Bryson pleaded guilty to defrauding Medicare. As part of his sentence, the District Court ordered him to pay restitution of $30,338.90 to the Government. Bryson challenges the District Court's method of calculating the restitution amount. We affirm.

* For several years, Bryson, a podiatrist, submitted fraudulent reimbursement claims to Medicare. On multiple occasions, he represented to Medicare that he had performed a reimbursable medical procedure; in fact, he had performed only routine, non-reimbursable services such as clipping toenails, removing corns, and applying topical medication. On August 17, 2005, Bryson pleaded guilty to one count of making false statements to Medicare. See 18 U.S.C. §§ 2, 1035. As part of the plea, he admitted to perpetrating a fraudulent reimbursement scheme from 1998 until 2001, and he identified seven patients for whom he submitted false information to Medicare. At the sentencing hearing, the District Court heard extensive testimony on the amount of loss caused by Bryson's scheme. The court sentenced Bryson to three years of probation and ordered him to pay restitution of $30,338.90 to the Medicare Fraud Unit. On appeal, Bryson challenges that restitution figure.

In its initial sentencing submission to the District Court, the Government proposed a restitution figure of $37,417.34. The Government explained that this amount was calculated based on four investigative sources: (1) the false claims filed by Bryson on behalf of the seven patients named in the plea documents; (2) a review of Bryson's files by an expert podiatrist, Paul Kinberg, which revealed information that was inconsistent with Bryson's representations to Medicare; (3) a tally of claims submitted to Medicare for which Bryson had insufficient, incomplete, or inconsistent notes, despite the Medicare requirement that physicians document treatment for which they seek reimbursement; and (4) surveys and interviews of former patients about what kind of medical treatment they received from Bryson. After its initial submission, the Government slightly lowered its figure to $37,132.61, explaining that the initial calculation had included restitution for conduct that occurred outside of the time period covered by Bryson's guilty plea. The Government also took pains to ensure that a patient included in multiple components was not double-counted; for example, a patient who indicated on a survey that he had received no treatment and who also was not mentioned in any notes in the patient's file was counted only once.

Bryson argued that the Government substantially overstated the amount of restitution. In his view, the proper figure was $20,050.40. This was allegedly an estimate that a prior prosecutor had suggested to Bryson at some earlier date.

During Bryson's sentencing hearing, Bryson and the Government called witnesses to justify their respective restitution figures. Bryson called two former patients, both in their 80s, to show that at least some of the interviews and surveys cited by the Government had not occurred. The first witness, Martha Feggins, testified that she had never completed a survey or been interviewed by government agents (although she repeatedly became confused on the witness stand and gave contradictory answers at different times during her testimony). The second witness, Hope Settles, testified that government agents had not interviewed her son (also one of Bryson's patients), as the Government had claimed.

The Government called Jason Marrero, a special agent with the Department of Health and Human Services who had investigated Bryson. Agent Marrero presented documentation showing that he had interviewed both Feggins and Settles's son in connection with his investigation. Agent Marrero also detailed the steps he had taken in his investigation of Bryson, which included reviewing Bryson's medical records, employing podiatrists to examine both Bryson's records and the physical condition of his patients, interviewing Bryson's patients, and sending surveys to Bryson's patients.

Agent Marrero explained the method he used for calculating the loss to Medicare attributable to Bryson's fraud. He clarified that the four-component approach initially described by the Government was in fact a three-component approach: The first component (fraudulent claims in connection with the seven patients named in the plea documents) was actually just a subset of the fourth component (interviews and surveys of all patients). Thus, the three components included in Marrero's final figure were: (1) Kinberg's review of Bryson's files; (2) the government review of Bryson's "no notes" files; and (3) patient surveys and interviews. During his testimony, Marrero re-calculated the appropriate restitution amount for the period from January 1998 to September 2001 and reached the same figure that the Government had stated in its revised submission (apart from a $100 reduction caused by a double-counting error). Agent Marrero also testified that the loss attributable to Bryson's entire scheme through April 2002 was $48,285.39.

The District Court then computed a final restitution amount. First, the court concluded that the time period for which Bryson should pay restitution was January 1998 to April 2002, even though Bryson had pleaded guilty to submitting fraudulent claims only through September 2001. The court explained that the Guidelines require the sentencing judge to consider all relevant conduct, and it was therefore appropriate to award restitution for Bryson's conduct "before and after [the] time period [pled to] and [which] is related and connected to the scheme that was pled to." Tr. of Dec. 15, 2005, Hr'g at 160. As Agent Marrero had testified, the restitution over this period was $48,285.39. Second, the court concluded that although the interviews of patients were reliable, the surveys "pose problems in that they do not provide the . . . same opportunity for the face-to-face discussion between the patient and the agent." Id. at 164. The court therefore subtracted the amount attributable only to the surveys ($15,040.70) from the total. Id. at 165-66. Third, the court excluded the restitution amount attributable to Bryson's false claims on behalf of Feggins ($2,905.79), because her "memory faults" threw into question the accuracy of her statements to investigators. Id. at 164, 167. The final figure was $30,338.90. Id. at 167.

The Government must prove at sentencing that its proposed restitution figure is supported by a preponderance of the evidence. 18 U.S.C. § 3664(e). We review the District Court's restitution determination under an abuse of discretion standard. United States v. Rezaq, 134 F.3d 1121, 1141 (D.C.Cir.1998). We examine the factual findings underpinning the restitution award only for clear error. United States v. Lawrence, 189 F.3d 838, 846 (9th Cir.1999); see also United States v. Leonzo, 50 F.3d 1086, 1088 (D.C.Cir. 1995).

Bryson raises five objections to the restitution calculation.

First, Bryson argues that it was improper to rely on Agent Marrero's analysis because he is "not a medical professional." Appellant's Br. at 19. At the time of his testimony, Agent Marrero had seven years of experience as an HHS special agent and had received training in Medicare regulations and billing. See Tr. of Dec. 6, 2005, Hr'g at 33. Moreover, Agent Marrero consulted expert podiatrists such as Kinberg throughout his investigation. It would be odd, to say the least, if a district court could not rely on testimony of a trained Medicare fraud investigator as the court determined the magnitude of a Medicare fraud.

Second, Bryson claims that the District Court erred by relying on a restitution figure submitted by the Government, because the District Court was presented with "a number of possibilities" during the sentencing hearings. Appellant's Br. at 14-15. Bryson cites United States v. George, 403 F.3d 470 (7th Cir.2005), in which the Seventh Circuit invalidated a restitution award when different witnesses had submitted widely varying estimates of restitution to the District Court. See id. at 473-74. But Bryson's reliance on George is mistaken. The problem in George was not the wide range of estimates itself, but rather the fact that the sentencing judge selected a seemingly arbitrary figure within that range without "mak[ing] findings of fact or articulat[ing] his reasons." Id. at 473. By contrast, here the District Court carefully explained its analysis, making an effort to ensure accuracy by excluding the amount attributable both to the surveys and to the Feggins interview.

Third, Bryson contends that the District Court unreasonably relied on interviews of Feggins and Settles's son, even though Bryson presented testimony that they were never interviewed. With respect to Feggins, the District Court in fact excluded from the final restitution award everything attributable to her interview. And with respect to Settles's son, the District Court reasonably credited the testimony and documentation of Agent Marrero and another agent (Agent Tucker) indicating that they had interviewed him.

Fourth, Bryson argues that the rest of the surveys and interviews the Government cited were unreliable because of the faulty memories of elderly patients and the technical nature of the questions asked. The District Court shared some of these concerns — which is precisely why it excluded all the survey results from the final calculation. As the District Court put it, a survey "doesn't provide an opportunity for the patient to ask a question" or "for the agent to see that there may be some hesitation on the part of the patient in regard to the proper answer to a question." Tr. of Dec. 15, 2005, Hr'g at 164. The District Court concluded, however, that the same problems generally do not exist with the interviews, during which there is face-to-face interaction between the agents and the patients. Id. The District Court's assessment was eminently reasonable, and we will not disturb it.

Fifth, Bryson asserts that the District Court erred by relying on the lack of notes in a patient's file as evidence of Medicare fraud. We are not convinced. Given Bryson's guilty plea to a scheme of Medicare fraud, the overwhelming evidence against him, and the documentation requirements for Medicare reimbursements, it was logical for the District Court to conclude that the lack of any documentation in a patient's file for a procedure indicates that Bryson did not perform the procedure.

Finally, Bryson has not argued that the restitution amount could have been incorrect because it was based in part on conduct that occurred outside the period covered by his guilty plea. We therefore do not consider that issue. See Cruz v. Am. Airlines, Inc., 356 F.3d 320, 333-34 (D.C.Cir.2004).

We affirm the judgment of the District Court.

So ordered.