Maryland Department of Health and Mental Hygiene 201 W. Preston Street Baltimore, Maryland 21201

Transcription

1 STATE OF MARYLAND DHMH Maryland Department of Health and Mental Hygiene 201 W. Preston Street Baltimore, Maryland Martin O Malley, Governor Anthony G. Brown, Lt. Governor John M. Colmers, Secretary January 17, 2007 The Honorable Ulysses Currie The Honorable Norman H. Conway Chairman Chairman Senate Budget and Taxation Committee House Appropriations Committee 3 West Miller Senate Office Bldg. 131 Lowe House Office Bldg. Annapolis, MD Annapolis, MD The Honorable Thomas M. Middleton The Honorable Peter A. Hammen Chairman Chairman Senate Finance Committee House Health and Government Operations 3 East Miller Senate Office Bldg. Committee Annapolis, MD House Office Bldg. Annapolis, MD RE: 2006 Joint Chairmen s Report (P. 114) Study of Pharmacy Dispensing Fees Dear Chairmen Currie, Middleton, Conway and Hammen: In keeping with the requirements of the 2006 Joint Chairmen s Report, the Department is submitting the enclosed report on the cost to pharmacies of dispensing prescription drugs to Medicaid patients. The report examines current research on Medicaid prescription dispensing fees (including dispensing fees in other states) and compares actual prescription drug acquisition costs to the reimbursement rates that will be in effect on January 1, 2007 as a result of the federal Deficit Reduction Act of If further information is required, please contact Jeff Gruel, Director of Maryland Pharmacy Programs, at (410) Enclosures Sincerely, Signature on file Susan Steinberg Chief Operating Officer cc: Anne Hubbard Tricia Roddy Jeff Gruel Toll Free MD-DHMH TTY for Disabled - Maryland Relay Service Web Site:

2 Medicaid Dispensing Fee Survey and Analysis The 2006 Joint Chairmen s Report required the Department of Health and Mental Hygiene to study and report on the cost to pharmacies of dispensing prescription drugs to Medicaid patients. This report examines current research on Medicaid prescription dispensing fees (including dispensing fees in other states) and compares actual prescription drug acquisition costs to the reimbursement rates that will be in effect on January 1, 2007 as a result of the federal Deficit Reduction Act of Based on the results of this study, the Department does not endorse a revision of the dispensing fee at this time. Instead, the Department recommends waiting until federal regulations are published in 2007 and the final federal upper limits have been released and implemented to gauge the impact of these measures on Medicaid prescription drug costs. Current Reimbursement Policy Nationally-published drug prices average wholesale price and wholesale acquisition cost tend to be higher than the actual drug acquisition costs of pharmacies, in part due to discounts and other incentives provided by drug manufacturers. Pharmacies receive reimbursement for Medicaid prescriptions based upon a dispensing fee plus an amount to cover the cost of the ingredient or product dispensed. Maryland s formula for determining the price to be paid for reimbursement of ingredients is based upon the lesser of four pricing formulae: Estimated Acquisition Cost (EAC) which is the lowest of: Wholesale Acquisition Cost (WAC) plus 8 percent, Direct Price plus 8 percent or Average Wholesale Price (AWP) minus 12 percent; or Federal Upper Limit (FUL) for multiple-source drugs; or Maryland State Maximum Allowable Cost (MAC) for multiple-source drugs; or The pharmacy s usual and customary charges. Since the Department receives rebates from drug manufacturers under the Medicaid Rebate Program, the Department realizes a net cost for drugs at a level representing their best price while reimbursing pharmacies for brand-name drugs at a rate above their acquisition cost (wholesale acquisition cost plus 8 percent). The dispensing fee that is assessed by the pharmacy itself is a fixed amount. Both components estimated acquisition costs and actual dispensing fees must be considered when determining whether pharmacists are paid appropriately. The second reimbursement formula, federal upper limits, applies only to multiple-source or generic drugs and is being revised in accordance with the Deficit Reduction Act of 2005 (DRA). Effective January 1, 2007, the new federal upper limits will consist of 250 percent of the average manufacturer price, replacing 150 percent of the lowest published price. Preliminary new FUL prices appear lower than the existing FUL prices. DHMH completed a comparison of these prices using the preliminary FUL figures issued by the federal government. The comparison revealed that in the aggregate, the Department may realize $2 million in annual savings. This reduction in the drug acquisition cost is much smaller than many have anticipated. One explanation is that the AMP price list does not take into consideration the Maryland Medicaid

3 MAC figures, which have historically been lower than most FUL prices. These savings cannot be assured, since federal regulations finalizing the methodology for calculation of the AMP will not be issued until mid-2007 (the Bush Administration recently announced a proposed rule on definition of AMP in accordance with the DRA; after a public comment period, a final regulation must be promulgated by July 1, 2007). Until then, the methodology of calculating the AMP is not final and therefore, AMP prices may fluctuate from currently issued prices. Dispensing Fees For retail pharmacies Maryland s dispensing fee is currently $3.69 for generic drugs and preferred brand name drugs; the dispensing fee for other brand name drugs is $2.69. For longterm care pharmacies, the dispensing fees are $4.69 and $3.69 respectively. A review of dispensing fees from other states indicates that Maryland s dispensing fee appears to be consistent with other states (see Attachment 1). Other state Medicaid retail pharmacy fees vary from a low of $1.75 in New Hampshire to a high of $7.25 in California. Cost of Dispensing Survey The Medicaid Pharmacy Program, through competitive bidding, has contracted with the University of Maryland School of Pharmacy s Pharmaceutical Health Services Research Department to analyze data collected from a recent survey regarding the cost to pharmacies of dispensing prescription drugs to Medicaid enrollees. The survey was posted on the Maryland Pharmacy website from October 6-23, and Maryland pharmacies were encouraged to complete the survey on-line. Out of approximately 1,100 pharmacies in the state, 387 submitted responses. Many of these were from chain pharmacies that submitted off-line composite entries for all their stores in Maryland. The survey instrument was based upon a survey used by the State of Maine and endorsed and provided to the Maryland Medicaid Pharmacy Program by the Maryland Association of Chain Drugstores. With the help of the School of Pharmacy, the survey was minimally expanded to meet the needs of the analysts. A copy of the survey analysis is attached (see Attachment 2) and it includes a copy of the actual survey. The analysis of the survey submissions reveals that the average cost of dispensing per prescription is $11.71 with a median cost of $ However, analysts of the data emphasize that due to limitations of the survey (i.e., data was self-reported), the findings should be interpreted with caution. In addition, no state or third party payer offers that amount as a dispensing fee. A dispensing fee increase approaching the median cost of $10.67 would cost the Maryland Medicaid program over $8 million (total funds) annually. Again, this does not mean that pharmacists are not receiving adequate payment. One needs to examine the profit levels that are obtained with the acquisition costs. Other Factors Complicating the pricing picture even more is the fact that several discount department stores have announced unusually low prescription prices for certain generic drugs. One chain discount department store offers prescriptions for over 300 different drugs at $4.00, including the price of ingredients and the dispensing fee. These new low prices are only a few pennies more than

4 Maryland Medicaid s current dispensing fee. This would appear to support the adequacy of the current Maryland Medicaid dispensing fee. Conclusion Based on preliminary analysis, it appears that the new federal upper limits will have a small reduction on Maryland Medicaid s reimbursement for generic drugs (approximately $2 million in annual savings). All components acquisition costs as well as the actual dispensing fee should be taken into consideration when determining whether pharmacists are adequately paid. Since Maryland s current dispensing fee appears to be consistent with other states fees, and in light of the latest trend by several chain discount department stores to offer generic prescriptions at very low prices, a revision of Medicaid s dispensing fee is not recommended at this time. Rather, it would be wise to wait until after federal regulations are published next July and the final federal upper limits have been released and implemented.

5 ATTACHMENT 1 STATE DISPENSING FEES 12/06/06 STATE REIMBURSEMENT FORMULA DISPENSING FEE ALABAMA Lowest of SMAC, FUL, WAC+9.2%, or AWP-10% $5.40 ALASKA AWP-5% $ $11.46 (long term ARIZONA (data pertains AWP-15% $2.00 to FFS population only; most AZ recip. in MCOs) ARKANSAS AWP-14% for brand; AWP-20% for generics $5.51; $2.00 for non MAC'd generics CALIFORNIA Brand or generic: AWP minus 17% $7.25-pharmacies; $8.00-SNF's; U&C COLORADO Brand: AWP-13.5%. Generic: AWP-35%, MAC/FUL, WAC+18% or U&C $4.00-retail; $ B pharmacies; $0.00-govt pharmacies CONNECTICUT AWP-14% $3.15 DISTRICT OF AWP-10% $4.50 COLUMBIA DELAWARE AWP-14% or 16% 3.65 [1 disp fee per pdt every 23 days unless drug is on exclude list (e.g., antibiotics)] FLORIDA Lower of: AWP-15.4% or WAC %, or state MAC or $4.23 federal MAC or the usual and customary GEORGIA Brand or generic: AWP-11% or MAC/FUL or most $5.13 for generics; $4.63 for brands favored price HAWAII AWP-10.5% $4.67-pharmacies; $0.50-physician dispensed IDAHO Lower of AWP-12%, SMAC, FUL or usual & customary $4.94; $5.54 for unit dose ILLINOIS Medicaid: AWP-12.% (brand) AWP-25% (generic) SeniorCare: AWP-14% (brand), AWP-25% (generic) INDIANA AWP-13.5% (brand) AWP-20% (generic) $4.90 IOWA AWP-12% $4.26 KANSAS AWP-13% (brand) AWP-27% (generic) $ Medicaid: $4.60 (generic) $3.40 (brand) SeniorCare: $2.25 (all) KENTUCKY AWP-15% for brand; AWP-14% for generics $4.50 for brand; $5.00 for generic LOUISIANA AWP-13.5% for independent pharmacies; AWP-15% for $5.77 chain pharmacies MAINE AWP - 15% $3.35-MaineRx; $2.35 MaineRx DEL (Drugs for the Elderly) MARYLAND WAC+ 8% or DP + 8% or AWP-12% For Community pharmacies: $3.69 for PDL & generics; $2.69 for brands MASSACHUSETTS Lower of MAC/FUL or WAC+5% (equates approx. to $3.00 AWP minus 16% for brands) MICHIGAN AWP-13.5% for independent pharmacies; AWP-15% for chain pharmacies (>5 stores) $3.77 MINNESOTA AWP-11.5%; Generics: AWP-12%+$3.65 or MAC/FUL + $ MISSISSIPPI Sole Source - Lower of U/C, AWP-12% or WAC + 9%; Generics - Lower of FUL, AWP-25% Effective : Sole Source - disp fee is $3.91; Generics - disp fee is $4.91; LTC - disp fee is $3.91 for all Sole Source and Generics

6 ATTACHMENT 1 STATE DISPENSING FEES 12/06/06 STATE REIMBURSEMENT FORMULA DISPENSING FEE MISSOURI AWP-10.43%; WAC+10%; MAC/FUL; UCR. Extensive MAC list. $ $3.95 enhanced fee subject to provider tax MONTANA AWP-15% Variable $2.00-$4.70 based on avg cost of filling a Rx as determined by the MT disp fee survey. $3.50 fee paid to out of state providers. NEBRASKA Brand or generic: AWP-11% or MAC/FUL + disp fee. $3.27-$5.00 NEVADA Brand or generic: AWP-15% or MAC/FUL $4.76 NEW HAMPSHIRE AWP-16% $1.75 NEW JERSEY AWP-12.5% $3.96 NEW MEXICO AWP-14% or MAC/FUL $3.65 NEW YORK For brands AWP-12.75%; for generics AWP-16.5% $3.50 Brand $4.50 Generic NORTH CAROLINA AWP-10%; extensive SMAC list $5.60 for generics & selected OTCs; $4.00 for brands; $0.00 fee for same-month refills. NORTH DAKOTA AWP-10% or WAC+12.5% $5.60 for generics; $4.60 for brands OHIO WAC+9%; Generics: pays for MAC'd generics at 65th $3.70 percentile of actual acquisition cost (averages AWP minus 60%) OKLAHOMA AWP-12% or MAC/FUL (State MACs are approx AWP- $ %) OREGON AWP-15%retail-AWP-11% Institutional $3.50 / $3.91 PENNSYLVANIA Lesser of FDB, MDX, Medispan pricing used; Brands = $4.00 Lesser of AWP-14%, WAC+7%; Generics = Lesser of AWP-25%, WAC+66%, FUL or State MAC RHODE ISLAND For either brand or generic: WAC+5% or FUL - $2.85 LTC; $3.40 Ambulatory whichever is lower SOUTH CAROLINA Lowest of AWP-10% or SMAC/FUL or U&C. State has $4.05 an extensive SMAC list. SOUTH DAKOTA AWP-10.5% $4.75 TENNESSEE AWP-13% or MAC/FUL $2.50 (TennCARE) TEXAS AWP-15% or WAC+12% $5.14 fee with variable add on UTAH AWP-15% 3.90 Urban; $4.40 rural VERMONT Lesser of 1) AWP % + dispensing; 2) HCFA FUL + $4.25 plus limited incentives dispensing; 3) VT (First Health) MAC (AB rated generics w/ at least 3 available, select products) + dispensing; 4) U&C including dispensing VIRGINIA AWP-10.25% $3.75 WASHINGTON AWP-14% (brand) AWP-50% (if > 5 mfg) $4.20-$5.20 WEST VIRGINIA AWP-12% $3.90 WISCONSIN AWP-13% $4.88--$40.11 WYOMING AWP-11% $5.00 2

7 ATTACHMENT 2 Analysis of Cost of Prescription Drug Dispensing in Maryland Report on the Cost of Prescription Dispensing (M00SO Pharmacy Dispensing Cost Analysis) C. Daniel Mullins, PhD, Principal Investigator Amy Davidoff, PhD, co-principal Investigator Francis B. Palumbo, PhD, Esq, co-investigator Françoise G. Pradel, PhD, co-investigator Julia Ju, PharmD, Lead Research Assistant The authors wish to acknowledge the technical assistance of Lisa Blatt, MA, for assistance in preparing this Report and Philip H. Cogan, RPh and Athos Alexandrou, MBA from Maryland Medicaid for their assistance in obtaining and interpreting survey responses, and Tammy Balzano of the DHMH Information Resources Administration for computer programming support. Produced for the State of Maryland Department of Health and Mental Hygiene By the University of Maryland School of Pharmacy December 7, 2006

8 TABLE OF CONTENTS CHAPTER 1 4 EXECUTIVE SUMMARY 4 CHAPTER 2 6 METHODS 6 Overview 6 Survey Development 6 Survey Data Collection 6 Computation of Cost of Dispensing (COD) 7 Table 1 Cost Allocation Methodology Used in Cost of Dispensing Analysis 8 CHAPTER 3 10 RESULTS 10 Characteristics of Responding Pharmacies 10 Table 2 Selected Characteristics of Responding Pharmacies 11 Table 3 Number and Characteristics of Prescriptions Filled 12 Estimates of Cost of Dispensing 12 Table 4 Cost of Dispensing and Prescription, Overall and by Pharmacy Characteristic 13 Pharmacy Characteristics Associated with COD Variations 14 Table 5 Estimated Associations Between Pharmacy Characteristics and Cost of Dispensing 15 CHAPTER 4 17 SUMMARY AND DISCUSSION 17 M00SO Pharmacy Dispensing Cost Analysis Page 2 of 35

9 APPENDIX A. SUPPLEMENTAL TABLES 18 Appendix Table 1. Additional Characteristics of Responding Pharmacies Relevant to Cost Computations 18 Appendix Table 2. Selected Statistics for Responding Pharmacies 19 Appendix Table 3. Reported Median and Mean Salaries for Pharmacists-in-Charge and Staff Pharmacists 20 Appendix Table 4. Reported Median and Mean Percent of Time Spent in Prescription Department for Pharmacists-in-Charge and Staff Pharmacists 21 Appendix Table 5. Reported Median and Mean Salaries for Pharmacy Intern, Pharmacy Clerk/Technician, Delivery and Janitorial Personnel 22 Appendix Table 6. Reported Median and Mean Percent of Time Spent in Prescription Department by Pharmacy Clerk/Technician, Delivery and Janitorial Personnel 23 Appendix Table 7. Reported Cost of Prescription Department Expenses 24 Appendix Table 8. Reported Total Store Expenses 25 Appendix Table 9. Regression Estimates of Association Between Average Cost of Dispensing and Pharmacy Characteristics 26 Appendix Table 10. Regression Estimates of Association Between Total Cost of Dispensing and Pharmacy Characteristics 27 APPENDIX B. SURVEY FORM INSTRUCTIONS 28 APPENDIX C. FEE SURVEY FORM 29 M00SO Pharmacy Dispensing Cost Analysis Page 3 of 35

10 Chapter 1 Executive Summary Most states pay a fixed dispensing fee for Medicaid prescriptions. There is considerable variance across states in the amount reimbursed with amounts ranging from $3 to more than $6 per prescription. A report by the United States General Accountability Office (GAO) 1 concluded that dispensing fee reimbursements in many states may be less than the actual dispensing costs to pharmacies. The Deficit Reduction Act of 2006 revises the Medicaid Program s reimbursement for generic drugs effective January 1, Under this Act the calculation of federal upper limits for reimbursement of generic drugs is being revised. Medicaid must use federal upper limits in its reimbursement methodology for generic drugs. Due to concerns that this revised federal limit will result in reimbursement for generic drugs below pharmacies costs, the Department has been instructed to conduct a study of the cost to pharmacies to dispense prescription drugs to Medicaid recipients. To determine the cost of dispensing a prescription in Maryland, the Department of Health and Mental Hygiene (DHMH) administered a survey to pharmacies in the State. Researchers at the University of Maryland School of Pharmacy used the self-reported data to calculate the cost of dispensing prescriptions and to examine the pharmacy characteristics that explain variation in costs. The survey contained questions on various components of costs related to prescription drug dispensing, including salaries for pharmacists and other personnel, rent, utilities, computers and software, and other costs of doing business. The survey also contained questions related to the number of prescriptions dispensed and the proportion of prescriptions dispensed that were reimbursed by Medicaid. The survey was posted on the DHMH website from October 6, 2006 through October 23, Pharmacy organizations and individual pharmacies were encouraged by DHMH to complete and submit this survey. Respondents were provided with instructions to assist in completing the survey. Several chain pharmacies in the state responded directly to DHMH on behalf of their individual units, using a consolidated spreadsheet instead of the web-based survey. A total of 387 valid responses were received, the vast majority of which (90%) came from Chain drug stores rather than Independent (9%) or Institutional pharmacies (1%). 1 U. S. General Accounting Office (GAO), Federal employees health benefits: effects of using pharmacy benefit managers on health plans, enrollees, and pharmacies. GAO January, M00SO Pharmacy Dispensing Cost Analysis Page 4 of 35

11 The average cost of dispensing per prescription was $11.71, with a median cost of $ The cost of dispensing per prescription was higher for pharmacies owned by corporations ($11.80) than for pharmacies for which the type of ownership was not a corporation ($7.34). Similarly, the average cost of dispensing per prescription was higher for Chains ($12.00) than for Independents ($9.02) or Institutions ($9.83). The cost of dispensing decreased as the volume of prescriptions filled increased and the cost of dispensing increases with the proportion of Medicaid business. Multivariable analysis confirmed the significance and direction of these results. Multivariate analyses were also performed to estimate the marginal cost of dispensing an additional prescription, as opposed to the average cost. The marginal cost is estimated to be $8.70. Due to the inability to adjust for underlying geographic differences in the cost of salaries and rent, these results must be interpreted with some caution. In summary, the cost of dispensing a prescription in Maryland varies based upon a number of factors and ranges from more than $7 to as much as $12, depending on ownership type, volume of prescriptions, and the percentage of prescriptions paid for by Medicaid. The reported costs of dispensing for the State of Maryland are higher than previously reported figures from the 1990s and are consistent with other figures that have been reported in the last few years for other states. M00SO Pharmacy Dispensing Cost Analysis Page 5 of 35

12 Chapter 2 Methods Overview Information on the characteristics of pharmacies in Maryland, and operating and overhead costs of the pharmacy within the context of the entire store or institution were collected by DHMH through an internet based survey. The data were stripped of identifiers and transferred to the Department of Pharmaceutical Health Services Research at the University of Maryland, School of Pharmacy (UMDSOP). Data were checked for internal consistency, categories were grouped, and overhead costs were allocated to the pharmacy in a manner consistent with recognized accounting standards. 2 The individual pharmacy cost of dispensing (COD) per prescription was computed. Then the mean and median of the pharmacy CODs were computed overall and by pharmacy characteristics. Multivariate regression was used to understand how the COD at the individual pharmacy level is affected by the pharmacy s characteristics. Multivariate regression was also used to compute the marginal, as opposed to the mean COD. Survey Development The survey was designed to capture information on the characteristics of pharmacies in the state and to gather information on their operating and overhead costs related to the cost of dispensing prescription medicine. DHMH drew on the experience of other state Medicaid programs to identify the relevant characteristics and costs, and structured them in a user-friendly format. The University of Maryland School of Pharmacy provided assistance to the DHMH in refining the final survey instrument. A copy of the final survey instrument is provided in Appendix C. Survey Data Collection Participation was open to all pharmacies (community and institutional) in the State of Maryland. The DHMH contacted potential participants via listserves, and a Newsletter to all pharmacies to inform them about the survey, and to encourage them to participate. The survey was posted on the DHMH website. Several chain pharmacies in the state responded directly to DHMH on behalf of their individual units. Data were submitted by these corporations on Excel spreadsheets as opposed to using the online submission process. These data were merged with the data submitted online prior to the analysis of the combined dataset. 2 Schafermeyer KW, Schondelmeyer SW, Thomas III J, Proctor KA. An analysis of the cost of dispensing third-party prescriptions in chain pharmacies. Journal of Research in Pharmaceutical Economics. 1992; 4(3): M00SO Pharmacy Dispensing Cost Analysis Page 6 of 35

13 Computation of Cost of Dispensing (COD) Pharmacies were asked to report their expenses from the 2005 tax year. Categories were provided for gross labor expense (salaries and benefits) for all pharmacists and pharmacy technicians, various categories of operating expenses for the prescription department, and various categories of other operating and overhead costs of the entire store that would be attributed to the prescription department. The survey also collected square footage and revenue statistics for the prescription department and the entire store, to be used to allocate other operating and overhead expenses to the prescription department. Labor expenses were adjusted based on the percent of time spent in the prescription department. Direct operating costs of the prescription department, collected from survey Table A, Section III, were included at 100% in the total COD. Other operating and overhead costs collected in survey Table B, Section III, were allocated to the prescription department based on either the area ratio or sales ratio. See Appendix C for a copy of the survey. Table 1 lists each category of cost and the basis for allocation. The area ratio was calculated by dividing total prescription department area (sum of square feet of prescription department, prescription register and patient waiting area, and stock room used for prescription drugs and containers) by the total store area (sum of square feet of total sales area and stock room for the entire store). The sales ratio was calculated by dividing the prescription department revenues by the total store revenues. Some pharmacies did not report the total store area or store revenues. When these respondents were contacted by DHMH, it was determined that these costs had already been allocated to the prescription department as part of each store s accounting process. The decision was made by DHMH to accept the reported costs as allocated. M00SO Pharmacy Dispensing Cost Analysis Page 7 of 35

14 Table 1 Cost Allocation Methodology Used in Cost of Dispensing Analysis Cost Category Allocation Basis Personnel Costs Adjusted by percent of time spent in prescription department, allocated at 100% Prescription Department Operating Costs 100% allocated to cost of dispensing Other Operating, Indirect or Overhead Costs Utilities Area ratio Depreciation Area ratio Property and real estate taxes Area ratio Building rent and equipment rent Area ratio Repairs Area ratio Operational and office supplies Area ratio Security Area ratio Insurance Sales ratio Interest Sales ratio Legal, accounting and professional fees Sales ratio Bad debts Sales ratio Credit card fees Sales ratio Advertising Sales ratio Central administration expenses Sales ratio Travel expenses Sales ratio Other taxes Sales ratio Other specified pharmacy expenses 100% allocated to cost of dispensing Some reported values appeared to be extreme or otherwise did not match the expected ranges, and steps were taken to adjust those values. For example, some pharmacies reported extremely large labor costs for individual pharmacists. Since respondents may have grouped together the labor costs for multiple individuals it would not have been appropriate to cap the cost as if they were reported for individual pharmacists. Instead, labor costs were summed for all pharmacists reported, and a labor cost per prescription filled was calculated. This value ranged from 0.70 to This statistic was capped at the 99 th percentile ($16.14 per prescription), and was used as the basis to cap total pharmacist labor cost included in the COD for each pharmacy. A similar process was used to cap extreme values for total pharmacy technician labor expense. Other examples where adjustments were made include: Exclusion of drug ingredient costs when reported by pharmacies. Several respondents reported allocation statistics such that the area or sales ratio exceeded a value of 1.0 (100%). In these situations, the median allocation statistic from reporting pharmacies was substituted. M00SO Pharmacy Dispensing Cost Analysis Page 8 of 35

15 Total costs were calculated as the sum of adjusted personnel costs, prescription department operating costs, and the other operating and overhead costs that were allocated to the prescription department. The unit of analysis in this study was the mean pharmacy COD per prescription, which was calculated by dividing the total prescription dispensing-related costs for each pharmacy by the total number of prescriptions dispensed by that pharmacy during the year. The mean and median pharmacy COD were calculated overall, and by characteristics including ownership, affiliation, physical setting, and size. Multivariate regression analyses were performed to examine the pharmacy characteristics and other factors that explain variation in pharmacy COD. Regression can be thought of as a statistical technique that allows the analyst to answer the question What happens to COD if we make a small change to a single characteristic of interest, but don t change any of the other characteristics? Regression analysis was also used to examine factors that affect the total pharmacy cost, and provides an estimate of the marginal cost of dispensing one additional prescription. The explanatory variables used in the models include percent of total prescriptions that are Medicaid, percent other third-party prescriptions, prescription volume (in quartile ranges or continuous), percent of total prescriptions that are new, as opposed to refills, independent versus chain pharmacy, whether emergency services are offered, pharmacy hours per week, whether there is an in-house charge system for prescription sales, and physical setting for the pharmacy (free standing, mass merchant, and other). We were not able to include information on type of pharmacy, or geographic locations, due to the large number of missing values. M00SO Pharmacy Dispensing Cost Analysis Page 9 of 35

16 Chapter 3 Results A total of 390 pharmacies responded to the survey, reflecting a sample of 31.5% out of 1240 pharmacies in Maryland. Three pharmacies were excluded from the analysis because they did not report the total number of prescriptions per year. Thus, the final sample size was 387. All results are based on the 387 valid responses, adjusted for outliers as described on page 8 of the Methods Section. Characteristics of Responding Pharmacies Table 2 provides characteristics of the responding pharmacies. Almost all (98%) pharmacies responding to the survey are organized as corporations, and 90% describe themselves as part of a larger chain. More than two out of five are located in shopping centers, with a similar number situated in free-standing sites. With respect to location, the largest number reported being in the outer suburbs, but information on this characteristic must be considered with caution due to the high item non-response. A high item non-response also was present for type of pharmacy, where all but one respondent reported being a retail provider, but responses were missing for almost half the observations. Relatively few pharmacies reported providing emergency services (4%), delivery services (8%) or in-house charge services (13%), and these characteristics were linked. For example, almost all pharmacies with emergency services were independent pharmacies, and none of the chain pharmacies provided such services. Among the 31 pharmacies providing delivery service, 30 of them were independent pharmacies and 1 of them was an institutional pharmacy. None of the chain pharmacies provided delivery services for prescription drugs. Among those pharmacies that allowed an in-house charge system for prescription sales, 25 of them were chain pharmacies, another 25 were independent pharmacies, and one was an institutional pharmacy. All pharmacies maintained an electronic patient profile system. Only 1% of pharmacies in our survey dispensed prescriptions to nursing home residents. Additional characteristics for reporting pharmacies are provided in Appendix A. Table 1. M00SO Pharmacy Dispensing Cost Analysis Page 10 of 35

17 Table 2 Selected Characteristics of Responding Pharmacies Characteristic Number of Pharmacies Percent of Non- Missing Responses Ownership Type Corporation % Non-Corporation 8 2% Ownership Affiliation Chain % Independent 35 9% Institutional 3 1% Physical Setting Medical Office Building 8 2% Shopping Center % Free Standing % Mass Merchant 30 8% Downtown 17 5% Other 3 1% Missing 16 Geographic Location Inner City 11 12% Inner Ring Suburbs 11 12% Outer Suburbs 49 53% Rural 21 23% Missing 295 Type of Pharmacy Retail % Long Term Care 1 1% Missing 191 Provides Emergency Services Yes 14 4% No % Provides Prescription Drug Delivery Yes 31 8% No % Provides In-House Store Charge System for Prescription Sales Yes 51 13% No % Maintains Electronic Patient Profile System Yes % No 0 0% Dispenses to Nursing Home Residents Yes 4 1% No % Table 3 provides information on the number and characteristics of prescriptions filled by the pharmacies. The mean total number of prescriptions was 45,662, but M00SO Pharmacy Dispensing Cost Analysis Page 11 of 35

18 the range was quite broad, with a minimum of less than 3,000 to over 220,000. Medicaid comprises a small proportion of prescriptions for the average pharmacy (6.8%), with other third party payers covering a much larger proportion on average (83.8%). Pharmacies reported that a relatively small proportion of prescriptions required prior authorization or some other contact with the Medicaid program (1%) or other third party payers (1.5%). Discussions with respondents suggest that many reported only the percent of filled prescriptions that required and received prior authorization; whereas other pharmacies estimated the proportion of prescriptions submitted that would be subject to prior authorization. Because of differences in reporting, we did not use this characteristic in the analysis. Additional characteristics for reporting pharmacies are provided in Appendix A. Tables 2-8. Table 3 Number and Characteristics of Prescriptions Filled Variables Number of Median Mean Pharmacies Total prescriptions ,545 45,662 New prescriptions , ,198 Refills ,666 19,268 Percent new prescriptions % 58% Percent Medicaid prescriptions % 6.8% Percent other third-party % 83.8% prescriptions Percent of Medicaid % 1.0% Prescriptions requiring prior authorization Percent of other third-party Prescriptions requiring prior authorization % 1.5% Estimates of Cost of Dispensing Table 4 provides estimates of the mean and median COD, overall and by pharmacy characteristic. The overall mean pharmacy COD is $11.71 with a range of $4 to $39 (data not shown). As expected, the median value is less ($10.67). We focus on median values because they are less affected by skewed distributions. The COD varies by pharmacy characteristics. The median COD for corporations ($10.74) is larger than for non-corporations ($7.54). Consistent with that trend is the finding that the median COD for chain stores ($11.01) is higher than for independents ($8.66). There were no clear patterns for physical setting or geographic location. Pharmacy size is associated with increases in the median COD. The COD for the lowest quartile in terms of pharmacy size was $12.30, compared to $9.69 for the highest quartile. The percent of prescriptions paid for by Medicaid was also associated with a trend towards increasing COD. The median cost for the lowest quartile is $9.05, and for the highest it is $ It should be noted that the large number of missing values for some categories, and the small number of pharmacies in many categories precluded statistical M00SO Pharmacy Dispensing Cost Analysis Page 12 of 35

19 comparison of the means for each category. The differences across groups should be interpreted as suggestive trends. Table 4 Cost of Dispensing and Prescription, Overall and by Pharmacy Characteristic Number of Pharmacies Median COD ($) Mean COD ($) All Pharmacies 387 $10.67 $11.71 Type of Ownership Corporation 379 $10.74 $11.80 Non-Corporation 8 $ 7.54 $ 7.34 Affiliation Chain 349 $11.01 $11.99 Independent 35 $ 8.66 $ 9.02 Institutional 3 $10.40 $ 9.83 Geographic Location Inner City 11 $ 8.43 $11.10 Inner Ring Suburbs 11 $13.45 $16.44 Outer Suburbs 49 $10.25 $12.64 Rural 21 $ 8.81 $10.11 Missing 295 $10.87 $11.51 Physical Setting Medical Office Building 8 $10.51 $12.25 Shopping Center 164 $10.93 $11.46 Separate Free Standing 149 $10.81 $11.89 Mass Merchant 30 $ 9.77 $12.34 Downtown 17 $11.30 $12.75 Other 19 $ 9.56 $10.21 Number of Prescriptions Quartile 1 (<30,000) 96 $12.30 $14.71 Quartile 2 (=> 41,545) 97 $10.00 $11.19 Quartile 3 (=> 56,619) 97 $11.09 $10.70 Quartile 4 (>56,619) 97 $ 9.69 $10.26 % Rx Covered by Medicaid Less than 2% 98 $ 9.05 $ % 105 $10.25 $ % 93 $12.48 $12.75 More than 9% 91 $12.26 $12.00 * Note that categories for Total and Medicaid prescriptions were based on quartiles of the distributions M00SO Pharmacy Dispensing Cost Analysis Page 13 of 35

20 Pharmacy Characteristics Associated with COD Variations Table 5 reports the results when multivariate linear regression is used to identify the association between each pharmacy characteristic and COD, while controlling for other characteristics. The first column in the table identifies the characteristic of the pharmacy being examined. For continuous variables the second column identifies the unit of measurement and the magnitude of the change that is being examined. For categorical variables, such as the size group, the second column indicates what comparison is being made in the analysis. The third column indicates the dollar impact on COD associated with the change in the characteristic. The symbols adjacent to the numbers in the third column indicate whether this result is deemed to be statistically significant, or if it is more likely just a chance finding. The results of the regression indicate that as the size of the pharmacy increases, the COD decreases. Relative to pharmacies in the lowest quartile of prescription volume, the COD for pharmacies in the second quartile are $3.12 less. The pharmacies with the highest prescription volume have even lower COD. Relative to pharmacies in the lowest quartile ranked by prescription volume, the COD for the highest quartile is $4.56 lower. The percent of prescriptions that are covered by Medicaid is associated with a higher COD. A 1 percentage point increase in the percent covered by Medicaid is associated with a $0.16 increase in the COD. There is not a significant effect associated with the percentage of prescriptions covered by other third party payers. New prescriptions are associated with a small increase in COD. For every 1 percentage point increase in the percent of prescriptions that are new, the COD increases by $.20. M00SO Pharmacy Dispensing Cost Analysis Page 14 of 35

21 Table 5 Estimated Associations Between Pharmacy Characteristics and Cost of Dispensing Estimated effect on average COD Independent Variables Unit of Measurement, in $, associated Unit of Change with 1 unit change in characteristic Size of pharmacy Ref group: 1 st quartile Rx volume 2 nd quartile Rx volume In 2 nd quartile compared to 1st -3.12* 3 rd quartile Rx volume 3 rd quartile compared to 1st -4.19* 4 th quartile Rx volume 4 th quartile compared to 1st -4.56* Percent Medicaid prescriptions 1 percentage point 0.16* Percent other third party prescriptions 1 percentage point 0.02 Percent new prescriptions 1 percentage point 0.20* Pharmacy affiliation, & emergency service provision Reference group: Chain pharmacy Independent pharmacy without emergency services Independent without emergency services, compared to chain -5.42* Independent pharmacy with Independent with emergency -5.74* emergency services services, compared to chain In-house charge system for Has in-house store charge 4.30* prescription sales compared to not having Pharmacy hours open per week 1 hour 0.01 Physical Setting Reference group: shopping center Free standing physical setting Free standing compared to shopping center Mass Merchant physical setting Mass merchant compared to Other physical setting R-Square 0.38 * Significant at p<0.05 shopping center Other setting compared to shopping center The combination of pharmacy affiliation and offering emergency services has a significant effect on COD. Recall that no chain pharmacies reported emergency services. Relative to chain stores, the COD for independent pharmacies without emergency services was $5.42 less. The COD of independent pharmacies with emergency services was $5.74 less than for chain pharmacies. Relative to pharmacies without in-house store charge systems, pharmacies with in-house store charge systems had CODs that were $4.30 more. Neither the number of hours the pharmacy was opened, nor the physical setting was associated with variations in COD. M00SO Pharmacy Dispensing Cost Analysis Page 15 of 35

22 The multivariate regression designed to explain variation in total pharmacy COD indicates that, given the current volume of prescriptions, an increase of one prescription increases the total COD by $8.70. This estimate is known as the marginal COD. The full set of estimates from the two regressions is provided on Appendix A. Tables 9 and 10. M00SO Pharmacy Dispensing Cost Analysis Page 16 of 35

23 Chapter 4 Summary and Discussion A total of 390 pharmacies responded to the survey. Of the 387 valid responses, 90% came from Chain drug stores, 9% from Independent pharmacies, and 1% from Institutional pharmacies. On average, pharmacies dispensed 45,662 prescriptions per year (median: 41,545). The average percentage of total prescriptions covered by Medicaid was 3.8%, with a median percentage of 6.8%. The average cost of dispensing per prescription was $11.71, the median cost of dispensing per prescription was $10.67, and the marginal cost of dispensing per prescription was $ The average cost of dispensing per prescription was higher for Chain drugstores ($12.00) than for Independent pharmacies ($9.02). The average cost of dispensing per prescription was lower for pharmacies reporting a low Medicaid prescription volume ($10.63 for a Medicaid volume less than 2% of total prescriptions filled) compared to those reporting a high Medicaid prescription volume ($12.00 for a Medicaid volume above 9%). Recent changes in prescription benefits for dual Medicare-Medicaid eligibles are likely to affect the volume of prescriptions paid for by Medicaid, and may affect the average complexity and cost. These patients are likely to use a complex array of prescription drugs, and pharmacists are likely to spend more time filling prescriptions and counseling patients than they would for less complicated Medicaid patients. Beginning in 2006, prescription drug coverage for dual eligibles shifted from Medicaid to other third party coverage. With this change, the association between Medicaid volume and cost of dispensing might be reduced. Limitations The findings in this report are based on an analysis of self-reported data. According to 2005 statistics from the National Pharmaceutical Council, 67% of pharmacies in Maryland were Chain pharmacies, 27% Independent pharmacies, and 6% Hospital/Institutional pharmacies. Compared to this distribution, Chain pharmacies were overrepresented in the DHMH study, while Institutional and Independent pharmacies were underrepresented. Therefore, the study sample may not be representative and findings should be interpreted with caution. M00SO Pharmacy Dispensing Cost Analysis Page 17 of 35

24 Appendix A. Supplemental Tables Appendix Table 1. Additional Characteristics of Responding Pharmacies Relevant to Cost Computations Characteristic Number of Pharmacies Percent of Non-Missing Responses Facility Ownership Pharmacy Owns 5 1% Rents from Related Party 39 10% Rents from Unrelated Party % Basis of Rent Payments Square Foot 15 8% % of Sales 25 14% Combination of Above Two % Fixed Amount 13 7% Dispenses to Nursing Home Residents Yes 4 1% No % Dispenses Unit Dose to Nursing Home Yes 2 1% No % Missing 209 Type of Unit Dosing Modified Unit Dose (bingo cards/blister packs) 1 50% Unit Dose 1 50% M00SO Pharmacy Dispensing Cost Analysis Page 18 of 35

25 Appendix Table 2. Selected Statistics for Responding Pharmacies Variables Number of Median Mean Pharmacies New prescriptions ,628 26,198 Refills ,666 19,268 Total prescriptions ,545 45,662 Annual total store 356 $3,866,155 $4,353,672 revenues Annual prescription 387 $2,687,699 $3,020,882 department revenues Percent Medicaid % 7% prescriptions Percent other thirdparty % 84% prescriptions Percent of Medicaid % 1.02% prescriptions requiring a prior authorization Percent of other thirdparty % 1.50% prescriptions requiring a prior authorization Percent of Medicaid 30 25% 32% prescriptions home delivered Pharmacy hours open per week Square feet of total 355 8,356 7,778 sales area Square feet of prescription department Square feet of prescription register and patient waiting area Square feet of stock room for entire store Square feet of stock room for prescription drugs and containers Area ratio * Sales ratio ** M00SO Pharmacy Dispensing Cost Analysis Page 19 of 35

26 Appendix Table 3. Reported Median and Mean Salaries for Pharmacists-in- Charge and Staff Pharmacists Number of Pharmacies Median Salary ($) Mean Salary ($) Pharmacist-incharge 61 $100,000 $115,434 Pharmacist A 382 $189,332 $184,395 Pharmacist B 46 $35,787 $47,067 Pharmacist C 28 $11,664 $38,070 Pharmacist D 21 $11,022 $33,828 Pharmacist E 18 $15,469 $37,978 Pharmacist F 13 $11,000 $23,186 Pharmacist G 11 $9,578 $19,180 Pharmacist H 8 $4,923 $18,195 Pharmacist I 7 $20,288 $17,712 Pharmacist J 6 $9,032 $17,248 Pharmacist K 5 $8,514 $11,747 Pharmacist L 2 $41,237 $41,237 Pharmacist M 2 $50,958 $50,958 Pharmacist N 2 $5,081 $5,081 Pharmacist O 1 $6,160 $6,160 M00SO Pharmacy Dispensing Cost Analysis Page 20 of 35

27 Appendix Table 4. Reported Median and Mean Percent of Time Spent in Prescription Department for Pharmacists-in-Charge and Staff Pharmacists Number of Pharmacies Median Time % Mean Time % Pharmacist-incharge % 95% Pharmacist A % 98% Pharmacist B 48 80% 88% Pharmacist C 39 80% 85% Pharmacist D 34 80% 85% Pharmacist E 32 80% 84% Pharmacist F 29 80% 83% Pharmacist G 28 80% 82% Pharmacist H 26 80% 81% Pharmacist I 26 80% 81% Pharmacist J 26 80% 80% Pharmacist K 25 80% 80% Pharmacist L 25 80% 80% Pharmacist M 25 80% 80% Pharmacist N 25 80% 80% Pharmacist O 25 80% 80% M00SO Pharmacy Dispensing Cost Analysis Page 21 of 35

28 Appendix Table 5. Reported Median and Mean Salaries for Pharmacy Intern, Pharmacy Clerk/Technician, Delivery and Janitorial Personnel Number of Pharmacies Median Salary ($) Mean Salary ($) Intern 98 $832 $5,741 Clerk A 224 $36,257 $40,434 Clerk B 52 $11,423 $16,024 Clerk C 48 $11,453 $14,654 Clerk D 38 $11,729 $18,744 Clerk E 32 $9,903 $14,152 Clerk F 27 $7,491 $11,863 Clerk G 23 $10,580 $16,554 Clerk H 17 $5,900 $13,637 Clerk I 14 $24,970 $22,872 Clerk J 15 $15,555 $14,691 Clerk K 13 $12,000 $14,066 Clerk L 11 $14,488 $14,746 Clerk M 9 $7,539 $12,609 Clerk N 8 $3,949 $9,203 Clerk O 7 $22,000 $20,262 Clerk P 6 $11,600 $12,046 Clerk Q 2 $13,160 $13,160 Clerk R 2 $16,836 $16,836 Delivery personnel Janitorial personnel $3, $645 M00SO Pharmacy Dispensing Cost Analysis Page 22 of 35

29 Appendix Table 6. Reported Median and Mean Percent of Time Spent in Prescription Department by Pharmacy Clerk/Technician, Delivery and Janitorial Personnel Number of Median Time Mean Time Pharmacies Clerk A % 97% Clerk B % 89% Clerk C % 88% Clerk D % 91% Clerk E % 90% Clerk F % 94% Clerk G % 89% Clerk H % 95% Clerk I % 92% Clerk J % 92% Clerk K % 91% Clerk L % 90% Clerk M % 92% Clerk N % 93% Clerk O % 95% Clerk P % 99% Clerk Q % 100% Clerk R % 100% Delivery 36 50% 53% personnel Janitorial personnel % M00SO Pharmacy Dispensing Cost Analysis Page 23 of 35

30 Appendix Table 7. Reported Cost of Prescription Department Expenses Number of Pharmacies Median ($) Mean ($) Prescription Computer 384 $561 $2,524 Expenses Prescription claim transmission 219 $3,458 $4,314 charges and switching fees Prescription supplies 384 $5,196 $6,968 Delivery expenses 30 $2,714 $11,819 Dues and publications 57 $445 $1,128 Pharmacist continuing 48 $505 $690 education costs Shrinkage 205 $5,758 $10,570 TP adjustments 366 $3,458 $6,821 Benefits 373 $45,000 $50,905 Bonuses 21 $2,679 $23,728 Liability insurance 53 $4,167 $5,467 Inventory service 210 $2,524 $2,015 Warehouse expense 37 $45,030 $39,831 Other prescription department 168 $350 $1,449 expenses Total prescription department expenses 387 $70,979 $81,860 M00SO Pharmacy Dispensing Cost Analysis Page 24 of 35

31 Appendix Table 8. Reported Total Store Expenses Number of Pharmacies Median ($) Mean ($) Electricity 185 $25,185 $24,731 Natural gas 136 $1,512 $1,894 Water 140 $505 $914 Telephone 381 $4,161 $4,381 Other utilities 208 $14,859 $14,242 Trash collection 199 $1,808 $2,134 Sewer $161 Depreciation 372 $30,651 $42,028 Taxes 189 $3,488 $9,041 Property taxes 190 $3,048 $4,236 Real estate taxes 137 $11858 $12,614 Other taxes 143 $81 $1,732 Rent 355 $120,698 $135,771 Building rent 183 $128,257 $141,480 Equipment and other rent 171 $224 $1,922 Repairs 350 $6,914 $13,057 Insurance 200 $1,162 $4,831 Workers comp and employee 186 $6,302 $7,373 Medical insurance 179 $22,405 $26,187 Any other insurance $3,438 Interest expense 19 $13,940 $12,764 Legal, accounting and other 189 $223 $1,800 professional fees Bad debts 200 $571 $2,154 Credit card fees 378 $14,515 $18,019 Operating and office supplies 378 $5,199 $5,832 Advertising 381 $45,016 $49,208 Central administration 361 $31,631 $123,550 expenses Travel expenses 129 $571 $1,778 Security 201 $25 $3,148 Monitoring system 154 $739 $1,141 Security personnel 51 $15,340 $19,296 Other pharmacy expenses not 150 $26,266 $26,907 included elsewhere Total other pharmacy related expenses 81 $128,497 $199,080 M00SO Pharmacy Dispensing Cost Analysis Page 25 of 35

32 Appendix Table 9. Regression Estimates of Association Between Average Cost of Dispensing and Pharmacy Characteristics Independent Variables β Coefficient Standard P value Error Size of pharmacy Ref group: 1 st quartile Rx volume 2 nd quartile Rx volume < rd quartile Rx volume < th quartile Rx volume < Percent Medicaid prescriptions < Percent other third party prescriptions Percent new prescriptions < Pharmacy affiliation, & emergency service provision Reference group: Chain pharmacy Independent pharmacy without emergency services < Independent pharmacy with < emergency services In-house charge system for < prescription sales Pharmacy open hour per week Physical Setting Reference group: shopping center Free standing physical setting Mass Merchant physical setting Other physical setting R-Square = 0.35 Dependent variable = average cost of dispensing per prescription M00SO Pharmacy Dispensing Cost Analysis Page 26 of 35