Scott v. Arizona Center for Hematology and Oncology PLC

• Court: District Court, D. Arizona
• Decided: April 29, 2020
• Docket: 2:16-cv-03703
• Status: Unknown

Opinion

1 WO 2 3 4 5 6 IN THE UNITED STATES DISTRICT COURT 7 FOR THE DISTRICT OF ARIZONA

9 United States of America, ex rel. J. Scott, No. CV-16-03703-PHX-DGC

10 Plaintiff, ORDER

11 v.

12 Arizona Center for Hematology and Oncology, PLC, et al., 13 Defendants. 14

15 Arizona Center for Hematology and Oncology, PLC, 16 Counterclaimant, 17

v. 18

19 J. Scott,

20 Counterdefendant.

21 22 Relator J. Scott has filed a qui tam action against Defendants Arizona Center for 23 Hematology and Oncology PLC, doing business as Arizona Center for Cancer Care 24 (“AZC”), and Drs. Terry Lee, Daniel Reed, and Christopher Biggs, alleging violations of 25 the False Claims Act (“FCA”), 31 U.S.C. § 3729 et seq. Doc. 47. AZC counterclaims for 26 breach of fiduciary duty. Doc. 122 at 62, 64.1 The parties cross-move for summary 27 judgment. Docs. 222, 233. Defendants also move to exclude the opinions of Drs. Abraham

28 1 Citations in this order are to page numbers at the top of each page. 1 Wyner and William Noyes (Docs. 225, 246), the government moves for leave to file a 2 statement of interest (Doc. 231), and Scott moves to file certain documents under seal 3 (Doc. 234). The motions are fully briefed, and oral argument was held by teleconference 4 on April 24, 2020. The Court will deny Defendants’ motions to exclude Scott’s experts, 5 grant Defendants’ motion for summary judgment in part, deny Scott’s cross-motion for 6 summary judgment, consider the government’s statement, and grant the motion to seal. 7 I. Daubert Motions. 8 Defendants move to exclude the opinions of Dr. Abraham Wyner, Scott’s statistical 9 sampling expert, and Dr. William Noyes, Scott’s medical billing expert. Docs. 225, 246. 10 Under Rule 702, an expert may offer “scientific, technical, or other specialized knowledge” 11 if it “will assist the trier of fact to understand the evidence,” provided the testimony rests 12 on “sufficient facts or data” and “reliable principles and methods,” and “the witness has 13 reliably applied the principles and methods to the facts of the case.” Fed. R. Evid. 14 702(a)-(d). The proponent of expert testimony has the ultimate burden of showing, by a 15 preponderance of the evidence, that the proposed testimony is admissible. See Cooper v. 16 Brown, 510 F.3d 870, 942 (9th Cir. 2007); Fed. R. Evid. 104(a). The trial court acts as a 17 gatekeeper to assure that the testimony “both rests on a reliable foundation and is relevant 18 to the task at hand.” Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 597 (1993). 19 A. Dr. Abraham Wyner. 20 Dr. Wyner uses statistical sampling to estimate alleged healthcare overpayments 21 received by AZC. He has produced three reports. See Docs. 225-5, 225-6, 225-7. 22 Defendants argue that (1) he is not qualified to estimate healthcare overpayments, (2) his 23 sampling methodology is not reproducible, (3) his samples are not representative, 24 and (4) his reports are replete with errors. Doc. 225 at 1. 25 1. Qualifications. 26 Defendants argue that while Dr. Wyner may have academic experience in statistics, 27 he has no experience estimating alleged healthcare overpayments and no knowledge of 28 relevant government guidelines. Docs. 225 at 6-7, 245 at 4. Defendants contend that he 1 has never worked for any government agency on a healthcare audit and cannot opine on 2 the amount of overpayments purportedly made by the government to AZC. Doc. 225 at 6. 3 Defendants rely on the Medicare Program Integrity Manual (“MPIP”) and the Office of 4 Inspector General Toolkit (“OIG Toolkit”) for their argument that an expert must have 5 specific experience in Medicare to opine on healthcare overpayments. Doc. 245 at 2. Scott 6 responds that Defendants’ characterization of Dr. Wyner’s expertise is too narrow and that 7 he is qualified to offer expert opinions regarding statistical analyses. Doc. 232 at 3-4. 8 Dr. Wyner has a Bachelor of Science degree in mathematics from Yale University 9 and a Ph.D in statistics from Stanford University. Doc. 225-5 at 6. He has served as a 10 professor of statistics at the Wharton School of Business at the University of Pennsylvania 11 for more than 20 years and has been the director of the undergraduate program in statistics 12 since 2005. Docs. 232 at 1, 225-5 at 2, 4. Dr. Wyner has authored works in a number of 13 academic journals and has rendered expert opinions on statistics. Doc. 232 at 4. His ability 14 to conduct a statistical analysis is not disputed by Defendants. 15 Dr. Wyner was retained to estimate healthcare overpayments to AZC. Doc. 225-4 16 at 9. Because evaluating all payments processed by AZC is not feasible due to their sheer 17 number, he recommended statistical sampling. Id. Dr. Wyner testified: 18 Statistical sampling . . . is the process by means of which we estimate a 19 quantity in a population whose value is impossible to measure without measuring every individual in the population. And so we do this as a matter 20 of economy and speed, and it allows us to learn something that would be impossible otherwise. 21 22 Id. at 10. 23 FCA cases often involve statistical sampling because, “in view of the enormous 24 logistical problems of [proving fraud in complex government programs], statistical 25 sampling is the only feasible method available.” United States ex rel. Martin v. Life Care 26 Ctrs. of Am., Inc., 114 F. Supp. 3d 549, 560 (E.D. Tenn. 2014) (citing Ill. Physicians Union 27 v. Miller, 675 F.2d 151, 157 (7th Cir. 1982)); see United States v. Fadul, No. DKC 11- 28 0385, 2013 WL 781614, at *14 (D. Md. Feb. 28, 2013) (“Courts have routinely endorsed 1 sampling and extrapolation as a viable method of proving damages in cases involving 2 Medicare and Medicaid overpayments where a claim-by-claim review is not practical”). 3 Defendants present no evidence that Dr. Wyner is not qualified to conduct such a 4 statistical analysis. See Fed. R. Civ. P. 702. Nor do the MPIP and OIG Toolkit require a 5 statistician with expertise in healthcare billing. They merely call for “consultation with a 6 statistical expert.” Doc. 225-3 at 8. The MPIP states: “The sampling methodology used 7 in estimations of overpayments must be reviewed and approved by a statistician or by a 8 person with equivalent expertise in probability sampling and estimation methods.” Id. The 9 MPIP provides a list of minimum qualifications for statistical experts, which includes: 10 • The possession of a “Bachelor’s degree . . . in Statistics or in some 11 related field (e.g., psychometrics, biostatistics, econometrics, mathematics) with significant coursework in probability and 12 estimation methodologies, and at least 6 years of experience applying 13 methods of statistical sampling and interpreting the results”; or

14 • The possession of a “Doctoral degree in statistics or in some related 15 field with significant coursework in probability and estimation methodologies, and at least 1 year of experience applying methods of 16 statistical sampling and interpreting the results.” 17 Id. at 8-9. 18 Dr. Wyner possesses both qualifications. The fact that he has no experience 19 healthcare billing does not render his statistical analysis inadmissible under Rule 702, 20 particularly when Rule 702 “‘contemplates a broad conception of expert qualifications.’” 21 Alsadi v. Intel Corp., No. CV-16-03738-PHX-DGC, 2019 WL 4849482, at *11 (D. Ariz. 22 Sept. 30, 2019) (citing Thomas v. Newton Int’l Enters., 42 F.3d 1266, 1269 (9th Cir. 23 1994)).2 24

25 2 The cases Defendants cite are not helpful. They deal with experts who sought to 26 testify outside their area of expertise. See, e.g., Gable v. Nat’l Broad. Co., 727 F. Supp. 2d 815, 833 (C.D. Cal. 2010) (expert in copyright law not qualified to opine on the substantial 27 similarity between two literary works as he had no “experience, knowledge, training, or education in the literary field”); Vaxiion Therapeutics, Inc. v. Foley & Lardner LLP, 593 28 F. Supp. 2d 1153, 1163 (S.D. Cal. 2008) (attorney with no special ethics training or experience who sought to serve as an expert in legal ethics). 1 2. The Reliability of Dr. Wyner’s Studies. 2 Defendants make three arguments regarding the reliability of Dr. Wyner’s studies. 3 First, Defendants contend that he failed to disclose the random seeds used to 4 generate his random samples. Doc. 225 at 9-10. Dr. Wyner used JMP software to conduct 5 his studies. Doc. 225-7 at 3. That software uses a random number generator, which is a 6 device that generates a random sample from a larger population displaying no 7 distinguishable patterns. Doc. 225 at 3. Dr. Wyner asserts that he did not save the random 8 seeds because there was no way to extract them from the JPM software. Doc. 225-4 at 9 21-22. Without the seeds, Defendants contend that they cannot replicate his samples or 10 verify they are random, thus rendering the samples unreliable under Rule 702. Doc. 225 11 at 10.3 12 Defendants do not assert, however, that Dr. Wyner failed to use a valid statistical 13 program in running his study or that the samples chosen by the software were not in fact 14 random. The OIG Toolkit states: “A sample is random if it is generated from a valid 15 random or pseudo-random number generator.” Doc. 225-1 at 12. Dr. Wyner used and 16 followed the instructions provided by the JMP software, which uses a built-in sampler that 17 is “completely random.” Doc. 225-7 at 3. Defendants do not challenge the validity of the 18 JMP software or its ability to produce a random sample. The Court therefore cannot 19 conclude that Dr. Wyner’s statistical analysis is unreliable. 20 Defendants also argue that Dr. Wyner failed to disclose his sampling frames. 21 Doc. 225 at 10. The database from which a sample is drawn is known as the sampling 22 frame. Doc. 225-1 at 14. Scott contends that Dr. Wyner’s sampling frames are taken from 23 Defendants’ own spreadsheets of AZC’s claims for payment. Doc. 232 at 8. Those 24 spreadsheets are identified in Dr. Wyner’s reports. Docs. 225-5 at 2, 225-6 at 2. 25

26 3 Each computer-generated random sample is associated with a “random seed,” 27 which is used to initialize the random number generator. Doc. 225-1 at 14. When the seed is entered into the random number generator, it will recreate the previous sample rather 28 than provide a new set of random numbers. Id. Without the seed, it is not possible to recreate the random sample. Id. 1 Defendants again rely on the MPIM and OIG Toolkit to support this argument, but 2 they point to no authority indicating that the toolkits are binding. Doc. 245 at 6-7. Indeed, 3 the OIG Toolkit explicitly states that it is merely a “training guide”: 4 [It] is not intended to establish formal standards or to restrict the methods 5 available to the MFCUs to complete their mission. Because statistics is a broad field with a wide variety of effective and valid methods, a sample or 6 estimate may be valid even if it does not follow the steps in this guide. 7 Doc. 225-1 at 3. 8 Nor does the failure to preserve certain documentation, including the sampling 9 frames Defendants identify, render the study unreliable. Defendants do not dispute that 10 their own spreadsheets were used as Dr. Wyner’s sampling frame. And the OIG Toolkit 11 explains that “[t]he failure to keep such documentation does not necessarily render a 12 sample invalid, but it can make the resulting estimate more difficult to defend.” Id. at 16. 13 Defendants may use this point for cross-examination, but the Court cannot conclude that it 14 renders Dr. Wyner’s testimony inadmissible under Rule 702. 15 Second, Defendants contend that Dr. Wyner’s samples are not representative. 16 Defendants’ expert, Dr. Salve, ran three tests to confirm this – the T-test for equality of 17 means, the Kolmogorov-Smirnov test, and the Chi-square test. Doc. 225 at 12-14.4 18 Defendants argue that Dr. Salve’s tests indicate that “it is likely that any extrapolations 19 from [Dr. Wyner’s] samples to their populations would result in a larger overpayment 20 estimate . . . than if the samples were actually representative of the populations from which 21 they were drawn.” Id. at 14. But Dr. Wyner explains in considerable detail why Dr. Salve’s 22 tests are not accurately applied to, and are based on a fundamental misunderstanding of, 23 his statistical analysis. See Doc. 225-7 at 2-8. Defendants do not address this detailed 24 response to Dr. Salve’s criticisms in their motion or reply, and the Court cannot conclude 25 that the criticisms prove the unreliability of Dr. Wyner’s work. 26 27 4 Scott argues that Dr. Salve is not credible because he is not a statistician and is a “quintessential hired-gun expert.” Doc. 232 at 6-7. But Scott does not move to exclude 28 the Salve opinions, and the Court’s task on summary judgment is not to make credibility determinations. 1 Third, Defendants argue that Dr. Wyner’s reports are “replete with errors that render 2 his opinions unreliable under Rule 702 and Daubert.” Doc. 225 at 15. The alleged errors 3 generally are numerical differences between the experts, such as 719 claims (Wyner) 4 versus 717 claims (Salve). Id. at 15-16. But if the Court finds Dr. Wyner qualified to 5 render his opinions and that he has applied reliable principles and methods to the facts of 6 this case reliably – as it does – then the Court’s task is not to determine whether his 7 conclusions are correct. Daubert, 509 U.S. at 595 (the “focus, of course, must be solely 8 on principles and methodology, not on the conclusions they generate”); see also Fed. R. 9 Evid. 702, advisory committee’s note to 2000 amendment (proponents of evidence “do not 10 have to demonstrate to the judge by a preponderance of the evidence that the assessments 11 of their experts are correct, they only have to demonstrate by a preponderance of the 12 evidence that their opinions are reliable” (citation and quotation marks omitted)). Dr. 13 Wyner’s opinions are sufficiently reliable to be admitted under Rule 702. Defendants will 14 be free to challenge the correctness of his conclusions at trial. As the Supreme Court has 15 noted, “[v]igorous cross-examination, presentation of contrary evidence, and careful 16 instruction on the burden of proof are the traditional and appropriate means of attacking” 17 an opposing expert’s opinions. Daubert, 509 U.S. at 596. 18 B. Dr. William Noyes. 19 Dr. Noyes is Scott’s expert on the American Medical Association Current 20 Procedural Terminology (“CPT”) code 77290. Doc. 246-9 at 3.5 Defendants contend that 21 Dr. Noyes, a radiation oncologist, is not qualified to render opinions regarding the propriety 22 of CPT code 77290 as it relates to the delivery of Stereotactic Radiosurgery (“SRS”) and 23 Stereotactic Body Radiation Therapy (“SBRT”). Doc. 246 at 1-2.6

24 5 CPT code 77290 permits a radiation oncologist to bill for a “complex simulation.” 25 Doc. 249 at 5-6. Such a simulation is a “dry run” of a specific treatment, without the actual administration of radiation, and is distinct from simple and intermediate simulations. 26 Doc. 223-11 at 10. Complex simulations typically happen on the first day of treatments administered over multiple days. Scott alleges that Defendants consistently failed to do 27 the work necessary to bill CPT code 77290. See Doc. 1 at 10. 28 6 SRS and SBRT deliver “very high, extremely potent doses of radiation,” with precision not available through other traditional forms of radiation therapy. Doc. 1 Defendants first claim that Dr. Noyes lacks practical experience in SRS and SBRT, 2 as he last performed SRS almost 25 years ago and SBRT almost a decade ago. Id. at 10-11. 3 Defendants contend that this is critical, as the field of radiosurgery has undergone sweeping 4 changes in the last decade alone. Id. at 12. Defendants also argue that Dr. Noyes lacks 5 knowledge of applicable guidelines governing billing for SRS and SBRT. Id. at 13.7 6 SBRT is one form of a general category of therapy called external beam radiation 7 therapy (“EBRT”), where radiation oncologists use a machine called a linear accelerator to 8 administer beams of radiation to cancer sites. Id. at 6. Another common form of EBRT is 9 intensity modulated radiotherapy (“IMRT”), which Dr. Noyes uses in his radiation 10 oncology practice. Id.; Doc. 246 at 11. 11 CPT code 77290 is used to bill for complex simulations regardless of the form of 12 EBRT used to treat the patient. Doc. 249 at 6. Dr. Noyes explains that “simulations are 13 performed for all forms of EBRT, and the basic steps in performing a simulation are the 14 same and not tied specifically to the method of EBRT that is being simulated.” 15 Doc. 249-2 ¶ 7. Defendants’ expert, Dr. Steinberg, agrees that the different kinds of 16 simulations are not tied to the specific forms of EBRT. Doc. 249-1 at 16. 17 Defendants argue that because Dr. Noyes is not currently using SRS or SBRT, he 18 cannot opine on whether Defendants properly billed for complex simulations when they 19 used SBRT. Defendants emphasized during oral argument that each patient, each case, 20 and each therapy session is different, and that highly individualized judgments must be 21 made as to whether and when complex simulations are needed as part of SBRT treatment. 22 But Dr. Noyes’s opinion is not based on his evaluation of the treatment Defendants 23 provided in each case, but on Defendants’ medical records – specifically, the fact that those 24 246-3 ¶ 11. SRS and SBRT are safer and less invasive forms of cancer treatment than 25 standard surgery, and are more cost effective than conventional radiation therapy. Id. ¶¶ 13-14. 26 7 Defendants particularly challenge Dr. Noyes’s conclusion that the services AZC’s 27 physicians provided in connection with each daily fraction of SBRT treatments do not constitute complex simulations, and that even if they do, “[t]here is no clinical justification 28 for an additional complex simulation to be performed prior to the delivery of each and every SBRT therapy.” Doc. 246 at 10. 1 records do not show that complex simulations were performed when Defendants billed for 2 them under CPT code 77290. See Docs. 246-9 at 5 (Defendants’ records either “showed 3 the use of cone beam computerized tomography (“CBCT”) scans for imaging and not 4 complex simulation,” or included “no documentation for any simulations”); 246-14 at 5 5 (“in the sample I reviewed, I saw no evidence in the documentation that [Defendants] were 6 doing complex simulations”). Scott contends that Defendants charged for complex 7 simulations as a matter of course with every SBRT fraction, and Dr. Noyes’s review of the 8 records seems to confirm that fact. He found such billings for every fraction, and found 9 they were always improper based on Defendants’ medical records. Doc. 246-9 at 5 (“I 10 found that in all the cases in the sample I reviewed CPT code 77290 was billed 11 improperly.”). Thus, Dr. Noyes’s opinion is not based on second-guessing the complex 12 simulations decisions made by Defendants in individual SBRT cases. It is based on the 13 fact, according to Scott and Dr. Noyes, that Defendants consistently bill for complex 14 simulations they did not perform. 15 What is more, Dr. Noyes performs hundreds of complex simulations per year and is 16 qualified to opine on whether Defendants’ medical records show complex simulations. 17 Challenges to his lack of expertise with SRS and SBRT technology can be raised on cross- 18 examination. See Wichansky v. Zowine, No. CV-13-01208-PHX-DGC, 2016 WL 19 6818945, at *3-4 (D. Ariz. Mar. 22, 2016) (denying a motion to exclude a medical billing 20 expert because her 30 years of general experience in medical billing and past engagement 21 by attorneys were sufficient qualifications); Contreras v. Brown, No. CV-17-08217-PHX- 22 JAT, 2019 WL 2080143, at *3 (D. Ariz. May 10, 2019) (denying a motion to exclude a 23 medical billing expert where the expert’s general experience in the medical field, and 24 familiarity with billing practices nationally, provided sufficient foundation in her area of 25 expertise). 26 Dr. Noyes is also qualified to opine on when doctors can bill under CPT code 77290. 27 He completed a four-year residency in radiation oncology and has been practicing radiation 28 oncology for more than 20 years. Doc. 249-2 ¶ 3-4. He has received specific training on 1 how to perform and document simulations, including when and how complex simulations 2 may be billed under CPT code 77290. Id. ¶¶ 10, 12. Dr. Noyes has also worked in the 3 development of CPT codes, including with the American Society of Therapeutic Radiation 4 Oncology (“ASTRO”) and the American College of Radiation Oncology (“ACRO”). 5 Id. ¶ 13. In particular, he was a member of ASTRO’s Code Utilization and Applications 6 Committee, the body responsible for considering issues relating to development and 7 utilization of CPT codes. Id. He served as a senior editor on two of their coding guides 8 and was responsible for answering questions submitted by insurance companies, the 9 Centers for Medicare and Medicaid Services (“CMS”), and the public. Id. 10 Defendants’ Daubert motion to exclude the opinions of Drs. Wyner and Noyes will 11 be denied.8 12 II. Summary Judgment. 13 A. Background Information. 14 AZC is a cancer treatment center with a radiation oncology department. Doc. 222 15 at 2-3. Radiation oncology involves the controlled use of radiation to treat cancer. The 16 process generally involves six steps: (1) consultation with the patient; (2) preparation for 17 the treatment, including simulation; (3) physics and dosimetry work to customize each 18 treatment; (4) treatment delivery; (5) treatment management; and (6) follow-up care. 19 Doc. 223 ¶¶ 1, 17. According to ASTRO, steps two and three are often repeated by medical 20 necessity during the treatment course. Id. ¶ 18. 21 22 23 8 Citing Rule 26(a)(2)(B)’s disclosure requirement, Defendants move to strike Dr. Noyes’s declaration submitted in opposition to Defendants’ motion as exceeding the scope 24 of his expert reports. Doc. 252 at 2. But the Court’s Case Management Order, consistent with Rule 26, states that expert reports must set forth “the testimony the witness is expected 25 to present during direct examination, together with the reasons therefor.” Doc. 74 at 3 (citing Rule 26 advisory committee’s note to 1993 amendments) (emphasis added). The 26 intent is to disclose trial testimony. Courts may receive additional evidence when considering Daubert motions, and often hear testimony directly from experts when 27 deciding such motions. Rule 26 therefore does not provide a basis for striking Dr. Noyes’s declaration in this Daubert motion. If Defendants believe information in the declaration 28 should be excluded from trial because it was not included in Noyes’s Rule 26 reports, they certainly may object at trial. 1 Some of AZC’s patients are insured under federal programs such as Medicare, 2 which require AZC to submit claims to be paid for treatment. Doc. 222 at 3. There are 3 rules and guidelines dictating the conditions under which Medicare and similar programs 4 will pay for particular medical services. Id. Radiation oncology billing involves the use 5 of CPT codes, which establish a uniform process for billing medical services. 6 Doc. 223 ¶¶ 24-25. The five-character CPT codes are used by payors to help determine 7 the amount of payment a doctor receives for services provided. Id. ¶¶ 24-25. 8 AZC hired Scott in 2008 to serve as the billing manager of its radiation oncology 9 department. Doc. 222 at 4. He was responsible for overseeing day-to-day billing 10 operations, including properly analyzing claims for accuracy and completeness, serving as 11 AZC’s expert on coding and billing processes, ensuring that AZC’s billing operations were 12 conducted in a manner consistent with payor rules and guidelines, and keeping up-to-date 13 with current coding, billing, and compliance requirements. Id. 14 Scott filed this lawsuit in 2016, alleging that AZC and its treating physicians, Drs. 15 Lee, Reed, and Biggs, submitted false claims for reimbursement from Medicare, Medicaid, 16 and Tricare in violation of the FCA. See Docs. 1, 47. The FCA authorizes individuals 17 such as Scott, known as “relators,” to file civil qui tam suits against persons who present 18 false claims to the government. 31 U.S.C. § 3730. Scott’s claims are based on five alleged 19 billing schemes: (1) Defendants falsely billed CPT code 77290 for complex simulations in 20 SBRT (“Scheme One”); (2) Defendants improperly billed CPT code 77470 for special 21 procedures they did not perform (“Scheme Two”); (3) Defendants filed claims for 22 medically unnecessary computerized tomography (“CT”) scans (“Scheme Three”); 23 (4) Defendants billed for inappropriate brachytherapy treatments (“Scheme Four”); and 24 (5) Defendants improperly billed both private and federal insurance programs for the same 25 treatment, resulting in overpayments that were not refunded (“Scheme Five”). Doc. 47. 26 The Court dismissed all of Scott’s claims based on Schemes Four and Five. 27 Doc. 76. The remaining claims are based on Schemes One, Two, and Three, and allege 28 that Defendants improperly billed CPT codes 77290 (a complex simulation in connection 1 with SBRT), 77470 (a special treatment procedure), and 77295/77301 (second CT scans in 2 conjunction with developing breast and prostate cancer treatment plans). Scott also claims 3 that Defendants fired him in retaliation for asserting these claims. Doc. 47 at 60. 4 B. Legal Standard. 5 A party seeking summary judgment “bears the initial responsibility of informing the 6 district court of the basis for its motion, and identifying those portions of [the record] which 7 it believes demonstrate the absence of a genuine issue of material fact.” Celotex Corp. v. 8 Catrett, 477 U.S. 317, 323 (1986). Summary judgment is appropriate if the evidence, 9 viewed in the light most favorable to the nonmoving party, shows “that there is no genuine 10 dispute as to any material fact and the movant is entitled to judgment as a matter of law.” 11 Fed. R. Civ. P. 56(a). Summary judgment is also appropriate against a party who “fails to 12 make a showing sufficient to establish the existence of an element essential to that party’s 13 case, and on which that party will bear the burden of proof at trial.” Celotex, 477 U.S. 14 at 322. Only disputes over facts that might affect the outcome of the suit will preclude 15 summary judgment, and the disputed evidence must be “such that a reasonable jury could 16 return a verdict for the nonmoving party.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 17 248 (1986). 18 C. Defendants’ Summary Judgment Arguments. 19 Defendants make five arguments: (1) statistical sampling cannot be used to establish 20 liability under the FCA; (2) Scott has no evidence that Defendants submitted false 21 claims; (3) Scott has no evidence of scienter; (4) Scott’s claims period for CPT code 77290 22 is overbroad; and (5) Scott has failed to mitigate his alleged retaliation damages. 23 1. Statistical Sampling. 24 Defendants argue that because Scott’s claims involve subjective, fast-specific, 25 clinical determinations by treating physicians, liability cannot be established through 26 statistical sampling. Doc. 222 at 6. Scott asserts that statistical sampling is widely used in 27 FCA cases to prove liability and damages, and is appropriate here. Docs. 237 at 4-8, 244. 28 1 Scott relies heavily on Tyson Foods Inc. v. Bouaphakeo, 136 S. Ct. 1036 (2016), a 2 case that is helpful but not conclusive. Tyson held that statistical sampling could be used 3 in a class action to determine the average amount of time employees took to don and doff 4 protective clothing – time for which they should have been paid by their employer – and 5 that a statistical class-wide analysis could be used to show that individual issues would not 6 predominate over common issues for purposes of class certification. This conclusion was 7 reached in part because the employer had failed to maintain records that could be used to 8 determine how long each employee took to don and doff the protective clothing. As the 9 Supreme Court noted, “respondents sought to introduce a representative sample to fill an 10 evidentiary gap created by the employer’s failure to keep adequate records,” and “there 11 were no alternative means for the employees to establish their hours worked[.]” Id. 12 at 1047. No similar gap exists here. Scott does not contend that the medical records for 13 each procedure are unavailable. 14 But the Supreme Court did not limit its holding to cases where evidence is 15 unavailable. Indeed, it specifically declined to adopt a general rule on the proper use of 16 statistical evidence. Id. at 1046 (“the Court would reach too far were it to establish general 17 rules governing the use of statistical evidence”). Tyson instead explained that “[w]hether 18 and when statistical evidence can be used to establish classwide liability will depend on 19 the purpose for which the evidence is being introduced and on the elements of the 20 underlying cause of action[.]” Id. (quotation marks and citation omitted). 21 Thus, while Tyson did not establish a general rule that approves Scott’s proposed 22 use of statistical evidence, neither did it reject statistical evidence as inadequate to meet a 23 burden of proof or as a denial of due process, as Defendants contend. Instead, Tyson treated 24 statistical evidence like other forms of proof, noting that “[a] representative or statistical 25 sample, like all evidence, is a means to establish or defend against liability.” Id. Courts 26 must determine whether the evidence is reliable: “Representative evidence that is 27 statistically inadequate or based on implausible assumptions could not lead to a fair or 28 accurate estimate of the uncompensated hours an employee has worked.” Id. at 1048-49. 1 If reliable, statistical evidence may be admitted at trial and the jury must then decide 2 whether it satisfies the plaintiff’s burden of proof: 3 Once a district court finds evidence to be admissible, its persuasiveness is, in 4 general, a matter for the jury. Reasonable minds may differ as to whether the average time Mericle calculated is probative as to the time actually 5 worked by each employee. Resolving that question, however, is the near- 6 exclusive province of the jury. 7 Id. at 1049. 8 A case more directly on point is Ratanasen v. California, Department of Health 9 Services., 11 F.3d 1467 (9th Cir. 1993). Dr. Ratanasen challenged a claim by the California 10 Department of Health Services (“DHS”) that he had “submitted payment claims to Medi- 11 Cal for services he had not actually rendered.” Id. at 1468. DHS’s claim was based on “a 12 sample of 300 Medi-Cal beneficiaries out of a total of 8,761 beneficiaries for whom 13 Ratanasen had submitted claims during the period in question.” Id. at 1469. Like 14 Defendants in this case, Ratanasen objected to the use “of a random sample, which was 15 then used to calculate an estimated overpayment,” arguing that “to reach a true 16 overpayment, each file would have to be examined on its own.” Id. 17 In addressing Ratanasen’s argument, the Ninth Circuit reviewed the decision of four 18 circuit courts of appeals and one district court that had approved the use of statistical 19 sampling in similar situations. Id. at 1469-71. The court reached this conclusion: “We 20 now join other circuits in approving the use of sampling and extrapolation as part of audits 21 in connection with Medicare and other similar programs, provided the aggrieved party has 22 an opportunity to rebut such evidence.” Id. at 1471.9 23 The Ninth Circuit’s decision in Ratanasen comports with the clear weight of 24 authority on using statistical evidence in medical fraud cases. See United States v. Rogan, 25 26 9 The Ninth Circuit also recently approved the use of statistical evidence in a class action to recover compensation from an employer. See Ridgeway v. Walmart Inc., 946 27 F.3d 1066, 1087 (9th Cir. 2020) (“Statistical examples . . . are a means of proving a case. . . . These types of evidence are only permissible when the evidence is reliable in 28 proving or disproving the elements of the relevant cause of action.”) (quotation marks and citation omitted). 1 517 F.3d 449, 453 (7th Cir. 2008) (rejecting as “a formula for paralysis” defendant’s 2 argument that the district court in a health care fraud case must address each of the 1,812 3 medical claims forms); United States v. Lahey Clinic Hosp., Inc., 399 F.3d 1, 18 n.19 (1st 4 Cir. 2005) (statistical sampling may be used to establish amount of Medicare overpayments 5 because “sampling of similar claims and extrapolation from the sample is a recognized 6 method of proof”); Chaves Cty. Home Health Serv. v. Sullivan, 931 F.2d 914, 922-23 (D.C. 7 Cir. 1991) (statistical sampling procedures may be used to determine health care provider’s 8 liability for claiming Medicare reimbursement with respect to services not covered by the 9 Medicare statute); Yorktown Med. Lab, Inc. v. Perales, 948 F.2d 84, 89-90 (2d Cir. 1991) 10 (use of statistical sampling and extrapolation to establish a laboratory’s liability for 11 overcharging Medicaid program is consistent with due process in light of the low risk of 12 error, the government’s interest in minimizing administrative burdens, and the 13 government’s interest in eliminating fraud); Mich. Dep’t of Educ. v. U.S. Dep’t of Educ., 14 875 F.2d 1196, 1205-06 (6th Cir. 1989) (random sampling and statistical methods of 15 extrapolating the sample results to a large universe of claims may serve as evidence of 16 whether the claims complied with statutory limitations on federal payments). 17 Defendants contend that Ratanasen is inapposite because it arose out of a 18 government audit and not an FCA case. But nothing in Ratanasen limits its holding to 19 audits. And contrary to Defendants’ assertion, many courts have allowed statistical 20 sampling in FCA cases. See, e.g., Rogan, 517 F.3d at 453; Life Care, 114 F. Supp. 3d 21 at 556; United States v. Americus Mortg. Corp., No. 4:12-CV-2676, 2017 WL 4083589, at 22 *4 (S.D. Tex. Sept. 14, 2017); United States v. Fadul, No. CIV.A. DKC 11-0385, 2013 23 WL 781614, at *14 (D. Md. Feb. 28, 2013); United States ex rel. Loughren v. Unum 24 Provident Corp., 604 F. Supp. 2d 259, 261 (D. Mass. 2009); United States ex rel. Harris 25 v. Bernad, 275 F. Supp. 2d 1, 8 (D.D.C. 2003). 26 Defendants argue that the use of statistical sampling would violate their right to due 27 process of law. Doc. 222 at 8. The argument seems to be that if Scott is permitted to prove 28 that Defendants made a false claim on a particular occasion without presenting direct 1 evidence of that false claim – if Scott instead can prove the false claim by showing that it 2 was statistically probable that Defendants submitted a false claim on that occasion – then 3 Defendants will be denied due process on that claim. The Court disagrees for at least three 4 reasons. 5 First, while it certainly is true, as Defendants emphasize, that Scott must “prove all 6 essential elements of the cause of action, including damages, by a preponderance of the 7 evidence,” 13 U.S.C. § 3731(d), nothing in the FCA or the Constitution suggests that a 8 preponderance of the evidence cannot be provided by reliable representative evidence. 9 AZC cites no case that so holds. 10 Second, Defendants’ argument seems to ignore the nature of the preponderance of 11 the evidence standard. Id. That standard considers probabilities – it does not require direct 12 proof. As Ninth Circuit Model Jury Instruction 1.6 explains: “When a party has the burden 13 of proving any claim or affirmative defense by a preponderance of the evidence, it means 14 you must be persuaded by the evidence that the claim or affirmative defense is more 15 probably true than not true.” (Emphasis added). Such probability can be proved by a 16 variety of means, including indirect evidence. 17 Third, Defendants’ argument would eliminate common forms of indirect evidence. 18 For example: 19 Evidence of a person’s habit or an organization’s routine practice may be 20 admitted to prove that on a particular occasion the person or organization acted in accordance with the habit or routine practice. The court may admit 21 this evidence regardless of whether it is corroborated or whether there was 22 an eyewitness. 23 Fed. R. Evid. 406. This rule specifically recognizes that indirect evidence may be used to 24 prove an event, even when there is no corroborating or direct evidence of the event. Indeed, 25 evidence of habit or routine is much like statistical proof – a defendant’s actions at other 26 times is used to prove how it most likely acted during the event in question. 27 The law on direct and circumstantial evidence makes the same point: 28 Evidence may be direct or circumstantial. Direct evidence is direct proof of 1 a fact, such as testimony by a witness about what that witness personally saw or heard or did. Circumstantial evidence is proof of one or more facts from 2 which you could find another fact. You should consider both kinds of 3 evidence. The law makes no distinction between the weight to be given to either direct or circumstantial evidence. It is for you to decide how much 4 weight to give to any evidence. 5 6 Ninth Circuit Model Jury Instruction 1.12. Simply stated, the law does not require direct 7 proof of a fact. Facts may be proven by indirect evidence if the jury finds the indirect 8 evidence sufficiently persuasive. 9 If sufficiently reliable to be admitted under Rule 702, the Court concludes that 10 statistical evidence may be admitted to prove Scott’s claims in this FCA case. Defendants 11 have provided no basis for the Court to conclude that such evidence would violate their 12 right to due process of law. See Yorktown, 948 F.2d at 89-90 (use of statistical sampling 13 and extrapolation to establish liability for overcharging Medicaid program is consistent 14 with due process in light of the low risk of error, the government’s interest in minimizing 15 administrative burdens, and the government’s interest in eliminating fraud); Life Care, 114 16 F. Supp. 3d at 570 (FCA medical billing case, holding that “Defendant will be afforded 17 due process by having the opportunity to depose the Government’s [statistical] expert, 18 challenge the qualifications of the Government’s expert, retain its own expert, and to 19 present all of this evidence at trial.”).10 20 Defendants asserted during oral argument that statistical sampling is inappropriate 21 in this case because each patient, each illness, and each treatment is different, and 22 generalization through statistics cannot possibly account for the many individual variables 23 that affect whether a particular procedure in a particular case was warranted. But as 24 discussed above with respect to CPT code 77290, Dr. Noyes does not purport to make 25 individualized determinations about whether a complex simulation was warranted by the 26 circumstances of a particular case. (Nor does Defendants’ expert, Dr. Steinberg.) Dr. 27 28 10 The Court acknowledges that there may be situations where statistical sampling is not appropriate. See, e.g., Cimino v. Raymark Indus., Inc., 151 F.3d 297 (5th Cir. 1998). 1 Noyes instead asserts that Defendants’ billings under CPT code 77290 were improper 2 because their own medical records show that they did not perform the complex simulation 3 for which they billed. He found this to be true in 100% of the cases he examined, a 4 percentage which tends to corroborate Scott’s assertion that complex simulations were 5 billed by Defendants routinely whenever SBRT fractions were administered. See 6 Doc. 246-9 at 11-34. Dr. Noyes’s opinions on CPT codes 77470, 77295, and 77301 are 7 also based solely on his review of medical records. Docs. 223-15; 223-16. Defendants 8 will be free at trial, if they choose, to cross examine Scott’s experts on how their statistical 9 analysis can account for individual patient variations. 10 One other point is important. The Supreme Court has observed that, in complex 11 cases such as this, “a representative sample is ‘the only practicable means to collect and 12 present relevant data’ establishing a defendant’s liability.” Tyson Foods, 136 S. Ct. at 1046 13 (citing Manual of Complex Litigation § 11.493, p.102 (4th ed. 2004)). Scott’s complaint 14 alleges that AZC submitted at least 4,000 false claims. Doc. 47 at 17-18. No reasonable 15 trial could include individualized proof of 4,000 separate occurrences of fraud. And as one 16 district court has observed: 17 Defendant’s position – that statistical sampling simply cannot be applied to an FCA case involving Medicare overpayment – is broad and potentially far- 18 reaching. If accepted, it would materially limit the efficacy of the FCA as a 19 tool to combat fraud against the government. . . . If the Court were to reach the conclusion urged by the Defendant – that a claim-by-claim review is 20 required in every FCA action and that statistical sampling is never 21 permissible – potential perpetrators of fraud would be emboldened by the fact that a claim-by-claim review is often impractical. 22 23 Life Care, 114 F. Supp. 3d at 571. 24 Defendants have had a full opportunity to develop their response to Scott’s 25 statistical evidence, and will have a full opportunity to present their defense at trial. The 26 jury will be instructed clearly on Scott’s burden of proof, and will decide whether Scott’s 27 evidence meets that burden. Due process will not be denied. 28 / / / 1 2. False Claims. 2 An FCA claim includes four elements: “(1) a false statement or fraudulent course 3 of conduct, (2) made with scienter, (3) that was material, causing (4) the government to 4 pay out money or forfeit moneys due.” United States ex rel. Hendow v. Univ. of Phx., 461 5 F.3d 1166, 1174 (9th Cir. 2006). “Evidence of an actual false claim is the sine qua non of 6 a False Claims Act violation.” United States v. Kitsap Physicians Serv., 314 F.3d 995, 7 1002 (9th Cir. 2002). Defendants contend that Scott cannot present evidence of any false 8 claims submitted by Defendants. 9 a. CPT Code 77290. 10 i. Objective Falsity Argument. 11 Scott contends that a complex simulation did not occur before each fraction of 12 SBRT treatment. Doc. 47 at 13-14. This contention is based on Dr. Noyes’s review of 13 AZC’s medical records. Doc. 254 ¶ 124. Dr. Noyes concluded that complex simulations 14 were not shown by the medical records, as it appeared that AZC physicians were merely 15 performing image guidance, which is not the same as a complex simulation. Id. ¶ 122; 16 Doc. 246-9 at 5. 17 Defendants contend that summary judgment is warranted in an FCA case where a 18 plaintiff cannot present evidence that a claim was “objectively false.” Defendants derive 19 this “objective falsity” requirement from United States v. AseraCare, 938 F.3d 1278 (11th 20 Cir. 2019), a case which held that “the mere difference of reasonable opinion between 21 physicians, without more, . . . does not constitute an objective falsehood” and cannot 22 support an FCA claim. Id. at 1301. Defendants also cite United States ex rel. Winter v. 23 Garden Regional Hospital and Medical Center, Inc., Case No. CV 14-08850-JFW, 2017 24 WL 8793222 (C.D. Cal. Dec. 29, 2017), a California district court case which held that “to 25 prevail on an FCA claim, a plaintiff must show that a defendant knowingly made an 26 objectively false representation,” and that “subjective medical opinions . . . cannot be 27 proven to be objectively false.” Id. at *6. 28 1 The Ninth Circuit recently reversed the district court decision in Winter and held 2 that the FCA does not require proof of objective falsity: “Under the plain language of the 3 statute, the FCA imposes liability for all ‘false or fraudulent claims’ – it does not 4 distinguish between ‘objective’ and ‘subjective’ falsity or carve out an exception for 5 clinical judgments and opinions.” Winter ex rel. United States v. Gardens Reg’l Hosp. & 6 Med. Ctr., Inc., 953 F.3d 1108, 1117 (9th Cir. 2020). Although the Ninth Circuit suggested 7 that its holding could be harmonized with AseraCare, it made clear that “to the extent that 8 AseraCare can be read to graft any type of ‘objective falsity’ onto the FCA, we reject that 9 proposition.” Id. at 1119 (citation omitted). The Third Circuit also recently rejected 10 AseraCare’s “objective falsity” requirement. See United States v. Care Alternatives, 952 11 F.3d 89, 99-100 (3d Cir. 2020). 12 As part of their objective falsity argument, Defendants contend that because their 13 expert, Dr. Steinberg, reviewed the same medical records as Dr. Noyes and found that 14 complex simulations did properly occur with each SBRT fraction, Scott cannot establish 15 falsity under the FCA. Defendants assert that “a reasonable difference of opinion among 16 physicians reviewing medical documentation ex post is not sufficient on its own to suggest 17 that those judgments – or any claims based on them – are false under the FCA.” AseraCare, 18 938 F.3d at 1297. 19 The Court does not agree. A “doctor’s clinical opinion must be judged under the 20 same standard as any other representation.” Winter, 953 F.3d at 1113. As the Ninth Circuit 21 explained in its recent decision: 22 A doctor, like anyone else, can express an opinion that he knows to be false, 23 or that he makes in reckless disregard of its truth or falsity. See 31 U.S.C. § 3729(b)(1). We therefore hold that a false certification of medical 24 necessity can give rise to FCA liability. We also hold that a false certification 25 of medical necessity can be material because medical necessity is a statutory prerequisite to Medicare reimbursement. 26 27 28 1 Id.; see also United States v. Paulus, 894 F.3d 267, 275 (6th Cir. 2018) (doctor “opinions 2 are not, and have never been, completely insulated from scrutiny,” and “may trigger 3 liability for fraud when they are not honestly held by their maker”). 4 Nor does the Court agree with Defendants’ contention that a contrary medical 5 opinion is never enough, standing alone, to prove the falsity of a medical billing claim. 6 The Ninth Circuit stated in Winter that “an opinion can establish falsity.” 953 F.3d at 1120. 7 And the Third Circuit recently held, in response to AseraCare, that “a difference of medical 8 opinion is enough evidence to create a triable dispute of fact regarding FCA falsity.” Care 9 Alternatives, 952 F.3d at 99-100. Thus, in FCA cases as in others, the “reliability and 10 believability of expert testimony . . . is exclusively for the jury to decide.” Paulus, 894 11 F.3d at 277 (citations omitted); see also Wyler Summit P’ship v. Turner Broad. Sys., Inc., 12 235 F.3d 1184, 1192 (9th Cir. 2000) (“Weighing the credibility of conflicting expert 13 witness testimony is the province of the jury.”). The opposing opinions of Drs. Noyes and 14 Steinberg create an issue of fact that must be resolved at trial. 15 What is more, Scott does not rely solely on the testimony of Dr. Noyes. He also 16 presents evidence that a 2013 audit of Defendants’ medical practice – by auditors selected 17 by Defendants themselves – found that billing of complex simulations for each SBRT 18 fraction under CPT code 77290 was improper. Doc. 223-23 at 9. The audit cited numerous 19 examples. See Doc. 236-16 at 2, 6-7, 10-12 (stating that it is “not appropriate . . . to bill a 20 daily simulation for isocenter location, as image guidance is included in the SBRT delivery 21 codes.”) (under seal). The audit also cited instances of Defendants improperly billing CPT 22 code 77290 that corroborate Dr. Noyes’ opinion. See, e.g., id. at 4 (“there is no 23 documentation in the chart indicating that a complex simulation occurred on this date of 24 service”); id. at 5 (“there is no documentation to indicate that a complex simulation was 25 performed on this date of service. It would not be appropriate to bill a complex simulation 26 for the boost physics plan.”); id. at 8-9 (“This was billed incorrectly as a complex 27 simulation.”); id. at 9 (“There was no documentation to support this charge[.]”); id. at 11 28 (“There is no documentation for the verification simulation prior to treatment. Normally 1 this would be a simple simulation.”). And Defendants’ own expert testified that he 2 personally does not bill under CPT code 77290 when administering fractions of SBRT and 3 does not know of any other radiation oncologists who does. Doc. 236-10 at 13, 83. The 4 jury must resolve the factual issue presented by this evidence. 5 ii. Alleged Lack of Authority. 6 Scott also argues that CPT code 77290 cannot be billed to federal healthcare payers 7 in combination with CPT code 77373 (delivery of SBRT treatment). Doc. 47 at 14. Scott 8 asserts that a reasonable jury could find that while Defendants had a uniform policy of 9 billing CPT code 77290 with every fraction of SBRT, they did not perform the services 10 encompassed by that code. Doc. 237 at 10. 11 Defendants argue that Scott identifies no governing authority to support his claim 12 that Defendants improperly billed CPT code 77290, and that Dr. Noyes’s conclusion that 13 Defendants engaged in routine image guidance is nothing more than a difference of opinion 14 based upon an ex post review of the medical records. Docs. 222 at 10, 247 at 9. Defendants 15 specifically argued during oral argument that no law prohibits them from billing CPT 16 code 77290 for every SBRT fraction. But this argument assumes that a complex simulation 17 was actually performed with every fraction. Surely a doctor commits fraud by knowingly 18 billing the government for a complex simulation he did not perform, even if there is no 19 clear billing guidance for that procedure. Even without specific guidance, Defendants may 20 submit false claims under the FCA. 21 As noted above, Scott has also presents evidence beyond Dr. Noyes’s conclusion 22 that billing CPT code 77290 with every fraction was improper. AZC selected American 23 Medical Accounting & Consulting (“AMAC”) to conduct a billing audit of its radiation 24 oncology practice in 2013. Doc. 223-17 at 5. AMAC’s audit was based on AZC’s charts 25 and documents, and relied on guidance from a number of resources, including the CPT 26 Codebook, the World Health Organization’s International Classification of Diseases, the 27 Healthcare Common Procedural Coding system, the ASTRO/ACR Guide to Radiation 28 Oncology Coding, various Medicare LCD policies, and CCI Edits. Doc. 223-18 at 5. 1 The audit revealed that Defendants were improperly billing under CPT code 77290 2 with each fraction of SBRT. Doc. 223-23 at 9. The individual billing review produced by 3 AMAC as part of the audit cited numerous examples of improper billing similar to Dr. 4 Noyes’s cited examples. See Doc. 236-16 at 2, 4-9, 10, 11, 12. And as also noted above, 5 Dr. Steinberg testified that he personally does not bill under CPT code 77290 when 6 administering fractions of SBRT and does not know of any other radiation oncologists who 7 do. Doc. 236-10 at 13, 83.11 8 When construed in Scott’s favor, as required at the summary judgment stage, this 9 evidence could support a reasonable jury’s finding that Defendants’ billings under CPT 10 code 77290 were improper. Summary judgment will be denied with regard Scott’s claims 11 arising out of this code. 12 b. CPT Code 77470.12 13 Scott alleges that Defendants improperly billed CPT code 77470 on patients 14 since 2009. Doc. 47 at 22. Defendants argue that summary judgment is warranted because 15 Dr. Noyes failed to conduct a full review of AZC’s patient charts, and the only evidence 16 Scott can present is controverted by their expert, Dr. Steinberg. Doc. 222 at 11. Defendants 17 again rely on the Eleventh Circuit’s AseraCare decision, which the Ninth Circuit declined 18 to follow. Winter, 953 F.3d at 1117-19. And as explained above, the Court disagrees with 19 Defendants’ assertion that an issue of fact cannot be created by a contrary expert opinion. 20 Defendants argue that Dr. Noyes’s factual review was inadequate, but he asserts that 21 he reviewed the necessary records for his opinion. Doc. 236 ¶¶ 133-38. Scott also notes 22 that Drs. Noyes and Steinberg disagree on only a relatively small percentage of the CPT 23 code 77470 cases. Id. The conflicting evidence presented by these doctors creates a 24 question of fact that must be resolved by the jury. See Wyler, 235 F.3d at 1192 (“Weighing

25 11 Defendants assert that Scott agreed with their decision to continue billing CPT 26 code 77290 for every SBRT fraction after the audit (Doc. 247 at 13), but this assertion is disputed (Docs. 236 ¶ 85, 236-7 at 61-62), presenting a factual question for the jury. 27 12 CPT 77470 is a “special treatment procedure” code billed only when the physician 28 performs extra work during the course of treatment due to underlying comorbidities or special circumstances. Docs. 47 at 22, 223-8 at 8. 1 the credibility of conflicting expert witness testimony is the province of the jury.”); see 2 also Care Alternatives, 952 F.3d at 100 (“a difference of medical opinion is enough 3 evidence to create a triable dispute of fact regarding FCA falsity.”). 4 c. CPT Codes 77295 and 77301.13 5 Scott alleges that it was not medically necessary for Defendants to perform a second 6 CT scan while treating patients for breast or prostate cancer. Doc. 47 ¶¶ 139-61. 7 Defendants argue that Scott relies solely on Dr. Noyes’s opinions, which are based on an 8 ex post review of medical records. Doc. 222 at 12. Again citing AseraCare, Defendants 9 contend that without more, summary judgment is warranted. Id. 10 The disagreement between the experts precludes summary judgment. Dr. Steinberg 11 opined that in cases of breast cancer, conducting a second CT scan is the standard of care 12 and medically necessary. Doc. 223-10 at 13. He opined that in the case of prostate cancer, 13 second CT scans were justified by documented and significant changes in prostate volume. 14 Id. at 14. Dr. Noyes reached a different conclusion, finding that only one set of records for 15 both breast and prostate cancer properly established medical necessity, and that the 16 remainder were improperly billed. Doc. 223-16 at 7-8. As already explained, a jury must 17 resolve this dispute. See Wyler, 235 F.3d at 1192; Care Alternatives, 952 F.3d at 100. 18 3. Scienter. 19 To act with scienter in an FCA case, a defendant must act with knowledge that the 20 claim for payment is false, or with deliberate ignorance or reckless disregard of whether it 21 is false. § 3729(b). Congress adopted this definition to make “‘firm . . . its intention that 22 the act not punish honest mistakes or incorrect claims submitted through mere 23 negligence.’” United States ex rel. Hochman v. Nackman, 145 F.3d 1069, 1073 (9th Cir. 24 1998) (quoting S. Rep. No. 99–345 at 7 (1986)). 25 26 13 CPT code 77295 is defined as a “3-dimensional radiotherapy plan, including dose- volume histogram,” and is based on the patient anatomy defined by three-dimensional 27 reconstructions obtained from trans-sectional imaging devices such as CT or MRI scans. Doc. 223-10 at 12, 65. CPT code 77301 is defined as an “intensity modulated radiotherapy 28 plan, including dose-volume histograms for target and critical structure partial tolerance specifications.” Id. at 65. 1 Defendants turn again to the “objective falsity” argument rejected by the Ninth 2 Circuit in Winter, 953 F.3d at 1119. The Court need not say more on this incorrect 3 argument. 4 The AMAC audit put Defendants on notice, as of November 27, 2013, that their 5 billing practices under CPT code 77290 were improper. This evidence could support a jury 6 finding that Defendants acted with scienter when they failed to change their billing 7 practices for CPT code 77290. Reckless disregard suffices for scienter and applies to “the 8 ostrich type situation where an individual has buried his head in the sand and failed to make 9 simple inquiries which would alert him that false claims are being submitted.” United 10 States v. United Healthcare Ins., 848 F.3d 1161, 1176 (9th Cir. 2016) (citations omitted). 11 Construed in Scott’s favor, the evidence creates a question of fact on whether Defendants 12 acted with scienter. 13 Defendants argue that Scott cannot present evidence that they did not provide the 14 services billed for under CPT codes 77470, 77295, and 77301. Doc. 222 at 14. But the 15 experts clearly disagree on this issue, creating a question of fact that precludes summary 16 judgment. See Wyler, 235 F.3d at 1192; Care Alternatives, 952 F.3d at 100. 17 4. Overbreadth of CPT Code 77290. 18 Defendants contend that Scott can present no evidence that they had reason to doubt 19 the propriety of billing CPT code 77290 before receiving the AMAC report on 20 November 27, 2013. Doc. 222 at 15. The Court agrees. Scott has failed to present any 21 evidence that Defendants knew or had reason to know of alleged impropriety in their 22 billings under CPT code 77290 before receiving the final AMAC report.14 23

24 14 During oral argument, Scott asserted that Dr. Biggs’s April 15, 2013 email is 25 evidence of a “guilty mind.” Scott notes in his statement of facts that Dr. Biggs’s email said the AMAC audit would be a “major pain” that “cannot turn into a witch hunt and/or 26 bashing session,” and that he favored “cancellation.” See Docs. 236 ¶ 156. But the email also notes that “there are better things to do with [$]11K.” Doc. 236-14 at 3. This 27 imprecise evidence would not support a jury finding that Defendants knew they were billing improperly under CPT code 77290 before November 27, 2013. Nor could such a 28 finding be based on Defendants’ general duty to be familiar with billing codes or the draft audit report that was subject to review and revision. 1 Scott’s reliance on the FCA’s reverse false claims provision is not convincing. Scott 2 alleges that Defendants knowingly avoided an obligation to refund overpayments in 3 violation of § 3729(a)(1)(G). But the Court previously dismissed claims arising out of this 4 argument, finding that Scott made no allegation that Defendants committed fraud with 5 respect to any credits. Doc. 76. 6 Because Scott has presented no evidence that Defendants acted with scienter in 7 billings under CPT code 77290 before November 27, 2013, the Court will grant summary 8 judgment with respect to those billings. Celotex, 477 U.S. at 322 (Summary judgment is 9 appropriate against a party who “fails to make a showing sufficient to establish the 10 existence of an element essential to that party’s case, and on which that party will bear the 11 burden of proof at trial.”). 12 5. Mitigation of Damages. 13 Scott alleges that Defendants fired him in retaliation for bringing this action. 14 Doc. 47 ¶¶ 317-23. AZC terminated Scott on March 29, 2018, asserting that he failed to 15 perform his responsibilities as billing manager. Doc. 223-27 at 2. Scott obtained another 16 billing position in July 2018, but quit two weeks later and took a lower paying job. 17 Doc. 223-4 at 28-30. Claiming that Scott failed to mitigate his damages, Defendants seek 18 summary judgment on damages for his retaliation claim. Doc. 222 at 15. 19 A discharged employee has “a duty to mitigate damages by seeking alternative 20 employment with ‘reasonable diligence.’” Caudle v. Bristow Optical Co., 224 F.3d 1014, 21 1020 (9th Cir. 2000). Defendants argue Scott has not, since July 2018, actively searched 22 for a job that would pay a market rate for his experience. Docs. 222 at 16-17, 247 at 19. 23 In response, Scott presents evidence that he applied for nearly 100 jobs and engaged in 20 24 interviews before receiving a single job offer at a lower salary than he received in the past. 25 Doc. 236-22 at 5-6. Defendants imply that Scott took a lower paying job to inflate his 26 mitigation damages, but Scott presents evidence that he took the lower paying job to obtain 27 health insurance. Doc. 223-4 at 30, 32. Scott’s expert opines that he engaged in reasonable 28 1 mitigation efforts. Id. at 8-9. Scott has raised a genuine is of fact that precludes summary 2 judgment on his mitigation efforts. 3 D. Scott’s Cross-Motion for Summary Judgment. 4 AZC alleges that Scott breached his fiduciary duties by committing serious 5 misconduct in bad faith and contrary to the best interests of AZC. Doc. 122 at 63. AZC 6 asserts three theories of liability against Scott: (1) he improperly billed for SRT with CPT 7 code 77373 instead of Healthcare Common Procedure Coding System (“HCPCS”) code 8 G0340, (2) he improperly wrote off patient account balances without obtaining approval 9 from physicians, and (3) he failed to reimburse overpayments to insurance company 10 payors. Id. 51-61. Scott seeks summary judgment on all three theories. 11 1. First Theory – Statute of Limitations.15 12 Scott argues that AZC’s first theory relating to the alleged failure to bill SRT using 13 HCPCS code G0340 is barred by the statute of limitations. Doc. 237 at 21. He contends 14 that Defendants were on notice as of November 27, 2013 – the date of the AMAC audit 15 report – that they could bill for SRT under a G code, and that they knew as early as October 16 2010 that he instead was billing under CPT code 77373. Doc. 237 at 22. 17 Arizona law governs AZC’s claim against Scott. See United Mine Workers of Am. 18 v. Gibbs, 383 U.S. 715, 726 (1966). Actions for breach of fiduciary duty must commence 19 within two years of the action’s accrual. A.R.S. § 12-542; CDT, Inc. v. Addison, Roberts 20 & Ludwig, C.P.A., 7 P.3d 979, 981 (Ariz. Ct. App. 2000). “The limitations period begins 21 to run when the plaintiff discovers the cause of action – that is, when the plaintiff knows, 22 or reasonably should know, that he has been harmed, that the harm was caused by the 23 defendant, and that the act or omission which caused the harm was wrongful.” Wichansky 24 v. Zowine, 150 F. Supp. 3d 1055, 1062 (D. Ariz. 2015). 25 Summary judgment is precluded by factual issues on when AZC knew or should 26 have known that Scott’s failure to bill SRT procedures under the G code harmed it and was

27 15 AZC claims that Scott improperly billed for SRT with CPT code 77373 instead 28 of HCPCS code G0340. Doc. 122 at 58. AZC contends that it consequently lost over $2 million from commercial payors. Doc. 248 ¶ 209. 1 wrongful. Id. In Walk v. Ring, 44 P.3d 990 (2002), the Arizona Supreme Court held that, 2 in determining whether the plaintiff was on notice to investigate, the question is whether a 3 plaintiff’s “failure to go forward and investigate [his possible cause of action] is not 4 reasonably justified.” Id. at 996. For example, a plaintiff would be reasonably justified in 5 declining to investigate a claim against a possible defendant if the plaintiff “subjectively 6 believed” that the defendant had done nothing wrong. Id. (citation omitted). The question 7 is whether a reasonable person in the plaintiff's position would investigate the claim. Id. 8 Even if AZC knew, prior to the limitations period, that Scott was not billing the 9 G code for SRT procedures, the claim for breach of fiduciary duty would not accrue until 10 AZC also knew or had reason to know that it could bill under that code and that Scott’s 11 failure to do so was wrongful. AZC presents evidence that Scott advised it to disregard the 12 conclusion in the AMAC report regarding billings under the G code, said the auditors had 13 misunderstood the AZC equipment, advised it that it could not bill SRT procedures under 14 the G code, and that it relied on Scott and trusted his advice as its longtime billing manager. 15 See Doc. 248 ¶¶ 215-20. Scott does not address this evidence in his reply, and it clearly 16 creates a question of fact as to when AZC’s cause of action accrued. Doc. 251 at 2-5. 17 Walk also notes that the accrual of a claim must account for any fiduciary duty that 18 existed between the plaintiff and the defendant. AZC asserts that Scott owed it fiduciary 19 duties. Walk states that “‘if the fiduciary nature of the relationship charges the fiduciary 20 with a duty to disclose his wrong to the plaintiff and he fails to disclose, the statute of 21 limitations will be tolled.’” 44 P.3d at 1000 (quoting Bourassa v. LaFortune, 711 F. Supp. 22 43, 46 (D. Mass. 1989)). Whether Scott breached his fiduciary duty to AZC in this manner 23 is not addressed by the parties and is for the jury to decide. 24 2. Second and Third Theories – Damages. 25 Scott argues that AZC cannot prove damages on its second theory that he improperly 26 wrote off account balances and on its third theory that he failed to reimburse overpayments 27 to insurance company payors. Doc. 237 at 23; see Doc. 122 at 51-61. AZC responds that 28 it does not seek to recover lost revenue, but instead seeks to recover $546,348.80 in 1 compensation and benefits paid to Scott. Doc. 247 at 23, 26. AZC relies on the “faithless 2 servant doctrine,” under which an employee who violates his fiduciary duties forfeits the 3 right to compensation during the period of faithfulness. Doc. 247 at 24. Under this 4 doctrine, “an agent is entitled to no compensation for conduct which is disobedient or 5 which is a breach of his duty of loyalty.” Restatement (Second) of Agency § 469 (1958); 6 see Johnson v. Pac. Lighting Land Co., 817 F.2d 601, 607 (9th Cir. 1987) (“In the absence 7 of contrary authority, Arizona courts follow the Restatement as the proper statement of 8 law.”) (citation omitted). 9 Scott’s initial cross-motion did not anticipate AZC’s assertion of the faithless 10 servant doctrine. Scott argued that AZC cannot prove and did not disclose damages 11 incurred as a result of his alleged improper write-offs or his alleged failure to reimburse 12 insurance overpayments. But this basis for summary judgment is rendered moot by AZC’s 13 disavowal of such damages. 14 The question, then, is whether Scott has provided a basis for summary judgment on 15 the compensation AZC seeks to recoup under the faithless servant doctrine. Scott does not 16 argue that the doctrine is unavailable under Arizona law or under AZC’s claim for breach 17 of fiduciary duty. Nor does Scott assert that AZC will be unable to present evidence that 18 he breached his fiduciary duties. Instead, Scott’s sole argument in his reply brief continues 19 to be that AZC cannot prove that it suffered losses “caused” by his allegedly wrongful 20 action. But this argument misunderstands the faithless servant doctrine. Compensation 21 under that doctrine is not recouped by an employer because it approximates the financial 22 harm “caused” by the employee’s actions. Compensation is recouped because it was paid 23 in vain – it was paid for faithful service that the employer did not receive. The employee’s 24 faithless service causes the loss. As one court has explained, “dishonesty and disloyalty 25 on the part of an employee which permeates his service to his employer will deprive him 26 of his entire agreed compensation, due to the failure of such an employee to give the 27 28 1 stipulated consideration for the agreed compensation.” Foley v. Am. Elec. Power, 425 F. 2 Supp. 2d 863, 863 (S.D. Ohio 2006).16 3 AZC claims that it should recover the compensation paid to Scott because it never 4 received the faithful service for which he was paid. If AZC proves what it alleges, Scott’s 5 faithlessness will supply the necessary causation. Thus, Scott has provided no basis for 6 granting summary judgment on AZC’s breach of fiduciary duty claim. 7 III. Other Matters. 8 A. Government’s Statement of Interest. 9 The government has not intervened, but seeks leave to file a statement of interest in 10 response to Defendants’ motion for partial summary judgment. Doc. 231. The parties do 11 not object. Id. at 1. “Pursuant to 28 U.S.C. § 517, the United States may submit a statement 12 in a case expressing its views on relevant issues in which it has an interest.” Wortman v. 13 All Nippon Airways, 854 F.3d 606, 617 (9th Cir. 2017). The Court will grant the motion 14 and consider the government’s statement of interest.17 15 B. Scott’s Motion to Seal. 16 Scott moves to seal portions of his cross-motion for summary judgment, statement 17 of facts, and response to Defendants’ summary judgment motion. Doc. 234. Scott has 18 lodged the proposed sealed versions of the documents with the Court and has filed redacted 19 public versions on the docket. See LRCiv 5.6(b)-(c). 20 The filings contain confidential business information related to an audit of 21 Defendants’ radiation oncology billing practices. The Court’s protective order defines 22 23 16 Scott’s reply also asserts that his alleged breaches of fiduciary duty are unlike those addressed in the faithless servant cases cited by AZC, and that employees who make 24 mistakes should not be penalized with the loss of their full compensation. Doc. 251 at 6-8. But he does not explain how these arguments entitle him to summary judgment, nor did he 25 assert them in his cross-motion. 26 17 Defendants request an opportunity to respond to the government’s statement, asserting that it contains “misleading and selective quotations.” Doc. 247 at 3 n.2. Because 27 the government’s statement makes many of the same arguments as Scott’s briefing, to which Defendants have fully responded, and the Court is able to evaluate the cases cited 28 by the parties, an additional response is not necessary. The outcome of this order has not been affected by the government’s statement. 1 confidential business information as “any trade secret or other confidential, strategic, 2 research, development, financial, or commercial information that, if disclosed, could 3 detrimentally affect the party’s business, commercial or financial interest.” Doc. 38. 4 Revealing information about Defendants’ billing practices and the results of the audit could 5 detrimentally affect AZC’s business or financial interests. Sealing the motion and 6 statement of facts will have little effect on the public’s ability to understand the issues 7 addressed in this order because lightly redacted copies have been filed in the public docket. 8 The Court finds compelling reasons to seal and will grant the motion. See Kamakana v. 9 City & Cty. of Honolulu, 447 F.3d 1172, 1179 (9th Cir. 2006). This order has cited to 10 publicly filed briefs and exhibits where possible, and has identified citations to sealed 11 documents. 12 IT IS ORDERED: 13 1. Defendants’ motions to exclude the opinions of Dr. Abraham Wyner 14 (Doc. 225) and Dr. William Noyes (Doc. 246) are denied. 15 2. Defendants’ motion for partial summary judgment (Doc. 222) is granted 16 with respect to billings under CPT code 77290 before November 27, 2013, 17 and otherwise is denied. 18 3. Scott’s cross-motion for summary judgment (Docs. 233, 237) is denied. 19 4. The government’s motion for leave to file a statement of interest (Doc. 231) 20 is granted. 21 5. Scott’s motion to file documents under seal (Doc. 234) is granted. The Clerk 22 of Court shall accept for filing under seal the documents lodged on the 23 Court’s docket as Docs. 235, 236. 24 6. The Court will hold a telephonic hearing on May 13, 2020 at 3:30 p.m. to 25 schedule a final pretrial conference and trial in this matter. Counsel who will 26 be trying the case shall participate in the telephonic hearing. Plaintiff shall 27 initiate a conference call to include counsel for all parties and the Court. 28 1 Plaintiff shall provide the dial in information to counsel for all parties and 2 the court no later than May 11, 2020 at 12:00 noon. 3 Dated this 29th day of April, 2020. 4

° David G. Campbell 7 Senior United States District Judge 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

-32-