what generic dispensing rate would the total cost of dispensing through community pharmacies equal the total cost of dispensing through the mail service pharmacy? Mean price of brandname prescriptions through community pharmacies $214.94 Mean price of generic prescriptions through community pharmacies $58.67 Mean savings from generic substitution in community pharmacies $156.27 Total costs through community pharmacies $6,902,252 Total cost through mail service pharmacies $6,401,624 Excess cost through community pharmacies $500,628 Divided by mean savings per generic substitution $156.27 Number of additional substitutions required for community pharmacy costs to equal mail service cost 3,204 Add number of generic prescriptions currently dispensed 17,515 Total number of generic prescriptions required for community pharmacy costs to equal mail service cost 20,719 Divided by total number of prescriptions dispensed 44,847 Generic dispensing rate required for community pharmacy costs to equal mail service costs 46.20% Vol. 45, No. 3 May/June 2005 www.japha.org Journal of the American Pharmacists Association 341

RESEARCH Community and Mail Service Pharmacy Costs Table 8. Break-even Analysis for Wastage Rates At what mail service wastage rate would the total cost of dispensing through mail service pharmacies equal the total cost of dispensing through community pharmacies? Data needed for calculation Assume the mail wastage rate was twice that of community pharmacies. Total cost of dispensing through community pharmacies was $500,628 greater than the cost of dispensing through mail service pharmacies in the original analysis. Let z equal the community pharmacy wastage rate and 2z equal the mail service wastage rate Formula and calculation of break-even point 2z (total mail service cost) z (total community pharmacy cost) = $500,628 2z(6,401,624) z(6,902,252) = $500,628 z = 0.084 or 8.4% and 2z = 0.168 or 16.8% of community and mail service pharmacy. The study was based on claims data from PCS Health Systems, a major PBM, and was published in 1988. The results indicated that mail service pharmacies provided lower unit costs on drugs than did community pharmacies but that unit cost savings were more than offset by higher usage among patients using mail service pharmacies. 10 PCS also retained the Gallup Organization to conduct a consumer survey of both mail service and community pharmacy patrons to determine why utilization was greater among mail service users. The results suggested that the difference was due to waste. About 15% of users of mail service pharmacies reported waste as compared with 24% of community pharmacy patrons. However, users of mail service pharmacies reported an average wastage of 129, compared with 36 for community pharmacy patrons. 10 Given the different outcome measures used, it is difficult to compare the results of the PCS study with our study. Also, the PCS study was conducted when third-party payment for prescriptions was much less prevalent and when copayments were much lower. If our results are typical of mail service pharmacy plans, then the current situation presents a substantial dilemma to health plans and mail service pharmacies. Health plans have turned to mail service pharmacy to provide lower drug costs. To encourage the use of mail service pharmacies, most health plans have provided patients with an economic incentive. Commonly this incentive is equal to one monthly copayment for every 3-months supply of prescription medicine. However, copayments have now become so substantial that when health plans offer this incentive to patients, they may realize higher costs than they would have through community pharmacies. So, this suggests that health plans may only realize savings through mail service pharmacies by eliminating or substantially reducing the economic incentive. This leads to the question of whether patients would use mail service pharmacies if they were not given economic incentives to do so. The available evidence indicates that they would not. Studies of patients reasons for using mail service pharmacy have found that the most important reason, by a wide margin over all other reasons, is that they pay less when using mail service. 6,12 Further, two reports have indicated that when offered mail service and community pharmacy services at the same price, patients overwhelmingly choose community pharmacies. Eagle Managed Care, a PBM owned by Rite-Aid, took over a Department of Defense health plan that had required the use of mail service pharmacy. Eagle Managed Care offered patients the choice of either mail service or community pharmacy at the same copay. A total of 90% of patients switched to community pharmacies when offered this choice. 13 Strategic Health Alliance, a PPO in Columbus, Ohio, reported that only 2% of maintenance prescriptions in its prescription benefits plan were dispensed through mail service pharmacy. Patients in this plan paid the same copayment for mail or community pharmacy service. 14 Health plans might also reduce, but not eliminate, the incentive. For example, in the plan we examined, mean patient copayment savings from using the mail service pharmacy were about $18 per prescription. A reduction of copayment savings to $9 per prescription would result in the health plan saving an average of $2 per prescription on mail service prescriptions (rather than losing an average of $7 a prescription as it does with the current copayment structure). But lower copayment savings would likely result in fewer patients using mail service pharmacy. This is problematic because the discounts that mail service pharmacies give to health plans on ingredient costs and dispensing fees are based on the prescription volume from the health plan. Thus, lowering patient copayment savings could reduce health plans savings on ingredient cost and dispensing fees. Limitations A major limitation of this study is that its results are based on one mail service pharmacy and one health plan. To determine the extent to which the plan we studied was representative, we compared it with the typical or average employer-sponsored health plan described in the Takeda Prescription Drug Benefit Cost and Plan Design Survey Report. 11 Average copayments in three-tier designs in employer-sponsored health plans were as follows. For a 3-month supply at community pharmacies, the copayments by tier were $23.04, $48.18, and $91.53. The comparable mail service copayments were $12.60, $26.01, $55.23. The mean copayment for mail service prescriptions in our sample was $37.35. For a comparable 90-day supply of medication dispensed at community pharmacies the mean copayment was $55.40. The mean differences between mail and community pharmacy copayments by tier for employer-sponsored health plans were about $10, $22, and $35. The $18 copayment difference in our plan was well within this range. Community pharmacy copayments varied in our sample from a low of $5/$10/$25 to a high of $12/$30/$50. The average community pharmacy copayment structure for employer- 342 Journal of the American Pharmacists Association www.japha.org May/June 2005 Vol. 45, No. 3

Community and Mail Service Pharmacy Costs RESEARCH based plans was $7.68/$16.06/$30.51 in 2001 and had shown substantial annual increases over the past several years. 11 The health plan we examined charged a mail copayment for a 90-day supply that was twice the community pharmacy copayment for a 30-day supply. This relationship of mail to community pharmacy copayments is common across the industry. These comparisons indicate that the copayment structure in the health plan we studied is comparable with the typical copayment structure for employersponsored health plans. The Report indicates that the average employer-sponsored health plan receives an AWP discount through mail service that was a little less than 5% greater than the discount on community pharmacy prescriptions. 11 The difference in the health plan we examined was 2%. We performed a breakeven analysis to take account of this difference. The results of this analysis indicated that the AWP discount difference would have to be greater than 6.6% for the cost of mail service prescriptions to be less than the cost of community pharmacy prescriptions for plan sponsors. We also compared dispensing fees and administrative fees in the plan we studied with those of employer-sponsored health plans. These plans paid average dispensing fees for 90-day supplies of $1.09 for mail service and $6.63 for community pharmacy. 11 The comparable fees for the health plan we studied were no dispensing fee for mail service and $6.00 for community pharmacies. The average administrative fees paid by employer-sponsored health plans were $0.38 for community pharmacy prescriptions and $0.23 for mail service prescriptions. The plan we studied charged the same administrative fee for mail and community pharmacy prescriptions. The effect of differences in dispensing and administrative fees on our results would be to minimally increase the cost of mail service prescriptions. Our results are also limited in that they apply only to mail service pharmacies that are not owned by PBMs. The trade press has recently reported that mail service pharmacies owned by PBMs, as compared with those not owned by PBMs, were substantially less likely to dispense generic products 15 and more likely to repackage and inflate the prices of brandname products. 15 The mail service pharmacy we studied was not owned by a PBM and we found no evidence that it engaged in either of these practices. Conclusion The results of our study indicated that use of a mail service pharmacy resulted in higher costs for the health plan we examined. This occurred because the copayments the plan lost through incentivizing patients to use mail service pharmacies were greater than the savings realized on ingredient costs and dispensing fees. These results highlight the importance for health plans of analyzing their claims data to determine the actual savings (or losses) that they realize through use of mail service pharmacies. References 1. National health expenditures aggregate amounts and average annual percent change, by type of expenditure: selected calendar years 1980 2001. Accessed at cms.hhs.gov/statistics/nhe/historical/t2.asp, October 23, 2003. 2. Levit KR, Smith C, Cowan C, et al. Trends in U.S. health care spending, 2001. Health Aff. 2003;22:154 64. 3. IMS Health Inc. DDD annual class-of-trade analysis 2002. Plymouth Meeting, Pa.: IMS Health Inc.; 2003. 4. IMS America. Class of trade analysis: a pharmaceutical market overview using IMS/DDD data. Totowa, N.J.: IMS America; 1992. 5. Horgan C, Goody B, Knapp D, et al. The role of mail service pharmacies. Health Aff. 1990;9:66 74. 6. Pharmaceutical Care Management Association Web site. The value of mail service. Accessed at www.pcmanet.org, April 17, 2002. 7. Codling M. Rapid growth for mail-order drug dispensing. Cambridge, Mass.: Arthur D. Little Decision Resources; 1987. 8. Konnor DD. The mail service pharmacy industry: growing by meeting the needs of managed health care. J Res Pharm Econ. 1990;2:3 14. 9. Enright SM. Mail-order pharmaceuticals. Am J Hosp Pharm. 1987;44:1870 3. 10. Sieben and Associates. Actuarial studies and reviews of mail order drug option experience. Scottsdale, Az.: PCS, Inc.; 1987. 11. Pharmacy Benefit Management Institute. The Takeda prescription drug benefit cost and plan design survey report: 2002 edition. Scottsdale, Ariz.: Pharmacy Benefit Management Institute, Inc.; 2002. 12. Birtcher KK, Shepherd MD. Users perceptions of mail-service pharmacy. Am Pharm. 1992;NS32:967 72. 13. Fried, L. Rite Aid levels paying field between drug-store and mail-order prescriptions. Drug Store News. 1998;20(8):CP4. 14. Patients prefer independents to mail order. Drug Topics. 1999;143(5):10. 15. PBM self-dealing could cost Medicare $30 billion. America s Pharm. 2003;125(10):12. CALL FOR SUBMISSIONS JAPhA Seeks Publishable Landscape or Nature Photographs Do you have one or more photographs from a recent trip that you would like to share with your colleagues in pharmacy? The JAPhA editors are seeking color photographs for the front cover of the journal and black-and-white photographs for inclusion at the ends of Science & Practice articles. The photographs should be artistic, high-quality shots of nature, landscape, or city scenes containing no identifiable people. If you have photographs that might fit this bill, e-mail the images to the Editor at mposey@aphanet.org. If your submission is chosen for inclusion, the editors will contact you for a high-quality image and the necessary permissions to use your work in the journal. Vol. 45, No. 3 May/June 2005 www.japha.org Journal of the American Pharmacists Association 343