Quidel Corporation v. Siemens Medical Solutions USA, Inc.

Opinion

5 6 7 8 9 10 UNITED STATES DISTRICT COURT 11 SOUTHERN DISTRICT OF CALIFORNIA 12 QUIDEL CORPORATION, Case No. 16-cv-3059-BAS-AGS 13 Plaintiff, ORDER DENYING 14 DEFENDANTS’ MOTION TO EXCLUDE THE SURVEY, 15 v. EXPERT REPORT, AND TESTIMONY OF MATTHEW G. 16 SIEMENS MEDICAL SOLUTIONS EZELL USA, INC., et al., 17 [ECF No. 135] Defendants. 18 19 20 Plaintiff Quidel Corporation and Defendants Siemens Medical Solutions USA, 21 Inc. and Siemens Healthcare Diagnostics Inc. have each retained expert witnesses to 22 opine on portions of this litigation. One of Plaintiff’s experts is Matthew G. Ezell, 23 who conducted a survey and wrote an expert report based on his findings from the 24 survey. Defendants seek to exclude the survey, report, and testimony of Mr. Ezell. 25 (“Mot.,” ECF No. 135.) For the foregoing reasons, the Court DENIES the Motion. 26 27 1 I. Relevant Background1 2 Plaintiff and Defendants each manufacture and sell assays (blood tests) that 3 can aid in the detection of Graves’ disease. Plaintiff sells the Thyretain Bioreporter 4 TSI Assay (“Thyretain”), and Defendants sell the IMMULITE 2000/2000 XPi TSI 5 Assay (“IMMULITE”). 6 Plaintiff alleges Defendants have engaged in false advertising and unfair 7 competition due to Defendants’ advertising of IMMULITE. Plaintiff’s claims stem 8 in part from a statement on Defendants’ website that says IMMULITE detects “TSI 9 only.” A “TSI only” assay is one that detects only thyroid stimulating 10 immunoglobins (“TSI”), as opposed to an assay that fails to differentiate between 11 thyroid stimulating and thyroid blocking immunoglobins (“TBI”). Assays that are 12 unable to differentiate between TSI and TBI are called “TRAb” assays. Plaintiff 13 alleges IMMULITE is not a “TSI only” assay, and that Defendants’ false advertising 14 caused customers to purchase Defendants’ product over Plaintiff’s product and thus 15 damaged Plaintiff. 16 Plaintiff engaged Mr. Ezell to conduct a consumer survey and provide expert 17 opinion regarding Plaintiff’s allegations that “Siemens’ literally and deliberately 18 false statements influenced the ‘purchasing decisions’ of the relevant audience of 19 Siemens’ misstatements.” (“Opp’n,” ECF No. 175, at 2.) Ezell surveyed “physicians 20 that specialize in endocrinology and who, as part of their practice, order assay tests 21 to assist in patient diagnosis.” (“Ezell Report,” Exhibit 12 to Declaration of Erik 22 Haas, ECF 135-3, at ¶ 7.) 23 For the survey, Ezell asked a test group of physicians to review an excerpt 24 from Defendants’ website regarding IMMULITE, and he asked a control group to 25 review an edited excerpt. (Ezell Report at ¶ 17.) The test group reviewed the 26 27 1 A more extensive background section is available in the Court’s order on the parties’ cross- 1 || following excerpt: TEST CELL HIGHLIGHTED MATERIAL The IMMULITE” 2000/2000 XPi TSI assay is the first automated and 3 semiquantitative TSI assay available today. TSH receptor antibody (TRAb) assays A detect both thyroid-blocking and -stimulating antibodies. However, blocking antibodies inhibit TSH stimulation of thyroid cells and lead to hypothyroidism. 5 The IMMULITE 2000/2000 XPi TSI assay detects thyroid stimulating antibodies, 6 the specific cause of GD pathology, with 98.5% specificity.

7 || The control group reviewed the following excerpt: 8 CONTROL CELL HIGHLIGHTED MATERTAL 9 The IMMULITE® 2000/2000 XPi assay is an automated and semiquantitative assay designed for the more specific detection and measurement of stimulating 10 antibodies, but has a potential to detect blocking antibodies and does not differentiate between blocking and stimulating antibodies. The IMMULITE® Il 2000/2000 XPi assay detects thyroid stimulating antibodies, the specific cause 12 of GD pathology, with 98.5% specificity. 13 14 The respondents were asked what message(s) were communicated by the 15 ||material they viewed. (/d.) They were then asked if the material communicated 16 || “anything about IMMULITE assay’s ability to detect TSI only” and if so, what. (/d.) 17 ||They were then asked whether they understood that IMMULITE does or does not 18 ||detect TSI only, or whether IMMULITE is a TRAD assay. (/d.) They were also 19 ||asked open-ended questions about what the material communicates about whether 20 || IMMULITE detects TSI only and about IMMULITE’s ability to detect TSI only. 21 ||Ud.) They were then asked whether they were likely to order both a TSI only and 22 || TRAb assay, and why. (/d.) 23 Ezell concluded that approximately 67.42% of the relevant universe is likely 24 ||to be misled or deceived by Defendants’ false message. (/d. § 8.) He defined a 25 || “false message” as one that states IMMULITE 1s a TSI assay, detects TSI only, or is 26 ||not a TRAb assay. (/d. § 19.) He concluded that Defendants’ webpage is likely to 27 ||mislead a substantial portion of the relevant universe “into believing (1) that 28 ||Defendants’ IMMULITE Assay is a “TSI assay,’ (2) that Defendants’ IMMULITE 1 Assay detects only thyroid stimulating antibodies, and/or (3) that Defendants’ 2 IMMULITE Assay is not a TRAb assay.” (Ezell Report at ¶ 9.) 3 Defendants move to strike Ezell’s report and opinions. 4 II. Legal Standard 5 The trial judge must act as the gatekeeper for expert testimony by carefully 6 applying Federal Rule of Evidence 702 to ensure specialized and technical evidence 7 is “not only relevant, but reliable.” Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 8 579, 589 & n.7 (1993); accord Kumho Tire Co. Ltd. v. Carmichael, 526 U.S. 137, 9 147 (1999) (holding Daubert imposed a special “gatekeeping obligation” on trial 10 judges). In exercising its gatekeeping function, a court “may, in an appropriate case, 11 exclude a flawed survey report from being received into evidence.” 6 McCarthy on 12 Trademarks and Unfair Competition § 32.1158 (5th ed. 2019). 13 Consumer surveys may be used as evidence to show that the alleged 14 misrepresentations have misled, confused, or deceived the consuming public. 15 Southland Sod Farms v. Stover Seed Co., 108 F.3d 1134, 1140 (9th Cir. 1997). The 16 Ninth Circuit has “held that survey evidence should be admitted as long as it is 17 conducted according to accepted principles and is relevant.” Fortune Dynamic, Inc. 18 v. Victoria’s Secret Stores Brand Mgmt., Inc., 618 F.3d 1025, 1036 (9th Cir. 2010) 19 (alteration omitted). “The admissibility threshold for survey evidence in the Ninth 20 Circuit is notably low.” Townsend v. Monster Beverage Corp., 303 F. Supp. 3d 1010, 21 1025 (C.D. Cal. 2018).2 22 III. Analysis 23 Defendants seek to exclude Ezell’s survey, opinion, and report for a variety of 24 reasons, each of which the Court discusses in turn. 25 26 27 2 Both parties spend a good portion of their briefs analyzing cases decided by courts in other 1 A. The Population Surveyed 2 Defendants first argue that Ezell surveyed an irrelevant population. The 3 universe for Ezell’s survey was “comprised of physicians that specialize in 4 endocrinology and who, as part of their practice, order assay tests to assist in patient 5 diagnosis.” (Ezell Report at ¶ 7.) Defendants assert Plaintiff has consistently alleged 6 that laboratories, not physicians, were misled by Defendants’ advertising. (Mot. at 7 11.) Defendants believe Ezell’s survey is irrelevant because physicians are the end- 8 users of the assays, but do not actually purchase the assays, and thus any survey 9 evidence of their impressions of Defendants’ website is immaterial. (Id. at 12.) 10 One of the criteria a court considers in assessing the validity and reliability of 11 a survey is whether “the proper universe was examined and the representative sample 12 was drawn form that universe.” Medisim Ltd. v. BestMed LLC, 861 F. Supp. 2d 158, 13 166 (S.D.N.Y. 2012). The party must show that those surveyed “are the relevant 14 audience for its false advertising claims.” Kwan Software Eng’g, Inc. v. Foray 15 Techs., Inc., No. C 12-3762 SI, 2014 WL 572290, at *5 (N.D. Cal. Feb. 11, 2014). 16 In Kwan, the court rejected a survey because the party did not show “that any of the 17 members of the survey are the people who would see the alleged misrepresentations” 18 or are those “whose decision to purchase the product could be influenced.” Id. 19 To support their argument that physicians are not relevant in this case, 20 Defendants first point to Plaintiff’s operative complaint. The complaint alleges that 21 laboratories, who are Plaintiff’s “direct customers[,]” “have an incentive to purchase” 22 Defendants’ cheaper product “and rely on the face of Siemens’ misleading 23 marketing.” (First Amended Complaint, ECF No. 12, ¶ 21.) In contrast, clinicians, 24 who are not Plaintiff’s direct customers, “rely on the results of the tests to treat 25 patients.” (Id.) Although not altogether clear, it appears Plaintiff is implying that 26 physicians base their purchasing decisions on a product’s merit, but laboratories make 27 decisions based on advertising. Defendants argue Plaintiff cannot now claim that 1 concretely define the relevant market of consumers in its complaint. Plaintiff is not 2 precluded from arguing the breadth of the relevant market at this time. 3 Defendants next point to the deposition of Plaintiff’s expert Jennifer Sipos, 4 who testified that at her institution, “when a TSI is ordered [by a clinician], there is 5 no indication on the report of which assay (Roche, Thyretain, Immulite, etc.) was 6 utilized” and the physician only receives the results from the test, i.e. the 7 measurement of TSI. (Exhibit 8 to Declaration of Erik Haas, ECF No. 137, at 9.) 8 Defendants argue this shows that the physicians do not distinguish between the 9 assays, and therefore their opinions of Defendants’ description of IMMULITE is not 10 relevant. 11 Plaintiff disagrees and argues that physicians do order specific assay tests. 12 (Opp’n at 5–6.) Plaintiff points to the deposition of Defendants’ employee, Carole 13 Dauscher. Ms. Dauscher testified that Defendants previously hired a marketing 14 agency to conduct a marketing campaign aimed at clinicians. (Exhibit 1 to 15 Declaration of T. Kevin Roosevelt, ECF No. 177, at 106:1–9.) According to Ms. 16 Dauscher, the agency marketed to physicians as opposed to laboratories because “it’s 17 really important to educate the physicians . . . [b]ecause if they don’t order the test, 18 then there’s . . . no point of having it in the laboratory.” (Id. at 106:15–25.) 19 Physicians were not the ones buying the assays, but they became informed of the tests 20 through marketing, and they could go to Defendants’ website or talk to the 21 laboratories for more information. (Id. at 107:1–6; 108:18–23.) Further, Dr. 22 Silberman, a director of Sonic/CPL laboratory testified that the clinicians are 23 “substantially” in charge of deciding which type of assay to run. (Exhibit 3 to 24 Declaration of T. Kevin Roosevelt, ECF No 177, at 117:14–22.) Clinicians consult 25 with the laboratory, and some clinicians “have the capability of [sic] and ask for [the 26 assay] by name . . . and in some circumstances they will specify.” (Id.) 27 The evidence shows the physicians are not simply end-users of the assay with 1 product. See Kwan Software Eng’g, 2014 WL 572290, at *4 (excluding survey 2 evidence because it had not been proven that those surveyed were “potential 3 purchasers of the product—those whose decision to purchase the product could be 4 influenced”). This is not a situation where all physicians blindly use whatever assay 5 the laboratory happens to carry, with no input into what assay they use on patients. 6 Given the conflicting testimony, it is possible the physicians’ opinions regarding the 7 products are relevant and their opinions could be influenced by marketing or website 8 information. 9 The Court finds that a survey of physicians could be relevant in this case. Of 10 course, the findings would not be relevant to the laboratories’ reaction to Defendants’ 11 marketing or website, but only to the physicians’ reaction. See Spraying Sys. Co. v. 12 Delavan, Inc., 975 F.2d 387, 394 n.5 (7th Cir. 1992) (finding when the surveys 13 targeted farmers rather than the “actual purchasers” of the product, the selection of 14 farmers as the relevant universe “limits the surveys’ probative value”). The Court 15 declines to strike the Ezell report for this reason. 16 B. The Questions 17 Defendants next argue that the survey questions are ambiguous. (Mot. at 13.) 18 Defendants find it problematic that Ezell used the terms “TSI only” and “TRAb 19 assay” in the survey without defining the terms. (Id.) Plaintiff appears to admit the 20 respondents received no definition of “TSI only.” (Opp’n at 13.)3 And Ezell admits 21 he did not ask the respondents what they understood “TSI only” to mean, (“Ezell 22 Depo.,” Exhibit 13 to Declaration of Erik Haas, ECF No. 135-3, at 212:24–25) nor 23 did he ask them how they defined a TRAb assay, (id. at 240:17–19). Of course, 24 whether the assay is a TSI only assay or a TRAb assay is a critical question in this 25 case, as it forms the basis for much of Plaintiff’s claims. 26 Plaintiff argues this is an issue for the jury. As the Ninth Circuit has explained, 27 1 when evaluating a survey, the court first asks: is the survey admissible, meaning “is 2 there a proper foundation for admissibility, and is it relevant and conducted according 3 to accepted principles?” Click Billiards, Inc. v. Sixshooters, Inc., 251 F.3d 1252, 4 1263 (9th Cir. 2001). Once the survey is admitted, the jury decides “follow-on issues 5 of methodology, survey design, reliability, the experience and reputation of the 6 expert, critique of conclusions, and the like.” Id. “Technical unreliability” issues go 7 to the weight of the survey, not its admissibility, and are issues for the jury. E. & J. 8 Gallo Winery v. Gallo Cattle Co., 967 F.2d 1280, 1292 (9th Cir. 1992). 9 The issue of ambiguity of survey questions is one that has been looked at by 10 various courts. Defendants point to Wallace v Countrywide Home Loans, Inc., No. 11 08-1463-JST (MLGx), 2012 WL 11896333, at *5 (C.D. Cal. Aug. 31, 2012), where 12 the court analyzed a survey that asked about the respondents’ “typical” work week. 13 The court determined the understanding of the word “typical” was not uniform and 14 it excluded the survey for this reason and various other reasons. In Townsend v. 15 Monster Beverage Corp., 303 F. Supp. 3d 1010 (C.D. Cal. 2018), the defendants took 16 issue with two surveys. They objected that one of the surveys asked suggestive or 17 vague questions, and also objected that the expert did not measure how respondents 18 had interpreted the statements in another survey. Id. at 1024. The court found that 19 these objections go to the weight rather than the admissibility of the surveys and 20 declined to strike the surveys. Id.; see also United States v. 400 Acres of Land, more 21 or less situate in Lincoln Cty. Nev., No. 2:15-cv-1743-MND-NJK, 2017 WL 22 4797517, at *5 (D. Nev. Oct. 24, 2017) (holding the objection that the survey 23 questions are ambiguous does not affect the survey’s admissibility); Brighton 24 Collectibles, Inc. v. RX Texas Leather Mfg., 923 F. Supp. 2d 1245, 1258 (S.D. Cal. 25 2013) (finding the survey questions to be “sloppy” and problematic, but holding this 26 issue can be explored through cross examination). 27 The bulk of cases hold that an objection regarding the phrasing of survey 1 category of “survey design.” The Ninth Circuit has specifically found that this issue 2 goes to the weight of the survey, not the admissibility. The Court declines to strike 3 Ezell’s survey on this basis. 4 C. Leading and Biased Questions 5 Defendants next argue that the survey questions led and biased the 6 respondents. The Ninth Circuit has held that an objection that a survey asked leading 7 or biased questions goes to the weight, not the admissibility of the survey. Southland 8 Sod Farms, 108 F.3d at 1143; see also Medlock v. Taco Bell Corp., No. 1:07-cv- 9 1314-SAB, 2015 WL 8479320, at *5 (E.D. Cal. Dec. 9, 2015) (“The Court finds that 10 Defendants’ criticisms of [the expert’s] wording of the questions goes to the weight 11 of the evidence, and not the admissibility of the survey.”). Surveys can be admitted 12 even if they contain “highly suggestive” questions, as long as the survey is 13 “conducted according to accepted principles and [is] relevant.” Fortune Dynamic, 14 Inc. v. Victoria’s Secret Stores Brand Mgmt., Inc., 618 F.3d 1025, 1037 (9th Cir. 15 2010) (citation omitted). For example, a survey that exposes the respondent to the 16 desired response before asking the critical connection may be “given little weight.” 17 McCarthy, at § 32.172. But this is an issue for the jury. The Court declines to strike 18 the survey on this basis. 19 D. Survey Format 20 Defendants argue that the website excerpt shown to the control group was 21 biased because it used “gratuitous language.” (Mot. at 21.) Defendants cite Bobrick 22 Washroom Equipment, Inc. v. American Specialties, Inc., No. CV 10-6938 SVW 23 PLA, 2012 WL 3217858, at *18 (C.D. Cal. Aug. 8, 2012), aff’d, 565 F. App’x 660 24 (9th Cir. 2014), where the court excluded a survey because the ads shown to the 25 control group were “substantially different” than the ads shown to the test group. For 26 this reason, the “survey format effectively predetermined its result.” Id. But Bobrick 27 is distinguishable. Here, the differences between the website excerpt shown to the 1 the result of the survey. This objection therefore goes to the “design” of the survey 2 and is an issue for the jury. The survey should not be excluded for this reason.4 3 E. Ezell’s Conclusions 4 Defendants finally take issue with Ezell’s conclusions. After coding all 5 responses, Ezell concluded that 81.87% of the test group respondents received a false 6 message, and 14.45% of the control group respondents received a false message. 7 First, Defendants argue Ezell does not disclose how he coded the respondents’ 8 answers. (Mot. at 22.) This is incorrect. Ezell stated: if the respondent “gave a false 9 message in terms of an open-ended response or a close-ended response, they would 10 be in category 1. And if there was no false message at all, they would be in category 11 2. And sometimes the respondent might say one thing in terms of their open-ended 12 response and somewhat contradict themselves in terms of their close-ended 13 response.” If so, they were put into category 3, which is “indeterminable.” (“Ezell 14 Depo.,” Exhibit 13 to Declaration of T. Kevin Roosevelt, ECF No. 175-1, at 161:2– 15 11.) Thus, Ezell explained how he coded the responses, and this is not a reason to 16 exclude the survey. 17 Defendants next argue that Ezell improperly coded the responses of the control 18 19 4 Defendants also bring up another objection to the control group excerpt. Defendants argue the 20 excerpt “was written entirely by counsel to Quidel, without [Ezell’s] input.” (Mot. at 20 (emphasis in original).) Citing Elliott v. Google, Inc., 860 F.3d 1151, 1160 (9th Cir. 2017), Defendants argue 21 this alone is reason to exclude the entire survey. The Court finds various flaws in this argument and encourages Defendants not to 22 overexaggerate facts or misinterpret cases in a misleading way. First, Ezell testified that while Plaintiff’s counsel drafted the website excerpt for the control group, Ezell reviewed it and agreed 23 it was appropriate. (See Ezell Depo. at 113:10–17.) Therefore, it is inaccurate for Defendants to 24 state that Ezell had no input on the issue. And second, Elliott does not hold that a survey should be excluded when it was designed by counsel, as Defendants state. (Mot. at 20 n.9.) Instead, the 25 surveys in Elliott were entirely designed and conducted by counsel “who is not qualified to design or interpret surveys.” 860 F.3d at 1160. Therefore, the surveys were stricken. But there is no 26 question that Ezell conducted the survey here, and the fact that one portion of the survey was drafted by Plaintiff’s counsel does not mean the entire survey should be excluded under Ninth Circuit 27 precedent. See McCarthy, at § 32:166 (“Attorney cooperation with the survey professional in 1 ||group. Defendants argue if Ezell had properly coded the responses, more 2 ||respondents in the control group would have been confused by the survey. For 3 ||example, Ezell classified 23 physicians in the control group as “indeterminable” but 4 □□ Defendants argue 17 of those 23 respondents should have been classified as having 5 ||received a “false message.” (Mot. at 22.) Defendants argue if Ezell had properly 6 ||coded these 17 respondents, almost 25% of the control group would have been 7 ||misled, and the entire survey would therefore have to be excluded. (/d. at 22—23.) 8 Again, a “critique of the [survey’s] conclusion” goes to the weight of the 9 ||survey, not its admissibility. See Clicks Billiards, 251 F.3d at 1263; see also 10 || Microsoft Corp. v. Motorola Inc., 904 F. Supp. 2d 1109, 1120 (W.D. Wash. 2012) 11 ||(concluding that criticisms of an expert’s conjoint analysis concerned □□□□□□□ of 12 ||methodology, survey design, reliability, and critique of conclusions, and therefore 13 ||[went] to the weight of the survey rather than admissibility”); Brighton Collectibles, 14 ||923 F. Supp. 2d at 1258 (holding if the objection is that the survey had a “sweeping 15 |}conclusion,” this is a weakness that can be explored through cross examination or a 16 ||contradictory expert opinion). 17 CONCLUSION 18 For the foregoing reasons, none of Defendants’ objections lead the Court to 19 ||conclude that Mr. Ezell’s survey or testimony should be excluded. The Court 20 ||DENIES Defendants’ Motion. (ECF No. 135.) 21 IT IS SO ORDERED. 22 || DATED: October 21, 2019 /) y oY 34 (yitlug (Aaphan 6 How. Cynthia Bashant United States District Judge 25 26 27 28