Dispute Centers on Whether Certain Medical Procedures Were Performed, Not Their Necessity; United States District Court for the District of Arizona No. CV-16-03703-PHX-DGC, U.S. ex rel. J. Scott v. Arizona Center for Hematology and Oncology PLC


The district court denied the defendants’ motion for summary judgment in a case alleging healthcare fraud. The defendants first challenged various aspects of the plaintiff’s use of statistical sampling to allege the submission of false claims, but the court found the defendants did not challenge the expertise of the plaintiff’s statistician or his methodology. The defendants also argued the plaintiff’s allegations amounted to mere disagreement over medical diagnoses, which they asserted could not form the basis of an FCA complaint. However, the court noted that the plaintiff did not assert that the defendants billed for procedures that were not necessary, but billed for work that was never done.

In these proceedings, the court considered multiple cross-motions from the plaintiff and defendants in a qui tam action alleging healthcare fraud. Relator J. Scott and the defendants— Arizona Center for Hematology and Oncology PLC dba Arizona Center for Cancer Care, and Drs. Terry Lee, Daniel Reed, and Christopher Biggs—filed cross motions for summary judgment. The defendants also moved to exclude expert testimony proffered by the relator and filed counterclaims against Scott for breach of fiduciary duty. The government moved for leave to file a statement of interest and Scott moved to file certain documents under seal.

First, the court considered the defendants’ motion to exclude the opinions of Dr. Abraham Wyner, Scott’s statistical sampling expert, and Dr. William Noyes, Scott’s medical billing expert. Wyner used statistical sampling to estimate alleged healthcare overpayments received by AZC. The defendants argued that (1) he is not qualified to estimate healthcare overpayments, (2) his sampling methodology is not reproducible, (3) his samples are not representative, and (4) his reports are replete with errors.

Specifically, the defendants argued that while Wyner had academic experience in statistics, he had no experience estimating alleged healthcare overpayments and no knowledge of relevant government guidelines. The court found Wyner’s credentials and ability to conduct a statistical analysis were not in dispute. The court also noted that government guidance do not require a statistician to have expertise in healthcare billing prior to conducting an analysis. Rather, the guidance says that statistical sampling of possible healthcare overpayments should be overseen by a statistician or someone with equal expertise. The court concluded that Wyner met the relevant educational and background qualifications suggested by the guidance.

Next, the defendants argued that they could not replicate Wyner’s studies, because he failed to disclose his exact methodology. Wyner explained that he used industry software to conduct his analysis and that the specific results of a random number generator could not be extracted from the program. The court also noted the defendants did not suggest the software or methodology was invalid or that the samples chosen were not random. Therefore, the court could not find the analysis was unreliable. The defendants also argued that Wyner failed to disclose the database from which his sample was drawn, but the court found the data was taken from the defendants’ own spreadsheets of claims for payment. While the lack of some supporting information may make the analysis more difficult to defend during cross, the court did not find this invalidated the analysis.

Next, the defendants argued that Wyner’s samples were not representative, which resulted in a larger overpayment estimate. In response, Wyner provided a detailed analysis of how the defendants’ expert misinterpreted his analysis and misapplied certain tests. Because the defendants did not reply to Wyner’s response, the court could not conclude that the analysis was unreliable.

Finally, the defendants argued that Wyner’ reports were filled with errors, such as the exact number of claims in contention. However, the court reasoned that if Wyner were qualified to make his analysis and if his methodology was sound, then his opinions were sufficiently reliable to be admitted and the defendants may challenge the correctness of his conclusions at trial.

Next, the defendants moved to exclude the opinions of Dr. William Noyes, Scott’s medical billing expert, arguing that he is not qualified to render opinions about current procedural terminology as it relates to the delivery of stereotactic radiosurgery and stereotactic body radiation therapy. They argued Noyes lacks current practical expertise with these procedures, which have undergone sweeping changes in the last decade, and lacked knowledge of billing guidelines for these procedures. They also argued that each patient and therapy session is unique, and complex judgments are made during treatment.

However, the court noted that Noyes’ opinions were not based on an evaluation of treatment choices, but on the defendants’ medical records. Specifically, Noyes’ testimony is that the medical records do not reflect any evidence the defendants performed the complex procedures for which they defendants billed the government. Scott’s complaint alleged that the defendants billed for certain procedures regardless of whether they were performed, and the court reasoned that Noyes’ opinion supported this contention. Thus, the court found that Noyes’s opinion was not based on second-guessing complex medical decisions, but on the fact that the defendants consistently billed for procedures they did not perform. Further, the court found that Noyes performs similar procedures regularly and was therefore qualified to say whether they were reflected in the defendants’ medical records.

Next, the court considered the motions for summary judgment.

Scott previously worked for AZC as the billing manager for its radiation oncology department, where he was responsible for overseeing day-to-day billing operations, including properly analyzing claims for accuracy and completeness, serving as AZC’s expert on coding and billing processes, ensuring that AZC’s billing operations were conducted in a manner consistent with payor rules and guidelines, and keeping up-to-date with current coding, billing, and compliance requirements.

The court previously dismissed two of Scott’s claims, leaving three allegations in contention. Scott alleged that: (1) the defendants falsely billed under an incorrect billing code; (2) the defendants improperly billed for special procedures they did not perform; and (3) the defendants filed claims for medically unnecessary procedures. Scott also alleged that he was fired in retaliation for asserting these claims.

In their motion for summary judgment, the defendants first argued that statistical sampling cannot be used to establish liability under the FCA. More specifically, they argued that because Scott’s claims involve subjective, fast-specific, clinical determinations by treating physicians, liability cannot be established through statistical sampling.

The court disagreed, finding that previous cases cited did not reject statistical evidence, but left it to the court to determine if the evidence is reliable. Contrary to the defendants’ assertion, many courts have allowed statistical sampling in FCA cases.

The defendants also argued that the use of statistical sampling would violate their right to due process of law. The appeared to argue that if Scott is permitted to prove that they made a false claim on a particular occasion without presenting direct evidence of that false claim, they will be denied due process. The court disagreed, first noting that neither the FCA nor Constitution suggest that a preponderance of the evidence cannot be provided by reliable representative evidence. Second, the court noted that the nature of the preponderance of the evidence standard allows for probabilities, which can be proved by a variety of means, including indirect evidence.

The defendants also argued that statistical sampling is inappropriate because each patient and treatment plan is different, and generalization through statistics cannot account for many individual variables that may affect whether a certain procedure is warranted. But again, the court explained that the case did not allege bad treatment decisions, but the submission of bills for procedures that never occurred. The court also noted that the Supreme Court reasoned that complex cases may result in statistical sampling being the only practicable means to collect and present relevant data to establish a defendant’s liability. In this case, Scott alleged the submission of at least 4,000 false claims. According to the court, no reasonable trial could include individualized proof of 4,000 separate occurrences of fraud.

Next, the defendants argued that Scott had no evidence they submitted false claims, because he could not show their claims were objectively false. The defendants argued that Noyes’ observations were negated by an analysis by their own expert witness, who found that the defendants’ medical records did show that certain procedures were performed. The defendants argued that Scott could not bring his claims based on a difference of medical opinion.

The court disagreed, noting that a doctor’s clinical opinion can be false, if the doctor knows it to be false when made, or if the doctor rendered the opinion with reckless disregard for the truth. For example, a false certification of medical necessity could give rise to FCA liability. The court also disagreed with the contention that a contrary medical opinion is never enough to prove falsity. Rather, the reliability and believability of expert testimony is a judgment for the jury, and therefore the opinions of the expert witnesses create an issue of fact that must be resolved at trial.

Further, the court noted that Scott did not rely solely on Noyes’ testimony but presented evidence that a previous audit of the defendants’ billing practices found evidence of improper billing for the procedures at issue, including a lack of documentation that certain services were performed. The court also noted that a physician surely commits fraud by knowingly billing the government for a complex simulation he did not perform, even if there is no clear billing guidance for that procedure.

Next, the defendants argued Scott had no evidence of scienter, again making their argument regarding objective falsity. The court again rejected this argument as incorrect. Further, the court noted that Scott’s evidence of the prior billing audit put the defendants on notice of the conduct at issue. The court reasoned this evidence could support a finding that the defendants acted with scienter when they failed to change their billing practices.

Next, the defendants argued that Scott can presented no evidence that they had reason to doubt the propriety of their billing practices before they received that audit. The court agreed, finding Scott provided no evidence the defendants were aware their billings were improper. The court therefore granted summary judgment with respect to those billings.

Finally, the defendants sought summary judgment on Scott’s claims of retaliation. AZC terminated Scott on March 29, 2018, asserting that he failed to perform his responsibilities as billing manager. Scott obtained another billing position in July 2018, but quit two weeks later and took a lower paying job. Claiming that Scott failed to actively search for a job that would pay commensurate with his experience and therefore mitigate his damages, the defendants seek summary judgment on damages for his retaliation claim.

In response, Scott presented evidence that he applied for nearly 100 jobs and engaged in 20 interviews before receiving a single job offer at a lower salary than he received in the past. Scott also explained he accepted the lower paying job to obtain health insurance. The court found Scott raised a genuine issue of fact that precluded summary judgment.

Next, the court considered Scott’s motion for summary judgment on AZC’s claims that he breached his fiduciary duties by committing serious misconduct in bad faith and contrary to the best interests of AZC. AZC alleged Scott created improper billings; improperly wrote off patient account balances without approval; and failed to reimburse overpayments to insurance company payors.

In his motion, Scott argued that the alleged improper billing is barred by the statute of limitations. Scott argued that the audit report cited above put the defendants on notice that he billed for procedures using certain billing codes. The court noted that Arizona law provides a 2-year statute of limitations on this claim. However, factual issues remained regarding when AZC knew or should have known the basis for the claim. AZC presented evidence that Scott advised it to disregard the conclusion in the AMAC report regarding certain billings, and that it relied on Scott and trusted his advice as its longtime billing manager. The court found enough facts were in dispute to preclude summary judgment.

Scott also argued that AZC cannot prove damages on their other claims. In response, AZC explained that it did not seek to recover lost revenue, but to recover $546,348.80 in compensation and benefits paid to Scott. The court found the basis for summary judgment mooted by the defendants’ disavowal of these damages. The court allowed the claim to proceed on a faithless servant doctrine.

Finally, the court addressed procedural matters. First, the court granted the government’s motion to file a statement of interest, noting that neither party objected. Next, the court granted Scott’s motion to seal portions of his cross-motion for summary judgment, statement of facts, and response to the defendants’ summary judgment motion. The court agreed, noting that the documents contain confidential business information related to an audit of the defendants’ radiation oncology billing practices which, if disclosed, could detrimentally affect the party’s business, commercial or financial interest. The court reasoned that sealing the motion and statement of facts will have little effect on the public’s ability to understand the issues addressed in this order because lightly redacted copies have been filed in the public docket.

FCA Scott v Arizona Center