Corcoran et al v. CVS Health Corporation

Northern District of California, cand-4:2015-cv-03504

Exhibit 25

Interested in this case?

Current View

Full Text

0 Exhibit 25 REDACTED VERSION OF DOCUMENT SOUGHT TO BE SEALED 0 1 IN THE UNITED STATES DISTRICT COURT 2 FOR THE NORTHERN DISTRICT OF CALIFORNIA 3 OAKLAND DIVISION 4 Christopher Corcoran, et al. Case No. 15-civ-03504-YGR 5 on behalf of themselves and all others similarly situated, CLASS ACTION 6 Plaintiffs, 7 v. REBUTTAL REPORT OF 8 CVS Pharmacy, Inc. PROFESSOR JOEL W. HAY, PhD 9 10 Defendant. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 REBUTTAL REPORT OF JOEL W. HAY 0 1 I. INTRODUCTION 2 1. I, Joel W. Hay PhD, previously submitted an expert report on December 9, 2016 ("Hay 3 Report"). 1 I have been asked by counsel for Plaintiffs to assess and respond to CVS Pharmacy, Inc.'s 4 December 9, 2016 expert report of Brett E. Barlag ("Barlag Report"). 5 2. In addition, I have updated some of my analyses set forth in the Hay Report in response 6 to arguments and statements in the Barlag Report, as well as additional information identified in 7 discovery. 8 3. A copy of my updated curriculum vitae is attached hereto as Exhibit A. 9 4. A listing of legal cases where I have testified at trial or by deposition in the last four years 10 is attached hereto as Exhibit B. 11 5. In conducting my analysis, I have reviewed the materials specifically identified in the 12 Hay Report, as well as those listed throughout this report and identified in Exhibit C. 13 6. I understand that discovery is continuing in this case, as is my analysis. I reserve the 14 opportunity to update my analysis as additional information becomes available. Further, I reserve the 15 right to supplement or modify my opinions, if warranted, and to prepare additional supporting materials, 16 such as summaries, graphical exhibits, or charts. 17 II. SUMMARY OF OPINIONS 18 7. Opinion 1: CVS frequently charged non-HSP cash customers HSP and prorated HSP 19 prices. CVS's transaction data shows that the pharmacy frequently charged cash customers in non-HSP 20 transactions the standard HSP prices or pro-rated HSP prices. For example, for products defined as 21 "Usual Drugs" in the Barlag Report, I find that the 2011-2015 standard HSP price of $11.99 was the 22 single most common price CVS charged in non-HSP cash transactions, across all quantities. 23 Furthermore, in nearly 5 million instances (one of every four non-HSP cash transactions), CVS charged 24 cash customers the HSP price, and in nearly 7 million instances (one out of every three non-HSP cash 25 transactions), CVS charged cash customers the HSP price or lower. Thus the HSP discounted prices 26 were widely available to cash customers outside the HSP program, and therefore not limited to HSP 27 members. 28 1 Names and terms identified in the Hay Report have the same definition in this Rebuttal Report. REBUTTAL REPORT OF JOEL W. HAY, PHD 1 0 1 8. Opinion 2: Even excluding the HSP program transactions from consideration, CVS 2 systematically inflated the U&C prices it reported in TPP transactions. In response to the statements 3 of several witnesses from whom CVS submitted declarations, as an alternate analysis, I have examined 4 the prices CVS charged customers in cash transactions not categorized as HSP program transactions in 5 CVS's transaction data. My analysis reveals that the prices CVS submitted to TPPs and PBMs as its 6 supposed U&C prices are significantly inflated even when compared directly to the prices CVS charged 7 in non-HSP cash transactions. CVS's reported U&C prices exceed the lowest first percentile of prices 8 for comparable non-HSP cash transactions in 98 percent of TPP transactions. Including HSP 9 transactions, the reported U&Cs also exceed the first percentile of cash transactions in 98 percent of 10 TPP transactions. Further, CVS's reported U&C is among the highest prices charged by CVS to cash 11 customers. Thus, whether one considers the HSP price to be the U&C or not, the transactions data shows 12 CVS systematically reported inflated U&C prices. 13 9. Opinion 3: CVS did not uniformly collect HSP membership fees from HSP members. 14 The Barlag Report asserts that prices available under the HSP program cannot be U&C prices because 15 members paid a membership fee. In my opinion, the HSP program prices needed to be included in 16 CVS's calculation and submission of U&C prices whether or not the HSP program required payment of 17 a membership fee. However, even Mr. Barlag admits the data demonstrates that not every HSP member 18 paid the HSP membership fee each year they were enrolled. Mr. Barlag theorizes that because CVS did 19 not collect the fee by adding it to patients' first transaction until September 29, 2009, that explains why 20 the transaction data shows not all fees were paid each year. In response, I re-analyzed CVS's transaction 21 data to assess the payment of HSP membership fees accounting for Mr. Barlag's assumption that CVS 22 data do not capture HSP membership fees until September 29, 2009, and found that, even accounting 23 for this limitation, at least 31.68 percent of HSP members did not pay the HSP membership fee at least 24 once every 365 days while actively enrolled in the program. Additionally, the fees CVS collected are 25 $1,710,422 less than what uniform collection would indicate. 26 10. Opinion 4: The HSP program was broadly available to cash customers. The Barlag 27 Report inappropriately concludes that the HSP program was not broadly available to cash customers 28 because individuals rationally chose to join or not join the HSP program. The rationality of joining the REBUTTAL REPORT OF JOEL W. HAY, PHD 2 0 1 program and whether individuals would save money is irrelevant to the program's availability. It is 2 undisputed that the program was available to cash customers in all class states during the class period. 3 Further, more than 3 million non-HSP member cash customers across all class states paid the HSP price 4 for their prescriptions, effectively giving them the benefits of the program. 5 11. Opinion 5: The transaction data for the named Plaintiffs' relevant purchases indicate 6 that Plaintiffs meet the class definition. Contrary to CVS's arguments, each plaintiff made a purchase 7 of one or more drugs covered by the HSP program using insurance. Using CVS's own data, it can be 8 proven that CVS submitted a U&C price to each plaintiff's TPP or PBM that was inflated above the 9 HSP price. Each plaintiff was charged a copay that exceeded the HSP price – which was CVS's true 10 U&C price for the drugs. Each plaintiff also has a qualifying transaction in the 1st percentile fee screen 11 model, a component of the alternate analysis that I have developed using only CVS cash transactions 12 not categorized as HSP transactions. Thus, each plaintiff was damaged under either analysis. 13 12. Opinion 6: Ongoing purchases by Plaintiffs are a small percentage of their 14 transactions and damages. I have been asked to analyze the number of purchases involving HSP drugs 15 that the plaintiffs made after they joined this lawsuit where they were overcharged. Six plaintiffs have 16 such transactions, and in sum they amount to less than 4 percent of the transactions of the named 17 plaintiffs as a whole, and 4 percent of the plaintiffs' total overcharge damages. 18 13. Opinion 7: Mr. Barlag mischaracterizes my contract analysis in his December 9, 2016 19 Report. In his Report, Mr. Barlag mischaracterizes my contract analysis when he suggests that I used 20 the "Condor Code alone to 'map' prescription purchases to contractual U&C definitions." 2 As clearly 21 stated in my December 9, 2013 Report I "mapped" condor codes to contractual U&C definitions per the 22 declaration of CVS's Senior Director of Payer Relations, Susan Colbert. In doing so, I took into account 23 the condor code, date, and address of the transaction as well as the PBM and payor relationships with 24 CVS. As a result of reviewing this information, I have revised the coding of the end dates of certain 25 CONDR_PLAN_NBRs in my analysis. This is discussed in more detail in section G, and updated 26 damages and class member counts are attached in Exhibit D. 27 14. Opinion 8: Mr. Barlag's analysis of the frequency of the HSP-based U&C price is 28 2 Barlag Report ¶ 103. REBUTTAL REPORT OF JOEL W. HAY, PHD 3 0 1 flawed and misleading. Mr. Barlag contends that my calculated HSP-based U&C prices are intrinsically 2 flawed and contrary to industry standards if my calculated U&C prices are not paid by cash customers. 3 3 However, Mr. Barlag's analysis of the frequency at which the HSP-based U&C I have calculated was 4 paid by cash customers itself is flawed, in that it treats as invalid damages from HSP-based U&Cs where 5 cash customers actually paid even less than the calculated HSP-based U&C. Adjusting for this flaw in 6 Mr. Barlag's analysis eliminates virtually the entirety of the resultant impact on damages that he purports 7 to calculate in his report. 8 15. Opinion 9: Mr. Barlag incorrectly asserts that further information is needed to 9 determine the impact of CVS's inflated U&C on the Plaintiff's copayments. Mr. Barlag makes the 10 claim in his December 9, 2016 Report that further information is needed to determine the impact of the 11 inflated U&Cs CVS submitted to TPPs on the Plaintiffs' copayments. Specifically, he identifies the 12 lack of information regarding theoretical deductibles, maximum out-of-pocket limits, copayment 13 reimbursements, "complete" records of all benefit claims, and complete adjudication logic. 4 The 14 information about which Mr. Barlag speculates pertains to actions that largely occur upstream of the 15 "lesser of" claims adjudication logic and the application of the U&C fee screen, thus rendering them 16 irrelevant to the impact of CVS's inflated U&C prices on Plaintiffs' and class members' copayments. 17 III. BASIS OF OPINIONS 18 16. In reaching my opinions below, I have reviewed and rely on pharmacy industry materials, 19 which I cite in this report, as well as documents that CVS has produced in discovery, and testimony 20 from CVS witnesses who have been deposed. I rely on the same retail pharmacy transaction data 21 provided by CVS as described in the Hay Report. I also rely on my more than two decades of experience 22 teaching and conducting research at the USC School of Pharmacy and teaching thousands of pharmacy 23 students about how drugs are sold, purchased and paid for. I further rely on my more than three decades 24 of experience in evaluating and researching drug prices and pharmaceutical market transactions. 25 17. In addition, I have considered CVS's Opposition to Plaintiffs' Class Certification, and 26 the materials attached to that filing. In particular, I have reviewed the declaration submitted by Susan 27 3 28 Barlag Report ¶¶ 120, 121. 4 Barlag Report ¶¶ 107-110 REBUTTAL REPORT OF JOEL W. HAY, PHD 4 0 1 Colbert, CVS's Senior Director of Payer Relations; the declaration of several employees of pharmacy 2 benefit managers ("PBMs"); and the report submitted by Brett E. Barlag, who purported to conduct 3 several analyses of CVS's transaction data in response to my October 3 Declaration and at CVS's 4 request. I also have been provided with updated information from counsel regarding contracts and 5 condor codes that fall within the definition of the class that Plaintiffs seek the Court to certify. 6 7 A. Opinion 1: CVS frequently charged non-HSP cash customers HSP and prorated HSP prices. 8 18. CVS frequently charged non-HSP cash customers HSP prices. As I indicated in my 9 December 9, 2016 Report, the most common price transacted among cash customers for Usual Drugs 10 were typically HSP prices. Table 1 shows that the most common non-HSP cash price for Usual Drugs 11 was the HSP price for the HSP quantity. 12 13 Table 1: Most common non-HSP cash prices for "Usual Drugs" at the HSP quantity. 5 14 Price Count 2008-2010 10.00 32,635 15 2011-2015 11.99 95,116 16 19. Mr. Barlag argues that my analysis is misleading because 90-day supplies are a small 17 percentage of the overall transactions. He argues that 30-day supplies represent the majority of 18 transactions. 6 I believe Mr. Barlag's criticism itself is misleading. For the entire class period, 2008 19 through 2015, $11.99 is the most common price for non-HSP cash purchases of 30-day supplies. In 20 fact, using CVS's definition of "standard" HSP prices for Usual Drugs, 7 CVS's transaction data shows 21 that during the class period, non-HSP cash customers paid the HSP price in one of every four non-HSP 22 cash transactions. If we include transactions where non-HSP cash customers paid even less than the 23 HSP price, we find that non-HSP cash customers paid the HSP price or less in one of every three non- 24 HSP cash transactions. The prevalence of the HSP price in non-HSP cash transactions clearly shows 25 26 5 "HSP Quantity" is defined as quantity dispensed under HSP formularies. Typically, these quantities represent a 90-day 27 supply. In my opinion, the $10 price point is equivalent to the $9.99 HSP price for Usual Drugs from 2008-2011. 6 Barlag Report ¶¶ 88-91. 28 7 $9.99 from 2008-2010 for HSP listed quantities and lower and prorated up for higher quantities. $11.99 from 2011-2015 for HSP listed quantities and lower and prorated up for higher quantities. REBUTTAL REPORT OF JOEL W. HAY, PHD 5 0 1 that patients did not need to pay a membership fee or even sign up for the HSP program to receive the 2 HSP price. 3 4 Table 2: Frequency of HSP Prices Paid by Non-HSP Cash Customers for "Usual Drugs." 5 Non-HSP Cash Transactions 19,840,468 6 Non-HSP Cash Transactions at HSP Price 4,962,640 7 Non-HSP Cash Transaction at or below the HSP Price 6,617,539 8 9 20. Using Mr. Barlag's methodology to identify unique customers, I identified 3,143,393 10 cash customers who received the HSP price in non-HSP cash transactions. In contrast, I identified 11 516,135 unique HSP members. More than six times as many non-HSP cash customers received the HSP 12 price for HSP-eligible prescription drugs than actual HSP members. These data indicate that CVS did 13 not limit the HSP price to HSP members. 14 21. Contrary to Mr. Barlag's statements in his report, in smaller quantities, the prorated HSP 15 price is among the most frequently transacted prices for non-HSP cash customers. During the class 16 period there were 233,103 non-HSP cash transactions for the Usual Drugs at one-third the HSP quantity 17 (30-day supply) that occurred at one-third the HSP price. From 2008 through 2010, $4 was the second 18 most common price and from 2011 through 2015 $4 was the third most common price for cash non- 19 HSP transactions for "Usual Drugs" at one-third the HSP quantity. This clearly shows that prorated 20 HSP prices were made available to the general public whether they signed up for the HSP program or 21 not. 22 23 Table 3: $4 Non-HSP Cash Prices for "Usual Drugs" at one-third the HSP quantity. 24 Price Count 2008-2010 4.00 202,353 25 2011-2015 4.00 219,392 26 27 22. Mr. Barlag attempts to explain away prices below $11.99 by stating they are charged for 28 REBUTTAL REPORT OF JOEL W. HAY, PHD 6 0 1 drugs that are over-the-counter ("OTC") drugs, or drugs that are part of a price matching program. 8 2 First, as Mr. Barlag is aware, all drugs on the HSP formulary, including OTC drugs, are within the class 3 definition and subject to the U&C screen. Mr. Barlag cannot dismiss these drugs simply because they 4 are OTC. Second, Mr. Barlag states that the "second group" of sub-$11.99 transactions is the product 5 of price-matching, and then attempts to ascribe to me a view that price-matched transactions do not need 6 to be taken into account when determining the U&C. This is simply false. Pharmacies are welcome to 7 match competitor's prices or reduce prices at the point-of-sale, but when they do, those prices are made 8 available to the general public and therefore are eligible as true Usual and Customary prices. In fact, 9 CVS is currently the subject of a lawsuit by the state of Texas, alleging that CVS submitted inflated 10 U&C prices not only by excluding the HSP program prices but also by excluding prices offered under 11 price matching or register adjustments. 9 12 13 B. Opinion 2: Even excluding the HSP program transactions from consideration, CVS systematically inflated the U&C prices it reported in TPP transactions. 14 23. In response to the statements of CVS's witnesses, as I noted in my December 9, 2016 15 Report, that frequency does not, in their view, play a role in calculating U&C prices, but instead the 16 price should be the lowest cash price, I have examined the U&C prices CVS reported to TPPs and PBMs 17 for the HSP products and found that in virtually all instances, they exceed the lowest cash price charged 18 by the pharmacy. 19 24. I have identified the cash prices CVS made available to the public by using as a measure 20 the first percentile of non-HSP cash prices at which CVS transacted for the same GCN, year, and 21 quantity dispensed. For 97.9 percent of TPP transactions, the reported U&C exceeds the first percentile 22 price of cash non-HSP transactions. Table 5 below shows that CVS submitted artificially inflated U&Cs 23 every year. 24 25 26 27 8 Barlag Report, ¶ 72. 28 9 Texas ex rel. Winkelman v. CVS Health Corp., In the United States District Court for the District of Travis County in the State of Texas, Civil Action No. D-1-GV-14000388, Corrected First Amended Petition, January 5, 2017. REBUTTAL REPORT OF JOEL W. HAY, PHD 7 0 1 Table 5: TPP Transactions where Reported U&C Exceeds 1st Percentile of Cash Prices for 2 Identical GCN, Year, and Quantity. 3 TPP Transactions where Number of TPP Reported U&C Exceeds 4 Transactions First Percentile of Cash % 2008 15,002,928 13,654,134 91.0% 5 2009 108,212,821 104,771,445 96.8% 6 2010 112,137,732 111,384,465 99.3% 2011 118,675,824 117,068,755 98.6% 7 2012 124,887,208 123,161,045 98.6% 2013 123,452,654 122,025,850 98.8% 8 2014 125,799,895 121,886,925 96.9% 9 2015 122,741,244 118,713,832 96.7% Total 850,910,306 832,666,451 97.9% 10 11 25. The reported U&C is among the highest prices charged by CVS to cash customers. An 12 analysis of the range of prices CVS charged cash customers for each GCN, QTY and trailing year shows 13 that on average CVS's submitted U&C prices were approximately 10 times greater than the lowest cash 14 prices CVS charged. For example, a cash customer purchased a 30-day supply of Lovastatin 40 MG 15 Tablets on January 23, 2009 at store number 573 for $4. During that same day at the same store CVS 16 submitted a U&C of $47.99 to Express Scripts for a transaction involving the exact same drug and 17 quantity. In fact, the scenario described above repeated itself at store numbers 651, 219, 3237 and 3920 18 on April 20, 2012, February 9, 2011, November 16, 2013 and March 18, 2015. 10 As CVS's data 19 indicates and as the table below reflects, this was not an isolated incident. Time and time again, CVS 20 submitted significantly higher U&Cs to third-party payors while at the same time charging cash 21 customers far lower prices. 22 23 24 25 26 27 10 These transactions involved different payors, indicating that the price disparity was not a unique feature of CVS's 28 relationship with one PBM or payor. REBUTTAL REPORT OF JOEL W. HAY, PHD 8 0 1 Table 6: Comparison of Cash Prices to CVS Submitted U&C. 11 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 C. Opinion 3: CVS did not uniformly collect HSP membership fees from HSP members. 19 26. The Barlag Report indicates that HSP prices differ from cash prices in that HSP members 20 paid a membership fee. 12 I disagree. In my opinion, the HSP program prices needed to be included in 21 CVS's calculation and submission of U&C prices whether or not the HSP program required payment of 22 a membership fee, for the reasons I explain in my December 9, 2016 Report. However, both Mr. Barlag 23 and I agree that the data demonstrates that not all members paid the fee for each year they were enrolled. 24 I have updated my analysis of the payment of HSP membership fees to account for the assertion in Mr. 25 Barlag's report that CVS did not collect enrollment fees via a $10 or $15 charge added to the price of 26 11 27 The "Lowest Cash Price", "Prorated HSP Price", "1st Percentile Price" and "U&C Submitted by CVS" are based on averages across all applicable drugs and quantities. The "HSP Price 2008-10" and "HSP Price 2011-15" are the list price 28 for standard quantities of the HSP program "Usual Drugs," and inserted in Table 6 as reference points. 12 Barlag Report, ¶ 84. REBUTTAL REPORT OF JOEL W. HAY, PHD 9 0 0 1 concludes that 84.52 percent of HSP members paid the fee at least once. 16 Thus, Mr. Barlag's own 2 analysis indicates that not all members paid the fee. Mr. Barlag attempts to explain away the 15.48% 3 of members who did not pay the fee at all by claiming these members might have paid the fee during 4 the time in which CVS was not collecting the fee in HSP transactions. 17 My analysis specifically 5 excludes these potential transactions. 18 The transaction data shows that 136,764 HSP members did not 6 pay the fee in at least one 365-day period of enrollment, and at least 36,829 members did not pay a 7 membership fee at all. 8 29. CVS was well aware that the membership fees were not being uniformly collected. CVS 9 created and reviewed reports demonstrating that fees were often discounted or not collected at all. 19 10 D. Opinion 4: The HSP program was broadly available to cash customers. 11 30. The Barlag Report apparently concludes that HSP prices were not broadly available 12 because individuals rationally chose to join or not join the HSP program. 20 As an initial matter, in my 13 opinion, the supposed differences that Mr. Barlag observes in the characteristics of HSP versus non- 14 HSP members have no bearing on whether HSP prices should be considered as U&C prices. Mr. Barlag 15 points to no authority that the "differences" he discusses in his report are recognized as meaningful in 16 the pharmacy industry, and I do not believe they are. 17 31. In addition, Mr. Barlag disregards the fact that CVS charged HSP and non-HSP members 18 alike the HSP prices (or prorated or approximate HSP prices, such as $10 compared to the $9.99 HSP 19 Usual Drug price from 2008-2010). And as noted above, CVS gave more than 3 million non-HSP 20 member cash customers the benefit of paying HSP prices. These facts, in my view, are far more 21 important than the extraneous considerations Mr. Barlag discusses to show the lack of meaningful 22 differences between the HSP and non-HSP cash customers. 23 32. Furthermore, the fact that cash customers chose to enroll in HSP when it made financial 24 25 16 Mr. Barlag does not explain how what constituted paying the membership fee. It is unclear whether he counted partial or excessive payments. 26 17 Barlag Report, ¶¶ 54-56. 18 27 Specifically, in this analysis I exclude from consideration transactions prior to 365 days after a member's first HSP transaction, for members with HSP transactions prior to September 29, 2009. 19 28 CVSC-0012217, CVSC-0022108, CVSC-0022109, CVSC-0031007, CVSC-0031009 20 Barlag Report, ¶¶ 60-62, 85, 87, 89, 97. REBUTTAL REPORT OF JOEL W. HAY, PHD 11 0 1 sense does not indicate a lack of availability of the HSP prices. Rather, the fact that the program was 2 available in all states including the class states during the class period shows it was broadly available to 3 everyone. The program was available to all customers. There were no limitations based on age, gender, 4 occupation, income, geography, time-of-day, days-of-the-week or any other exclusion for any member 5 of the general public. 6 33. Moreover, as Mr. Barlag himself admits, CVS had no intention or desire to actively 7 encourage customers to join the HSP program. While available to the general public, HSP was a purely 8 defensive program to protect CVS market share from encroachment by lower-priced competitors such 9 as the Walmart $4 generics program for their price-sensitive customers. 21 In reaching his opinions, Mr. 10 Barlag appears to assume that non-HSP members made fully-informed, rational decisions, not to join 11 the program, but CVS's lack of marketing and promotion of the HSP program, coupled with their failure 12 to systematically explain to their customers what their cost effective options were, and the reality that 13 many customers never even learned of the HSP program, undercuts this assumption and thus his 14 opinions. 15 E. Opinion 5: The transaction data for the named Plaintiffs' relevant purchases indicate that Plaintiffs meet the class definition. 16 34. In its Opposition, CVS makes several assertions that a number of Plaintiffs did not have 17 transactions with CVS that qualify for the class definition. I disagree with that assertion. As explained 18 in my initial report, I have examined CVS's transaction data for drug purchases that the named Plaintiffs 19 made in this litigation. 20 35. Using information about Plaintiffs and their purchases from Excel spreadsheets that CVS 21 produced in this litigation, I identified in my class composition and damages models those purchases of 22 Plaintiffs that were qualifying transactions: a purchase of a drug offered through the HSP program that 23 a Plaintiff made with insurance associated with a plan in the class, for which the Plaintiff was charged 24 a copayment that exceeded CVS's HSP price or prorated HSP price applicable at the time. A list of all 25 such transactions is included in Exhibit E to this report. 26 36. Additionally, I confirmed that for each of the relevant transactions of the named 27 Plaintiffs, CVS in fact charged Plaintiffs a copayment that exceeded U&C as defined by the alternate 1st 28 21 Barlag Report ¶¶ 52, 60, 61. REBUTTAL REPORT OF JOEL W. HAY, PHD 12 0 1 percentile fee screen approach, which I describe further in this Rebuttal Report. A list of all such 2 transactions is included in Exhibit F to this report. 3 37. I also was asked to identify which of the Plaintiffs only had qualifying transactions based 4 on prorating below the standard HSP prices, i.e., where the Plaintiff's transactions within the class 5 definition only involved copayments inflated above a prorated HSP price for quantities below the HSP 6 quantity. The Plaintiffs are Christopher Corcoran, Amanda Gilbert, Zachary Hagert and Walter Wulff. 7 38. Mr. Barlag incorrectly concludes that in addition to the above four Plaintiffs, four other 8 Plaintiffs (Zulema Avis, Gilbert Brown, Carolyn Caine, and Toni Odorisio) only had qualifying 9 transactions involving copayments above a prorated HSP price for quantities below the HSP quantity.22 10 Mr. Barlag appears to reach this conclusion by ignoring the transaction data he spent almost his entire 11 report discussing and instead relying exclusively on interrogatory responses by plaintiffs who were 12 identifying transactions based on incomplete and preliminary information. In short, Mr. Barlag's 13 contention is not supported by CVS's transaction data. 14 F. Opinion 6: Ongoing purchases by Plaintiffs are a small percentage of their transactions 15 and damages. 16 39. I have also been asked to analyze the number of purchases involving HSP drugs that the 17 plaintiffs made after they joined this lawsuit, using their insurance, involving HSP drugs, where they 18 were overcharged, while the HSP program was ongoing. Using the HSP price-based class composition 19 and damages approach, six plaintiffs have such transactions, and in sum they amount to less than 4 20 percent of the transactions of the plaintiffs as a whole, and 4 percent of the plaintiffs' total overcharge 21 damages. These figures are even smaller using the alternate 1st percentile fee screen approach. A list 22 of all such transactions is included in Exhibit G to this report (using the HSP-based class composition 23 and damages approach), and in Exhibit H to this report (using the 1st percentile fee screen approach). 24 G. Opinion 7: Mr. Barlag mischaracterizes my contract analysis in his December 9, 2016 25 Report. 26 40. In his Report, Mr. Barlag mischaracterizes my contract analysis when he suggests that I 27 used the "Condor Code alone to 'map' prescription purchases to contractual U&C definitions." 23 As 22 28 Barlag Report ¶ 98 23 Barlag Report ¶ 103. REBUTTAL REPORT OF JOEL W. HAY, PHD 13 0 0 1 2 3 4 Absent contrary evidence from CVS, in my opinion it is appropriate to include transactions associated 5 with this condor code in my class composition and damages calculations, given how CVS has identified 6 this code in its transaction data. 7 43. Mr. Barlag criticizes my report for including damages for 8 because I did not consider that, in his opinion, the 27 9 However, the contract Mr. Barlag references did not take effect until. There were only 10 11 28 12 13. 29 14 44. Mr. Barlag states that I may have 15. However, as discussed above, damages for 16 CONDR_PLAN_NBRs associated with these entities were only included during time periods and/or for 17 states where qualifying contracts were in effect. I have 18 19 20 21 H. Opinion 8: Mr. Barlag's analysis of the frequency of the HSP-based U&C price is flawed and misleading. 22 45. Mr. Barlag contends that my calculated HSP-based U&C prices are intrinsically flawed 23 and contrary to industry standards if my calculated U&C prices are not paid by cash customers.30 24 However, Mr. Barlag's analysis of the frequency which the HSP-based U&C was paid by cash customers 25 26 26 CVSC-0343101 27 27 Barlag Report ¶ 104 28 CVSC-0343087, CVSC-0285949 28 30 Barlag Report ¶¶ 120, 121 REBUTTAL REPORT OF JOEL W. HAY, PHD 15 0 1 is flawed in that it seeks to invalidate damages for HSP-based U&Cs where cash customers actually paid 2 even less than the calculated HSP-based U&C. 3 46. Mr. Barlag erroneously excludes transactions where cash customers paid less than the 4 HSP-based U&C price when assessing frequency and validity of the HSP-based U&C prices I have 5 calculated. If cash customers were offered prices even lower than the HSP-based U&C price, an HSP- 6 based U&C price certainly is an appropriate metric for the U&C prices that CVS ought to have 7 submitted, i.e., CVS should have submitted U&C prices at least as low as the HSP-based prices, where 8 CVS is charging cash prices for the same drugs that are at or below the HSP-based price. 9 47. Correcting for this error in Mr. Barlag's analysis eliminates more than 95% of the 10 damages he claims relate to unpaid HSP-based U&Cs. What remains is a negligible sum of damages 11 related to a small pool of HSP-based U&Cs for very rare combinations of drugs and quantities. These 12 are so rare that on average there are only 17 cash transactions for each combination in the entire data 13 set. This pool of outliers only accounts for approximately one percent of cash transactions. 14 48. Mr. Barlag states that my "own HSP-based prices are intrinsically flawed and contrary 15 to industry standards" if they are not routinely paid by cash customers. 31 Putting aside the flaws in his 16 analysis of how routinely they were paid, Mr. Barlag's assertion displays a fundamental 17 misunderstanding of the definition of U&C and the industry standards. I evaluated the frequency of 18 CVS's submitted U&C prices to determine if they made up some overwhelming majority of cash prices 19 thus rendering the HSP price an outlier among CVS's cash transactions. I found just the opposite. 20 CVS's submitted U&C prices were not just far higher than the lowest cash price CVS charged to the 21 general public (whether considering HSP or non-HSP cash sales), but they were also infrequently paid 22 by cash customers. 23 24 I. Opinion 9: Mr. Barlag incorrectly asserts that further information is needed to determine the impact of CVS's inflated U&C on the Plaintiff's copayments. 25 49. Mr. Barlag makes the claim in his December 9, 2016 Report that further information is 26 needed to determine the impact of the inflated U&C's CVS submitted to TPPs on the Plaintiffs' 27 copayments. Specifically, he asserts that there is a lack of information regarding theoretical deductibles, 28 31 Barlag Report ¶ 121 REBUTTAL REPORT OF JOEL W. HAY, PHD 16 0 1 maximum out-of-pocket limits, copayment reimbursements, "complete" records of all benefit claims, 2 and complete adjudication logic. 32 This information pertains to actions in the adjudication process that 3 largely occur upstream of the "lower of" claims adjudication logic, and the application of the U&C fee 4 screen renders them irrelevant to the impact of CVS's inflated U&C on Plaintiffs' and class members' 5 copayments. 6 50. Deductibles and maximum out-of-pocket limits, where applicable, are considerations 7 upstream from the potential copayment being subject to the "lesser of" pricing logic. Thus, these 8 considerations are irrelevant to my analyses. "Complete" records of all benefit claims, in this context, 9 are only relevant to determining the patient's status within his or her deductible phase and/or his or her 10 maximum out-of-pocket limit and are thus further upstream even than the deductible and maximum out- 11 of-pocket limit. They, too, are irrelevant to my analyses. Copayment reimbursements by a TPP or 12 employer, as vaguely described by Mr. Barlag, can occur after the lesser of logic in non-coordination of 13 benefit transactions. 33 However, such reimbursement to patients, if it occurs, is irrelevant to the effect 14 of CVS submitting inflated U&Cs to TPPs on Plaintiffs' and class members' copayments. By definition, 15 copayment reimbursements occur outside of the claims adjudication process and thus do not impact any 16 of my analyses relating to the U&C prices CVS actually submitted versus what it ought to have 17 submitted, or to my calculations of class composition and damages. 18 51. All relevant transactions under the class definition are subject to "lower of" claims 19 adjudication logic and the U&C fee screen. The U&C fee screen is a common-sense solution designed 20 to prevent insured patients from having to pay more than uninsured patients. The potential copayment, 21 whether that figure includes the deductible or whether it has been reduced due to some other cost-sharing 22 arrangement such as the maximum out-of-pocket limit, is compared to the pharmacy's submitted U&C 23 price, and whichever is lower is the copayment figure that the TPP or PBM instructs the pharmacy to 24 charge the patient at the point of sale. This is the final check in the process, thus guaranteeing the patient 25 receives the benefit of his or her insurance, unless the pharmacy submits an inaccurate and inflated 26 U&C, as CVS has done in this case. 27 32 28 Barlag Report ¶¶ 107-110 33 Both Mr. Barlag and my analysis account for potential coordination of benefits effects on the copayment. REBUTTAL REPORT OF JOEL W. HAY, PHD 17 0 1 52. The evidence of the uniformity of the "lower of" adjudication logic in CVS's insured 2 transaction business is clear in the transaction data. As stated in my previous December 9, 2016 Report 3 only a de minimis number of transactions in CVS's data are ever processed where the copayment exceeds 4 the submitted U&C. 5 *** 6 53. I reserve the right to revise or supplement my opinions and their bases as a result of 7 additional information provided to me through discovery, and further analysis of CVS's transaction data, 8 contracts, and other documents and information. 9 I declare that the foregoing is true and correct to the best of my knowledge and belief. 10 11 Dated: January 27, 2017 12 13 Joel W. Hay 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 REBUTTAL REPORT OF JOEL W. HAY, PHD 18