Launch Price Tracker Methods

Background

Dr. Peter Bach first evaluated the shocking growth in the launch prices of oncology drug and explained the incentives that lead this growth to continue in a 2009 New England Journal of Medicine article. Since then, the Drug Pricing Lab (DPL) has continued tracking the price of newly approved anticancer drugs on a quarterly basis. DPL has expanded the scope to include select non-oncology drugs to illustrate how rising launch prices are pervasive across a variety of drug classes. Over the course of its lifetime, the oncology launch price tracker has been utilized as a tool to communicate and quantify the issue of rising drug costs to policy makers, researchers, and the general public.

Inclusion criteria 

Oncology tracker:  

For selection of drug classes in the oncology launch price tracker, we reviewed each announcement on the FDA’s list of Hematology and Oncology Approvals and Safety Notifications to determine if (1) the announcement was for a drug, (2) if the drug was newly approved, and (3) if the drug was indicated for oncology. Brand name drugs that met all three criteria were added to the oncology launch price tracker. Drugs were included if they were approved beginning in 1965, the year Medicare was signed into law.

Generic, biosimilar, and supportive care drugs were excluded.

 

Non-oncology tracker: 

Selection of therapeutic classes:

For selection of drug classes in the non-oncology launch price tracker, we included drug classes that contributed to the highest spending in the US. To do this we compared spending for the top 100 drugs across two measures: those with highest Medicare spending in 2017 from the Medicare Drug Spending Dashboard and those with the highest Wholesale Acquisition Cost (WAC) sales in quarter 2 of 2019. We used the most recently available data from both sources at the time of extraction. For each drug that appeared in both the top 100 list for Medicare spending and WAC sales we assigned a therapeutic drug category and class as defined in the United States Pharmacopeia Drug Classification (USP-DC) Files

There were 54 drugs that appeared in both Medicare spending and WAC sales lists, and of those, antineoplastic (oncology) drugs were most common. Excluding oncology drugs, the most frequently occurring drug categories were immunological agents, antivirals, and blood glucose regulators. Categories listed in the USP-DC files are broadly defined, so we narrowed the scope of our inclusion criteria to the USP Therapeutic Classes represented in our list of top drugs by spending. The drugs present in the immunological agents category included the drug classes of ‘immunosuppressants’ and ‘immunological agents, other’ and excluded others such as ‘vaccines’ and ‘immunoglobulins’. The classes within the therapeutic category of blood glucose regulators that were present in the list of top spending included ‘antidiabetic agents’, ‘glycemic agents’, and ‘insulins’. The therapeutic category of antivirals included ‘anti-HIV agents’ and ‘anti-hepatitis agents’. The drug class of ‘multiple sclerosis agents’ was also included as we found that it follows the same exponential increase in launch prices as the other non-oncology drug classes of interest.

Selection of drugs: 

Beginning in 1990, drugs that were newly approved according to the Drugs@FDA database and whose launch indication fell within one of the USP Categories and Classes of interest were included (Table 1). Drugs were matched to USP Category and Class based on drug indication, mechanism of action, and formulary key drug types found on the FDA website. When available, the USP Category and Class were confirmed using the USP Drug Classification 2021 File. All generics, biosimilars, and brand name drugs with subsequent indications that match our USP categories of interest are excluded from our analysis.

Table 1. Drug Categories and Classes Incorporated in Non-Oncology Tracker and the Number of Drugs Included Between 1990 and 2020

Data extraction: 

After establishing the list of drugs included in each tracker, we extracted the following summary variables for each drug from the launch label (found on Drugs@FDA):

  • Approval year
  • Manufacturer
  • Accelerated approval status
  • Route of administration
  • Launch indication
  • Recommended dosing
  • Packaging – as well as noting whether the drug is packaged in single-dose or single-use vials

If the launch label was unavailable, we extracted the above information from the oldest available label.

 

Dosing Calculations:

To estimate monthly price, we first calculated the recommended dose administered over the first 12 weeks of therapy before dividing by 2.77, the average number of months in 12 weeks. We calculated dosing for the first 12 weeks of therapy to account for the fact that some drugs have dose-escalation or induction phases. A known limitation of this method is that it extends the price of one-time administered drugs (e.g. CAR-T therapies) over the course of a 12-week period.

To calculate the monthly cost over the treatment course specified, we divided the treatment course by the following values to arrive at monthly price:

  • 4.3 weeks in a month
  • 30.4 days in a month

 

Insulin: 

Most of the drugs in our classes of interest have clearly defined dosing guidelines which we used to calculate the recommended dose over the course of the first 12 weeks of therapy. However, the recommended dose for insulin drugs is dependent on clinical patient characteristics. For these drugs, we reviewed existing literature for reported average doses and the drug labels for the reported mean insulin dose administered in clinical trials. We were only interested in the dosing estimate on the needs of patients with type 2 diabetes as they make up more than 90% of diabetes patients in the United States. Based on our review, we found that the mean daily dose ranged from 0.3-1 U/kg. To provide conservative pricing estimates, calculations assumed a dose of 1,000 U/month, equivalent to one vial per month. We confirmed with a clinician that this estimate is reasonable.

Glycemic Agents: 

Unlike insulin drugs, which utilize the same mechanism of action and the same unit of medicine, glycemic agents follow a variety of different dosing recommendations that are based on individual patient blood glucose response. To provide a conservative price estimate, we calculated the 12-week dose based on the lowest recommended dosing regimen for patients with type 2 diabetes.

 

Price Calculations at Launch:

Gross price calculations:

Estimated gross monthly prices for drugs are based on the Medicare reimbursement rule at the time of FDA approval. Since 2005, Medicare has reimbursed at 106% of the average sales price (ASP) for Part B drugs. ASPs are published in quarterly files released by CMS. There is a two-quarter lag between the collection of this data and its reporting by CMS. For the purposes of extracting prices at launch, we pulled the first-available ASP. For Part D drugs, current prices were retrieved from IBM Micromedex Red Book’s historic record of wholesale acquisition cost (WAC). Payment limits for prior years vary and are described briefly in Table 2 below.

In all cases, the relevant payment limit is applied to a 12 week dosing regimen for an “average” adult weighing 70kg, or with a body surface area of 1.7 meters squared, and divided by 2.77 to arrive at a monthly price (there are, on average, 2.77 months in 12 weeks). The prices shown are for the listed drug only, costs for supportive care or administration fees are not included.

Table 2. Medicare Reimbursement Rule Used to Determine Drug Price

Net price calculations:

Class-level Medicare rebates are not publicly available, so we searched existing literature to estimate them for the prescription drugs in the non-oncology tracker. Blood glucose regulators face the highest average level of rebating, with manufacturers paying rebates around 60% of the gross cost.  The average rebate paid by manufacturers for the remaining therapeutic classes (antivirals, immunological agents, and multiple sclerosis agents) was approximately 13% of gross cost. We applied these discounts to our gross price estimates for Part D drugs to calculate prices net of rebates.

We did not account for Medicare rebates in the oncology tracker given the low level of rebating in protected classes.

Median monthly price:

To track trends in these drug prices over time, we calculated the median monthly price for all drugs approved within a 5-year interval. In the non-oncology tracker, the 5-year median monthly price was calculated within each USP Therapeutic Category.