Medicaid Pharmacy Savings Opportunities:
National and State-Specific Estimates
This study explores how more efficient pharmacy benefits management—apart from drug manufacturer rebates—could save Medicaid an additional $74.4 billion over the next decade.These savings opportunities are compelling as the federal government and states strive to protect benefits and prepare for the Medicaid expansion authorized by the Affordable Care Act starting in 2014.
Today, Medicaid remains an outlieramong the nation’s large pharmacy benefits programs. In most states, Medicaid pharmacy payments are set by statute, not negotiated with pharmacies.
This report estimates the savings available if each state Medicaid program —both fee for service and managed care—applied pharmacy best practices. Some states may contract with a pharmacy benefit manager (PBM) or Medicaid Managed Care Organization (MCO), but impose restrictions on how they design formularies or negotiate with pharmacies. Likewise, some states may use a fee-for-service (FFS) approach that is more efficient than others.
Regardless, states tend to save more when pharmacy reimbursements and dispensing fees are negotiated by PBMs or MCOs rather than set by state governments. States also save more when they resist a “one-size-fits-all” approach to formularies and instead allow them to be flexible and actively managed.
Experience in manystates indicates that Medicaid pharmacy benefits can be more actively managed without compromising quality or access to medications for the unique and vulnerable populations that Medicaid serves. Likewise, widely varying payment levels—and per member per month (PMPM) costs—among state Medicaid FFS programssuggest that there is substantial room for improvement.
Key Savings Opportunities
Optimizing pharmacy benefit management tools and strategies in state Medicaid programs nationwide could save a total of $74.4 billion across the ten-year period 2014–2023, including$43 billion in federal savingsand $31.4billion state savings. Components of these potential savings include:
- $23.5 billion saved by increasing the use of generic drugs: State-administered Medicaid programs that do not already manage formularies are less effective at encouraging the use of generic medications. The average generic dispensing rate in the Medicaid FFS setting is 73%, compared to an average generic dispensing rate exceeding 80% in the Medicaid MCO setting.
- $12.5 billion saved by negotiating market-based pharmacy dispensing fees:Medicare Part D and commercial sector plans negotiate market-based pharmacy dispensing fees that average less than half those set by officials in most state Medicaid programs.
- $33.4 billion saved by using limited pharmacy networks: In most state Medicaid programs, every drugstore in the state is entitled to participate. State Medicaid programs could achieve greater savings by using a competitive process and negotiating better discounts from select drugstores that wish to participate in a limited pharmacy network.
- $2.7 billion saved by encouraging the use of more affordable, preferred brands: Unlike PBMs, state Medicaid FFS programs generally do not aggressively encourage the use of more affordable, preferred brands through active formulary management.
- $2.3 billion saved by reducing drug diversion, polypharmacy, fraud and waste: Medicaid plans that are more actively managed detect patterns of fraud through use of tools like step therapy, audits, and pharmacy lock-in programs to help detect and avoid inappropriate utilization.
Rebates from Brand DrugManufacturers are Unrelated to Medicaid Pharmacy Payments
The statutory and supplemental rebates paid to Medicaid by brand name manufacturers are determined separately from pharmacy dispensing fees and ingredient cost reimbursements. Increasing generic drug utilization will reduce drug manufacturer rebates, but still generate net savings of $23.5 billion.
Although an increase in the use of generics reduces the use of brand drugs and the related rebate income they generate for states, the net savings to Medicaid FFS programs is nonetheless enormous, as reflected in our savings estimate.
States Transitioning to Greater Pharmacy Benefits Management
Case Study: New York State
Many state Medicaid programs transitioning to greater pharmacy benefits management to lower costs without reducing the number of enrollees or the quality of pharmacy benefits.The initial results of New York transition to a managed pharmacy benefit in 2012 have been promising:
- New York Medicaid saved an estimated $425 million in 2012, four times more than the state originally estimated.
- PBM management techniques are expected to increase the generic drug dispensing rate (GDR) to more than 80% for the managed Medicaid population.
- Consistent with the commercial sector and Medicare Part D, pharmacy dispensing fees have been reduced from $3.50 to an average of approximately $1.75 under the PBM approach.
State-by-State Medicaid Savings
State-by-State Medicaid Savings from Optimal Use of PBM Tools and Strategies
Over the past several years, Medicaid fee-for-service (FFS) pharmacy programs have expanded their cost management of prescription drugs, but most still have not achieved the same level of pharmacy benefits management as found in either Medicare Part D, Medicaid MCOs, or the commercial sector.
State administered Medicaid FFS programs commonly reimburse pharmacies more for dispensing fees and ingredient costs than do MCOs and Part D plans. Additionally, Medicaid MCOs have demonstrated lower utilization and higher generic fill rates than in the Medicaid FFS setting. Medicaid FFS programs could achieve substantial savings if they were to move toward the reimbursement and utilization levels found in Medicaid and commercial MCOs and Medicare Part D plans. Improving management of the FFS pharmacy benefit would likely entail more competitive pricing policies as well as stronger management of the Medicaid FFS pharmacy benefit as typically occurs in the private sector. While there would be new costs associated with increasing management functions, the potential savings would more than offset these new administrative costs.
To estimate the potential impact of increased pharmacy benefit management in the Medicaid FFS setting, we modeled the impact of moving Medicaid FFS to levels typically found in Medicaid and commercial MCOs and Part D plans in four key areas: dispensing fees, generic dispensing, preferred brand dispensing, and reducing waste. Additionally, we calculated savings if both Medicaid FFS programs and MCOs used limited pharmacy networks consistent with access standards used in Medicare Part D.
We modeled these changes in a step-wise fashion so that the savings attributable to each step reflect the impact of changes made in prior steps. For example, the savings estimated for improving the generic mix reflect the decrease in dispensing fees and ingredient costs made in prior steps. Lastly, we subtracted the anticipated administrative costs of implementing more active benefit management activities.
Baseline 2012 Medicaid generic dispensing rates and pharmacy dispensing fees are presented in Exhibit 1. The Medicaid savings estimates from optimal management of the pharmacy benefit are shown in Exhibit 2 across the upcoming ten-year period. Projections are shown for each state and each component (e.g., competitive dispensing fees, optimal use of generics, etc.). Across the ten-year timeframe 2014–2023, potential nationwide Medicaid savings of $74.4 billion are projected if the pharmacy benefit in each state were optimally managed.
The degree to which various components will yield savings in each state varies considerably by component. For example, states with a relatively large dispensing fee have greater savings potential in that component. Nationwide, the largest component would be limited pharmacy networks, accounting for 45% of the total $74.4 billion savings, with the next-largest component being optimal use of generics (32% of national savings).
Exhibit 3 presents the federal and state government share of these overall savings within each state. The distribution between the federal and state share in any given state is determined by the state-specific Medicaid matching rate. Nationwide, 58% of the overall ten-year potential savings of $74.4 billion would accrue to the federal government.
Exhibit 1. Baseline Medicaid FFS Statistics and Potential Savings Opportunities
|State||Estimated Percentage of Medicaid Prescriptions in Fee-For-Service, 2014-2023||Generic Dispensing Rate in Fee-For- Service, 2011||Dispensing Fee inFee-For-Service,
|Alaska||100%||70.1%||$3.45 – $11.46||22.6%|
|District of Columbia||50%||70.0%||$4.50||22.4%|
|Idaho||100%||74.0%||$11.51 – $15.11||15.3%|
|Oregon||29%||77.0%||$9.68 – $14.01||14.7%|
Generic dispensing rates are from Drug Utilization Review 2011 Annual Report provided to CMS by each State. Website link is:
Medicaid dispensing fees also from CMS website:
Nationwide dispensing fee average excludes states that use a cost-plus fee structure (AL, ID, and OR).
Exhibit 2. Potential Medicaid Savings Distribution by PBM Tool/Strategy and by State, 2014–2023
|State||Competitive Dispensing Fee Savings||Optimal
|Reduced Polypharmacy, Fraud, Waste, and Abuse||Limited
|District of Columbia||12%||14%||2%||2%||69%||$334,274,878|
|Component’s Share ofTotal US Savings||16.8%||31.7%||3.6%||3.1%||44.8%||100.0%|
* State already optimizes pharmacy benefits management in this area.
**This state applies AAC benchmarks and pays higher dispensing fees. The cost impact of this approach is not yet clear.
Exhibit 3.Federal and State Savings from Optimal Use of PBM Tools and Strategies in Medicaid, 2014–2023
|State||Total Savings||Federal Savings||State Savings|
|District of Columbia||$334,274,878||$233,992,415||$100,282,463|
Savings estimates were derived in each state across the ten-year timeframe 2014–2023.The modeling effort assembled baseline Medicaid pharmacy usage costs for calendar year (CY) 2011, and then estimated how these figures would progress in each year through 2023 in the absence of strengthened benefits management. The Menges Groupthen applied an array of pharmacy cost management savings factors to these baseline figures to derive estimated savings in each state and year.The methodology used is described in detail below.
Baseline Data Compilation: The Menges Group downloaded and utilized CMS data on Medicaid fee-for-service (FFS) pharmacy usage and costs from two data sources: State Drug Utilization data files and Medicaid Statistical Information Systems (MSIS) data files.The information in these data files was used to establish a baseline volume of Medicaid prescriptions, the brand/generic mix of these prescriptions, and the total expenditure of these prescriptions.For each state, the following baseline information was obtained, tabulated, or estimated for CY2011:
- Number of Medicaid FFS prescriptions: brand, generic, and total
- Published information on each state’s Medicaid FFS dispensing fee
- Average ingredient cost for Medicaid FFS prescriptions – brand, generic, and total
- Average total unit cost for Medicaid FFS prescriptions – brand, generic, and total
- Total expenditures for Medicaid FFS prescriptions – brand, generic, and total
- Volume of prescriptions paid for by Medicaid Managed Care Organizations (MCOs) and associated expenditures; these were estimated by compiling MSIS data on level of capitation payments in each state, and estimating the pharmacy component of those expenditures
Annual Trending of Costs and Usage from 2011 to 2023:The Menges Group trended the baseline costs and usage to CY2023 using the following annual assumptions:
- Prescription volume was trended upward at an annual rate of 1.866%.This trend factor represents the average annual growth in the national Medicaid population from 2003 to 2007 (this time period was selected to avoid the impacts of the recent economic recession thatcreated a particularly high rate of growth in Medicaid enrollment).No changes in baseline prescription volume per beneficiary were assumed.Note that prescription volume increases attributable to the Medicaid expansion population under the provisions of the Affordable Care Act (ACA) were factored in separately as described further below.
- The mix of Medicaid FFS prescriptions is projected to continually evolve toward generics.Each state’s 2011 generic dispensing rate was increased by 1percentage point
- The State Drug Utilization data were typically available into CY2011.FY 2009 MSIS data were used to establish overall Medicaid FFS costs and prescription volume in roughly ten states where the State Drug Utilization files were deemed to be incomplete.Alabama, Idaho, and Oregon have moved to a new model of Medicaid pharmacy reimbursement that is based on average acquisition cost plus a much higher dispensing fee.each year through 2023.The generic dispensing rate was capped at 88% if and when our trending assumptions took a state to this level.
- Prescription drug ingredient unit costs were trended upward by 5.5% annually, which, when used in conjunction with our population growth trend, ties closely on average to the CMS National Health ExpenditureMedicaidprescription cost trend during the upcoming decade.
- Dispensing fees were not trended from their current levels because these fees are not typically tied to any inflation index and do not normally change each year.
Manufacturer Rebates:To derive net governmental savings, it is necessary to factor in the manufacturer rebates that drug manufacturers must pay to the federal government under the provisions of the ACA. These rebates vary from drug to drug but are far higher, on average, for brand medications than for generics.The savings estimates presented in this paper account for the estimated level of these rebates.Note, however, that changes in the dispensing fees paid to pharmacies or in the unit prices paid to pharmacies for ingredient costs (e.g., the savings projected through competitive contracting with low-cost networks) do not impact the level of rebates the federal government receives.
ACA Expansion Population: The initial baseline was adjusted to account for the ACA-authorized expanded Medicaid population beginning in 2014.While each state has the option to participate in the expansion, The Menges Group has assumed that all states would accept the full Affordable Care Act Medicaid expansion. To estimate the impact of the Medicaid expansion population, we relied on data from Kaiser Family Foundation on the expected increases of Medicaid enrollment. We estimated an average of 15 prescriptions per year per recipient for the incoming Medicaid population based on the Tennessee waiver program, which included a population similar to the expected expansion population. We assumed that the enrollment would be a gradual process for states until 2016, with 70% of incoming eligibles being enrolled on average during CY2014, 90% of incoming eligibles being enrolled on average during CY2015, and 100% of the ACA’s expansion population being enrolled throughout CY2016 and thereafter.The Menges Group estimated that the expansion population’s prescriptions would be divided between FFS and MCOs in the same proportion as occurs in each state for their existing adult, nondisabled population.The cost and the growth of Medicaid expansion population prescriptions were trended at the same rate as the current Medicaid population.
Medicaid Managed Care Program Changes:Additional state-specific adjustments to baseline volume and costs were made to account for known Medicaid managed care expansions and pharmacy carve-in program changes.A decrease in Medicaid FFS prescriptions (and a corresponding increase in Medicaid MCO prescriptions) was factored into the following states based on known expansions in Medicaid managed care since the point in time for which that state’s Medicaid FFS prescription volume was tabulated:California, Kentucky, Louisiana, New Hampshire, New York, and Texas.In addition, New York, Ohio, Texas, and Utah have recently changed from a Medicaid managed care pharmacy carve-out model to a carve-in approach.This creates a shift in Medicaid prescription volume from the FFS setting to the MCO setting in each of those states.It is likely that the use of the Medicaid MCO model will continue to grow between now and 2012.It is also likely that additional carve-out states will convert to a carve-in model. However, because it is not known in which states these programmatic changes will occur orwhen they will occur, The Menges Group did not factor these potential changes into our modeling.
The Menges Group projects that roughly $24 billion will be paid to pharmacies in the Medicaid FFS setting nationwide during 2013, with these FFS costs exceeding $430 billion across the ten-year timeframe 2014–2023.Brand medications are projected to represent only 30% of all Medicaid FFS prescriptions during 2013, but are projected to represent 73% of all Medicaid payments to pharmacies.
Impact of “Actual Acquisition Cost” Methodologies Uncertain
Unlike those programs, Medicaid FFS sets ingredient costs reimbursementsadministratively rather than through marketplace negotiations with pharmacies. Contrary to conventional wisdom, states paying high dispensing fees have often paid high ingredient costs as well.
A number of states are experimenting with a new methodology to set rates in Medicaid FFS. This approach relies upon officials setting ingredient cost reimbursements based on data pharmacies report as their Actual Acquisition Cost (AAC) for the drug. Because a pharmacy’s AAC is based on data self-reported by the pharmacy and because states that implement this approach typically increase dispensing fees considerably, it remains unclear if the AAC approach will lower overall costs.
In fact, some Medicaid pharmacy administrators have suggested that the AAC model would increase overall costs because any decrease in the ingredient costs may not counterbalance the associated increase in dispensing fees. Other experts have suggested that AAC fails as a price benchmark because costs are determined from invoice prices at a given point in time that become immediately outdated.
Savings from optimal management of the pharmacy benefit were assumed to occur from the following areas:
Dispensing Fee:Each state’s Medicaid FFS dispensing fee was assumed to drop to $2.00 under an optimally managed program.In prior work for the Pharmaceutical Care Management
CMS, “Medicaid Covered Outpatient Prescription Drug Reimbursement Information by State,” available at http://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Benefits/Prescription-Drugs/Downloads/StateReimbChart2Q2012.pdf
A 2010 study conducted by The Lewin Group, “Potential Federal and State-by-State Savings if Medicaid Pharmacy Programs were Optimally Managed,” found no inverse relationship between a state’s Medicaid dispensing fees and its ingredient payment rates to pharmacies.
“Managing Medicaid Pharmacy Benefits: Current Issues and Options.” Kaiser Commission on Medicaid and the Uninsured, September 2011.
Curtiss, et. al., “What Is the Price Benchmark to Replace Average Wholesale Price (AWP)?” Journal of Managed Care Pharmacy, September 2010.
Association (PCMA), modeling the impacts of New York’s change to a carve-out, it was evident from several MCO interviews that the average dispensing fee achieved by the PBMs (with which the MCOs subcontract) was between $1.50 and $2.00.The $2.00 figure was used in this report as a conservatively high estimate of the average dispensing fees that would occur if this aspect of the FFS Medicaid pharmacy benefit were optimally managed.No additional dispensing fee savings are assumed on prescriptions paid for by Medicaid MCOs.No dispensing fee savings are assumed in the three states (Alabama, Idaho, and Oregon) that have moved to an average acquisition cost payment approachbecause existingdata are currently inadequate to predict the long-term results of this approach.
Drug Mix:Each state’s baseline generic dispensing rate in a given year is assumed to increase steadily over the previous year due to patent expirations.These generic dispensing rates were increased by 5percentage points in each state with the introduction of the active pharmacy benefits management. In addition to the volume shift toward generics, The Menges Group has also factored in an estimate that when brands are used, the mix will shift toward relatively low-cost brand medications.This was factored into the model by assuming a 2.5% reduction in the average ingredient unit price of all brand medications relative to the baseline.These drug mix savings were applied only to Medicaid FFS prescriptions; it is assumed that the Medicaid MCO drug mix is already being actively managed by these health plans’ PBM subcontractors.
Reduced Polypharmacy, Fraud, Waste, and Abuse: A 1% reduction in baseline FFS prescription volume in each state is assumed.This reduction reflects an expectation that in an activelymanaged environment, PBMs will better detect and prevent polypharmacy, fraud, waste, and abuse, including excessive and inappropriate prescribing in areas such as narcotic painkillers.
Competitive Contracting with Limited Pharmacy Networks:Medicaid programs in both the FFS and MCO settings tend to contract with all willing pharmacies.This is consistent with how Medicaid FFS programs contract with physicians, hospitals, and all other provider types.However, it is not consistent with how managed care programs conduct provider contracting in the commercial and Medicare arenas.Medicaid MCOs typically contract with a subset of area providers for most non-pharmacy services (channeling patient volume to their network providers), but Medicaid MCOs typically have not yet utilized a competitive network contracting approach with pharmacies.
There is considerable evidence that competitive pharmacy contracting is yielding savings in other sectors without jeopardizing access to needed medications, as conveyed below.
- Some employers see average savings in the 13% to 18% range from limited pharmacy networks.
- A Millimanstudyconcluded that an employer could achieve savings up to 13% with a
“Walmart’s Booming Preferred Network Models,” Drug Channels, August 25, 2011,http://www.drugchannels.net/2011/08/walmarts-booming-preferred-network.html
limited pharmacy network.
- A leading commercial-sectorlimited pharmacy network guarantees clients a 10% savings compared with an open network.
The transferability of this approach to the Medicaid arena appears promising.A large Medicaid MCO in Florida has implemented a limited pharmacy network consisting of a mass merchant and major chain pharmacy in order to achieve savings of approximately 10%, inline with a major employer that implemented the same network.
The specific potential savings estimates from limited pharmacy contracting in this report are based a sliding scale of the degree to which different percentage reductions in a state’s full pharmacy network will yield pharmacy price discounts in return for the channeling of patient volume that this approach delivers.The estimated savings range from 2% to 10% depending on the concentration of pharmacies within a given state. For each state, we applied the maximum level of savings that could be achieved from a limited pharmacy network that met Tricare’s pharmacy network access standards. Under this approach, each state was applied a potential price savings from selective contracting, on top of all other savings components described above, of 2%, 5%, or 10%.
“The Value of Alternative Pharmacy Networks and Pass-Through Pricing: An Actuarial Analysis,” Milliman, 2010.
“CAT Rolls Out Preferred WAG-WMT Pharmacy Network,” Drug Channels, December 1, 2009.