Medicare & Medicaid Guide P 45,901 United States of America v. Samarjeet S. Sidhu Loren Arden Gifford, Medical Doctor

DeMOSS, Circuit Judge: Appellants Dr. Loring Gifford 1 and his employee Samarjeet Sidhu were each convicted of conspiracy to commit mail fraud and other criminal conduct relating to their practice of submitting false claims for medical services to various private insurers and government programs. Sidhu appeals both his convictions and his sentence. Gifford appeals only the district court’s determination of his sentence. We affirm. BACKGROUND Appellant Loring Gifford is a psychiatrist. Prior to his conviction in this case, Gifford operated a profitable practice in El Paso, Texas. Gifford advertised as a specialist in the area of addiction and pain management, and frequently prescribed morphine. Indeed, Gifford prescribed a significant portion of the morphine prescribed in the State of Texas. Many of Gifford’s patients came in daily to receive injections of morphine, alone or in combination with other pain-killing injections. Some of Gifford’s patients became addicted to morphine as a result of his treatments, and required further treatment to withdraw from the medications Gifford provided. Gifford systematically defrauded government programs, such as Medicare, Medicaid, and CHAMPUS, and private insurers. Specifically, government programs and private insurers were billed for services that were either (1) not performed, (2) not performed as billed, or (3) performed by non-physicians, for whose services Gifford was not* entitled to be reimbursed. In but one example, Gifford routinely billed sixty-minute psychotherapy sessions to patients who came to the office for injections. Gifford billed these sessions using a code that contemplates face-to-face psychotherapy. Nonetheless, patients were billed without regard to whether the patient was seen by Dr. Gifford. In several eases, patients were billed for psychotherapy sessions notwithstanding the fact that Gifford was out of town, or even out of the country. Eventually, patients were billed without regard to whether the patients came to the office. In one case, Gifford billed psychotherapy to a patient who was no longer living. Gifford also billed far more psychotherapy than could have reasonably been performed. For example, Gifford personally billed more than 5,800 hours as patient visits for the year 1993. Gifford routinely billed between forty and sixty psychotherapy appointments in a single day, all supposedly lasting between thirty and sixty minutes per session. Gifford does not dispute that he engaged in these fraudulent billing practices. Appellant Samarjeet Sidhu worked for Gif-ford. Sidhu, who was trained as a physician in Mexico but failed the Texas medical exams, performed biofeedback services on Gif-ford’s patients. Sidhu billed for biofeedback using a code that contemplates the measurement and regulation of body temperature, which is generally accomplished with the aid *648 of a biofeedbaek machine. Several patients testified that they never saw Sidhu’s biofeedback machine, and that Sidhu generally just talked to the patient, performing more of a counseling role. Sidhu ran the office in Gifford’s absence. Gifford’s staff were instructed to call Si-dhu “Dr. Sidhu.” Patients were referred to Sidhu in Gifford’s absence, and Sidhu was assigned the task of judging whether the patients needed medication, counseling, biofeedback or all three. Many of the services performed by Sidhu in Gifford’s absence were billed as psychotherapy sessions with Gifford. Sidhu received computer print-outs detailing Gifford’s billings, and assisted Gifford’s efforts to collect on the fraudulent billings by meeting and corresponding with patients and insurers. Office staff testified that Sidhu was involved in collecting the fraudulent billings on a daily basis. Sidhu does not dispute that insurers were systematically defrauded by billings produced in Gifford’s practice. Gifford and Sidhu were tried before a jury and convicted on multiple counts relating to the fraud. 2 Gifford was convicted of conspiracy to commit mail fraud, aiding and abetting mail fraud, mail fraud, obstruction of justice, and engaging in a monetary structuring transaction. On appeal, Gifford does not challenge the facts establishing his guilt. Rather, Gifford attacks almost all of the fact findings used by the district court to determine his sentence. Sidhu was convicted of conspiracy to commit mail fraud, aiding and abetting mail fraud, and making false statements to the FBI. On appeal, Sidhu challenges both his conviction and his sentence, arguing: (1) that the evidence was insufficient to support his conviction; (2) that his trial counsel was constitutionally deficient; and (3) that the district court incorrectly calculated the monetary loss attributable to Sidhu’s crimes, thus arriving at an erroneous base offense level. Gifford’s and Sidhu’s appeals will be addressed separately. SIDHU’S APPEAL I. Sufficiency of the Evidence to Support Sidhu’s Conviction Sidhu’s convictions must be affirmed if a rational trier of fact could have found the essential elements of each offense beyond a reasonable doubt. See United States v. Brown, 29 F.3d 953, 958 (5th Cir.1994). All inferences and credibility determinations must be resolved in favor of the jury’s verdict of guilty. United States v. McCord, 33 F.3d 1434, 1439 (5th Cir.1994). A. Conspiracy to Commit Mail Fraud Sidhu’s conviction for conspiracy to commit mail fraud in violation of 18 U.S.C. § 371 must be supported with sufficient evidence that Sidhu and Gifford agreed to commit mail fraud, and that either Gifford or Sidhu committed an overt act in furtherance of the agreement. United States v. Gray, 96 F.3d 769, 772-73 (5th Cir.1996), cert. denied, — U.S. -, 117 S.Ct. 1275, 137 L.Ed.2d 351 (1997); United States v. Pettigrew, 77 F.3d 1500, 1519 (5th Cir.1996); United States v. Mackay, 33 F.3d 489 (5th Cir.1994). Sidhu admits that he knew Gifford was systematically defrauding insurers by submitting fraudulent claims. Sidhu acknowledges that he performed acts that furthered Gifford’s fraud. Therefore, the second element of the conspiracy offense is not challenged. Rather, Sidhu claims that the government failed in its burden to demonstrate an agreement between he and Gifford to commit mail fraud. The agreement forming the basis of a conspiracy is rarely expressed, and may be inferred from circumstantial evidence. E.g., Pettigrew, 77 F.3d at 1519. The key issue is whether Sidhu knowingly and voluntarily joined Gifford’s course of action. Id.; see also Gray, 96 F.3d at 772-73. The jury’s inference that Sidhu and Gifford agreed to commit mail fraud is supported by ample evidence. Gifford entrusted his practice to Sidhu’s care when Gifford was out of *649 the office. Sidhu checked to see how payment would be made before seeing patients. Sidhu instructed the staff to bill for injections using a particular code. Patients testified that Sidhu’s methods were highly irregular and the jury heard evidence that strongly supports the proposition that Sidhu was not performing biofeedbaek at all. For example, Sidhu billed for biofeedback that was performed without the benefit of his biofeedback machine. One patient testified that Sidhu slept during a purported biofeedback session while the patient listened to music. Another patient testified that she was only with Sidhu for a brief time period, and that she was forced to flee when he began touching her inappropriately. Other patients saw Sidhu after taking injections of morphine, which induced prolonged sleep and would have rendered the patient unable to actively participate in biofeedback. Although the jury heard conflicting testimony on the issue of proper biofeedback technique, there is sufficient evidence in the record to establish that at least some of Sidhu’s billings for biofeedback were fraudulent. Sidhu also performed a variety of services for patients in Gifford’s absence, including filling out and dispensing pre-signed prescription forms, determining whether a patient needed prescribed medication, and counseling patients. Many of Sidhu’s services were subsequently billed by Gifford as face-to-face psychotherapy sessions. Sidhu obtained knowledge of these fraudulent billings because Sidhu was provided with computer print-outs of Gifford’s billings, which Sidhu was periodically held responsible for collecting. Sidhu does not dispute that he knew the billings to be fraudulent. Thus, Sidhu performed an important role in furthering the conspiracy by fraudulently billing for biofeedback services that were improperly or incompletely performed, by performing services that could be billed as though provided by a physician, and by assuming a role of responsibility for the practice from Gif-ford. Of equal importance, Sidhu inquired into the status of collections on a daily basis. Sidhu actively collected on debt he knew to be fraudulent by meeting and corresponding with insurance companies and patients. Si-dhu met with insurance adjustors. Sidhu was also involved in collecting funds from patients. One patient testified that Sidhu presented her with a $10,000 bill from Gif-ford but offered to excuse the debt, provided that the patient would sign a release exonerating Gifford of liability for misdiagnosing the patient with a terminal illness. Thus, Sidhu performed an important role in completing the fraudulent transactions by actively pursuing collection of the fraudulent bills. Sidhu maintains that he cannot be held liable because he did not personally create any fraudulent billings. Sidhu is correct that Gifford generally circled, or at least approved, the codes that would be billed to a particular patient or insurer. That does not, however, end the inquiry. The record contains sufficient evidence to support an inference that at least some of Sidhu’s billings for biofeedback were fraudulent. Moreover, Si-dhu may be held liable for the reasonably foreseeable conduct of his co-conspirator, Gifford. See, e.g., United States v. Sanchez, 961 F.2d 1169 (5th Cir.1992). In a related argument, Sidhu suggests that his conduct was not voluntary because, having failed his medical exams, he needed his job. Sidhu’s contention that he needed a job, or that he acted at Gifford’s direction, establishes only Gifford’s superior role in the conspiracy. Such an allegation is insufficient to establish a necessity defense that would excuse Sidhu’s criminal conduct. See United States v. Willis, 38 F.3d 170, 175 (5th Cir.1994) (duress defense requires, inter alia, that the defendant reasonably believe there is a threat of death or serious bodily injury that leaves the defendant no reasonable alternative to violating the law), cert. denied, 515 U.S. 1145, 115 S.Ct. 2585, 132 L.Ed.2d 834 (1995). Sidhu performed work knowing that the services would be fraudulently billed. Sidhu assisted the fraudulent scheme by collecting on those fraudulent billings. There is sufficient evidence in the record to support the jury’s inference that Sidhu and Gifford agreed to commit mail fraud. Accordingly, Sidhu’s conviction for conspiracy to commit mail fraud will be affirmed. *650 B. Aiding and Abetting Mail Fraud Sidhu’s conviction for aiding and abetting mail fraud in violation of 18 U.S.C. § 1341 must be supported with sufficient evidence that Sidhu: (1) voluntarily associated with the criminal enterprise; (2) voluntarily participated in the venture; and (3) sought by independent action to make the venture succeed. United States v. Casilla, 20 F.3d 600, 603 (5th Cir.1994). Evidence sufficient to support a conspiracy conviction is typically sufficient to support an aiding and abetting charge. Id.; United States v. Mergerson, 4 F.3d 337, 342 (5th Cir.1993). To the extent Sidhu’s independent action comprises a separate element of the aiding and abetting offense, that element is satisfied with evidence that Sidhu independently used the mail to collect on what he knew to be fraudulent billings. There is, therefore, ample evidence to support the jury’s verdict with respect to Sidhu’s conviction for aiding and abetting mail fraud, and that conviction will be affirmed. C. Making False Statements to the FBI During the course of the FBI’s investigation into Gifford’s billing practices, Sidhu was interviewed by the FBI. In that interview, Sidhu falsely stated that his only role in Gifford’s practice was to perform biofeedback. Specifically, Sidhu claimed that he did not do any counseling, that he had no knowledge of the procedures used when Gifford was out of the office, and that he had no role in dispensing prescriptions. Sidhu was prosecuted and convicted for making these false statements to the FBI, in violation of 18 U.S.C. § 1001. Sidhu admits he made the statements. Sidhu concedes, as he must, that there is evidence in the record to support the conclusion that statements were false, at least as they were understood by the FBI. Sidhu’s primary position is that exculpatory statements denying his role in the offense should not be punishable by 18 U.S.C. § 1001. As Sidhu recognizes, however, we decided that issue in United States v. Rodriguez-Rios, 14 F.3d 1040 (5th Cir.1994) (en banc), which eliminated the “exculpatory no” exception to § 1001 liability. Even assuming we were sympathetic to Sidhu’s position, a panel of the Court is without authority to reconsider the decision of the en banc Court in Rodriguez-Rios. Sidhu also contends that his misleading statements were not material. Statements are material within the meaning of § 1001 when they have the natural tendency or capacity to deceive, affect, or influence the federal agency. Kungys v. United States, 485 U.S. 759, 769-71, 108 S.Ct. 1537, 1546, 99 L.Ed.2d 839 (1988); United States v. Swaim, 757 F.2d 1530, 1534 (5th Cir.1985); United States v. McIntosh, 655 F.2d 80, 82 (5th Cir.1981). The issue of materiality was properly presented to the jury, which found that Sidhu’s false statements to the FBI in the course of an active investigation were material to that investigation. Having reviewed the record, we find no error in that determination. Sidhu’s conviction for making false statements to the FBI in violation of 18 U.S.C. § 1001 will be affirmed. II. Effective Assistance of Counsel Sidhu next contends that his convictions cannot stand because he received constitutionally defective assistance of counsel. Specifically, Sidhu complains that his counsel should have (1) moved to sever Sidhu’s case from Gifford’s and (2) objected to the district court’s calculation of the loss attributable to Sidhu’s offenses. Sidhu’s claim of ineffective assistance is not ripe for appellate review. Generally, a claim for ineffective assistance of counsel is not reviewed on direct appeal when, as here, there has been no development of the issue in the district court. United States v. Rinard, 956 F.2d 85, 87 (5th Cir.1992); United States v. Lewis, 902 F.2d 1176, 1180 (5th Cir.1990). Even if the Court were to review the issue, Sidhu could not demonstrate either constitutionally deficient performance or the type of prejudice required to establish an ineffectiveness claim. Sidhu’s counsel moved for a judgment of acquittal notwithstanding the verdict. In that pleading, counsel raised both the sufficiency issues raised by Sidhu on *651 appeal and the prejudice to Sidhu as a result of the joint trial. United States v. Capote-Capote, 946 F.2d 1100, 1104 (5th Cir.1991) (counsel’s failure to file a motion to sever is not error unless the defendant can demonstrate specific compelling prejudice against which the district court was unable to afford protection). Sidhu’s counsel also made arguments concerning the amount of loss attributable to Sidhu in Sidhu’s objections to the PSR and at sentencing. As a result, Sidhu’s relatively less culpable role was developed and accounted for, both at trial and in sentencing. Sidhu’s generalized assertions of prejudice are insufficient to meet the rigorous standards governing claims that trial counsel provided ineffective assistance. III. Loss Calculation used in Sentencing Sidhu objects to the amount of monetary loss attributed to his offenses for purposes of determining his base offense level. The district court’s loss calculation is reviewed for clear error. United States v. Krenning, 93 F.3d 1257, 1269 (5th Cir.1996). The “loss need not be determined with precision. The court need only make a reasonable estimate of the loss, given available information.” U.S.S.G. § 2F1.1, comment, (n. 8). Sidhu argues that the loss attributed to him should have been lower because he was the less culpable co-conspirator. Specifically, Sidhu claims he should be sentenced only for harm that he directly caused or intended. We disagree. Gifford’s conduct was reasonably foreseeable to Sidhu, and in furtherance of their jointly-undertaken criminal activity. U.S.S.G. § lB1.3(a)(1); United States v. Carreon, 11 F.3d 1225, 1232-34 (5th Cir.1994). Therefore, Sidhu may properly be sentenced for Gifford’s conduct in furtherance of the conspiracy. Moreover, the district court properly accounted for Sidhu’s relative culpability by: (1) limiting Sidhu’s loss calculation to the period during which he was employed by Gifford, and (2) treating Sidhu as a “minor participant.” As a result, Sidhu was sentenced to 37 months using an offense level of 21 and criminal history category of I. Gifford was sentenced to 120 months using an offense level of 30 and a criminal history category of II. Sidhu was ordered to pay a fine in the amount of $2,000. Gifford’s fine was waived, but he was ordered to pay restitution in the amount of $150,-899.27. Sidhu’s objection to his sentence is general. He believes he should have received an across-the-board discount based upon his position in the conspiracy. Sidhu does not, however, make any specific arguments about which amounts cannot be fairly attributed to him. Likewise, Sidhu failed to offer specific evidence rebutting the PSR in the district court. The district court’s reliance on the PSR, as adjusted to accommodate certain of Sidhu’s objections in the district court, was reasonable. See United States v. Ayala, 47 F.3d 688, 690 (5th Cir.1995); United States v. Angulo, 927 F.2d 202, 204 (5th Cir.1991). Sidhu has not demonstrated that the district court’s allocation of loss, and ultimate determination of a base defense level, is clearly erroneous. Sidhu’s sentence will be affirmed. GIFFORD’S APPEAL Gifford challenges only his sentence. Gif-ford’s convictions were grouped into three offense groups, pursuant to the grouping rules of sentencing guidelines 3D 1.2 and 3D1.3. Gifford was then sentenced using the following PSR recommendations: Conspiracy to Commit Mail Fraud and Mail Fraud Counts Base offense level 2Fl.l(a) 6 Amount of loss 2Fl.l(b)(l) 12 More than minimal planning 2Fl.l(b)(2) 2 Aggravating role 3Bl.l(a) 4 Vulnerable victim 3Al.l(b) 2 Abuse of position of trust 3B1.3 2 Obstruction of justice 3C1.1 2 Adjusted offense level 30 Obstruction of Justice Counts Base offense level 2J1.2(a) 12 Threat of injury 2J1.2(b)(l) 8 Aggravating role 3Bl.l(a) 4 Vulnerable victim 3Al.l(b) 2 Abuse of position of trust 3B1.3 2 Adjusted offense level 28 Financial Transaction Counts Base offense level 2S1.2(a) 17 Specified activity 2S1.2(b)(l)(B) 2 Aggravating role 3Bl.l(a) 4 Vulnerable victim 3Al.l(b) 2 Abuse of position of trust 3B1.3 2 Obstruction of justice 3C1.1 2 Adjusted offense level 29 *652 Gifford’s ultimate term of imprisonment was determined using the highest offense level, pursuant to the grouping rules of guideline 3D1.4. Thus, Gifford was sentenced using a base offense level of 30. Gif-ford articulates six issues for review. Taken together, these issues effectively challenge most of the adjustments made to determine Gifford’s base offense level. I. Adjustments to the Obstruction of Justice and Financial Transaction Offense Groups Gifford’s first two issues raise arguments that relate solely to the district court’s determination of the base offense level applicable to his convictions for obstruction of justice and engaging in an unlawful financial transactions. Specifically, Gifford challenges the district court’s findings that (1) Gifford’s obstruction of justice count involved a threat of physical injury, and (2) that the financial transaction offenses involved an obstruction of justice. See U.S.S.G. § 2J1.2(b)(1) (allowing eight-level adjustment if offense involved threat of physical injury); § 2S1.2(b)(l)(B) (allowing two-level adjustment if defendant knew funds were proceeds of unlawful activity). After reviewing the record and the arguments of the parties with respect to each of these issues, the Court finds no basis for holding that the district court’s fact findings are.clearly erroneous. Although we find no error with respect to the adjustment of the obstruction of justice offense group for threat of physical injury or the financial transaction offense group for obstruction of justice, we note that any such error would also be harmless. Granting Gifford relief with respect to these issues would change only the guideline range applicable to the obstruction of justice and financial transaction offense groups. Granting relief would not, however, have any effect with respect to the controlling guideline range. Guideline 3D1.2 provides rules for grouping certain convictions into offense groups. Gifford’s convictions in this case were grouped into three offense groups: the mail fraud offense group; the obstruction of justice offense group; and the financial transaction offense group. Gifford does not challenge these groupings. Guideline 3D1.4 provides for determination of a “combined offense level.” A combined offense level is determined by taking the highest of the base offense levels applicable to the various offense groups, and adding points to that level based upon the base offense levels of the other offense groups. The district court made fact findings which yielded a base offense level of 30 with respect to the mail fraud offense group. As developed more fully below, we affirm the district court’s determination of that base offense level. The district court was then required, pursuant to guideline 3D1.4 to add points to that base offense level to account for the obstruction of justice and financial transaction offense groups. In this case, however, any addition to the base offense level of 30, the level applicable to the mail fraud offense group, would result in a guideline range that was higher than the highest of the statutory maximum sentences allowed by law. 3 For that reason, the combined offense level was effectively limited to 30 by the statutory maximum sentence that could be imposed, and the base offense levels for the obstruction of justice and financial transaction offense groups played no role in determining Gifford’s guideline range. Gif-ford’s arguments that relate solely to the obstruction of justice and financial transaction offense groups present no error. Moreover, assuming the Court were to find error with respect to those adjustments, the error is without effect as to the application of the sentencing guidelines in Gifford’s case. Accordingly, the district court’s adjustment to the obstruction of justice offense group for threat of physical injury and the district court’s adjustment to the financial transac *653 tion offense group for obstruction of justice will be affirmed. II. Calculation of the Loss Attributable to Gifford’s Fraud Gifford’s third issue challenges the district court’s calculation of the loss attributable to his fraud. The base offense level for fraud and deceit is six. U.S.S.G. § 2Fl.l(a). The district court added twelve points to the base offense level applicable to the mail fraud offense group because the loss exceeded $1.5 million. See U.S.S.G. § 2Fl.l(b)(l)(M). The loss calculation was initially prepared by the FBI and later incorporated into the PSR. The district court adopted the calculations in the PSR, finding that Gifford intended the insurance carriers to suffer a loss exceeding $2,020,419.79. Testimony at trial established that the FBI calculated the loss attributable to Gifford by extracting numbers from Gifford’s own computer. The FBI properly limited its analysis to billings made between January 1993 and August 1994, the time period defined in the indictment for the conspiracy. The PSR reports that the FBI also limited its analysis to those billing codes that required Dr. Gifford to perform or be present for the procedure. Finally, the FBI limited its analysis to billings actually submitted to an insurance carrier. The FBI calculated that seventy-four percent of Gifford’s 1993 billings and sixty-eight percent of Gifford’s billings between January and August 1994 were fraudulent. To reach that conclusion, the FBI relied upon witness interviews and other evidence to estimate that Gifford was generally available for work an average of fifty hours per week. Allowing Gifford fifty hours per week, the FBI calculated that Gifford could have worked 2,600 hours per year. From this figure, the FBI subtracted time that Gifford was known to be out of the office, based upon travel receipts and other evidence. Using this method, the FBI estimated that Gifford was available for work 2,017 hours in 1993 and 1,236 hours between January and August 1994. The FBI then compared the amount of time that Gif-ford was determined to be available for work with Gifford’s actual billings. Time-sensitive billing or other procedures that were in excess of the time that Gifford could have been available for work were determined to be fraudulent. Gifford does not seriously quarrel with the government’s methodology. Rather, Gifford argues (1) that he never intended to recover the face amount of the fraudulent claims, and (2) that the loss should be reduced to reflect that the services performed by non-physician employees had some value. A. Gifford’s Intent to Collect on the Fraudulent Claims The private insurers billed by Gif-ford generally paid only eighty percent, while some government programs paid as low as fifty percent of the amounts billed for Gif-ford’s services. Gifford claims that he never intended to collect fraudulently billed amounts that were not covered by the insurers. Thus, Gifford argues that the overall amount of loss attributed to his fraud should be reduced by between twenty and fifty percent. 4 We disagree. The record contains testimony establishing that Gifford accepted cash payments from patients that were intended to supplement the amounts paid by insurance companies. One of Gifford’s patients testified that cash payments of $100 per week *654 were made to cover the gap between Gif-ford’s billings and available insurance coverage. Sidhu was involved in collecting money directly from patients for Gifford’s bills. Finally, Gifford secured credit card numbers from patients for the purposes of billing directly when the insurance companies did not pay. Thus, the record establishes that Gif-ford did intend, and did in fact accept payment from patients that was in addition to the amounts billed to insurance companies. Therefore, the district court’s reliance on the PSR, which used the face amount of the fraudulent billings as the measure of loss, was not clearly erroneous. Of equal importance, Gifford has not provided the Court with a record or an argument on appeal that is specific enough to support relief on this ground. Gifford did not specify in the district court and has not specified on appeal which billings are subject to reduction or by what percentage those billings should be reduced. There is, therefore, no basis for finding that the district court’s reliance upon the PSR in this complicated insurance fraud case was clearly erroneous. United States v. Ayala, 47 F.3d 688, 690 (5th Cir.1995); United States v. Angulo, 927 F.2d 202 (5th Cir.1991) (“in the absence of rebuttal evidence, the sentencing court may properly rely upon the PSR and adopt it”). Gifford’s premise that he never intended to collect fraudulently billed claims that were not paid by the insurers is contradicted by the record. Moreover, Gifford has not presented any specific evidence or argument that would allow relief on this ground. For each of these reasons, the district court’s reliance on the PSR, which determined loss according to the face value of fraudulently submitted claims, will be affirmed. B. Accounting for the Value of Services Rendered by Non-Physicians With respect to Gifford’s fraudulent billings for services that were provided by another non-physician employee, Gifford argues that the loss calculation should be reduced by the value of the services actually rendered by the non-physician. Gifford failed to develop this argument in the district court. Specifically, there is no evidence that would enable the Court to distinguish between fraudulent billings for services that were not performed, fraudulent billings for services that were improperly or incompletely performed, and fraudulent billings for services performed by non-physician employees. Moreover, there is no evidence concerning the value, if any, of services performed by non-physicians, and no evidence to establish that Gifford could have relied upon the codes billed to seek reimbursement in any amount for the services provided by non-physician employees. Gifford recognizes that the record is inadequate, but requests that the Court vacate his sentence and remand for determination of those issues. Gifford had ample opportunity to develop his rather specific challenges to the district court’s allocation of loss in the district court. See United States v. Bachynsky, 949 F.2d 722, 732-33 (5th Cir.1991). Gifford bore the burden of producing specific rebuttal evidence to support his argument that some discrete portion of the fraudulent billings was not in fact fraudulent because valuable services that could have been reimbursed were rendered by non-physician employees. United States v. Ayala, 47 F.3d 688, 690 (5th Cir.1995); United States v. Angulo, 927 F.2d 202 (5th Cir.1991) (“in the absence of rebuttal evidence, the sentencing court may properly rely upon the PSR and adopt it”). Gifford failed in this burden. We are a court of error. We will therefore decline Gifford’s invitation to remand for development of an argument that should have been developed in the district court. Based upon the existing record, the district court’s calculation of the loss attributable to Gif-ford’s fraud was not clearly erroneous and will be affirmed. III. Adjustment Based Upon Gifford’s Aggravating Role Gifford’s fourth issue challenges the district court’s finding that Gifford played an aggravating role in the offenses, which resulted in the addition of four points to his base offense level. Guideline section *655 3B1.1 allows adjustment of the base offense level when the criminal defendant is found to be “an organizer or leader of a criminal activity that involved five or more participants or was otherwise extensive.” An aggravating role determination is reviewed for clear error. See United States v. Allibhai 939 F.2d 244, 252 (5th Cir.1991). No one contends that there were more than five participants in the criminal activity. Eather, the government argues that Gifford’s scheme to defraud was “otherwise extensive” because it required the “unknowing services” of multiple outsiders. See Allibhai 939 F.2d at 253 (unknowing services of bank employees, as well as the scope and duration of the conspiracy made money laundering scheme “otherwise extensive”); U.S.S.G. § 3B1.1, comment, (n. 3) (“a fraud that involved only three participants but used the unknowing services of many outsiders could be considered extensive”). Gifford created and managed an extensive scheme that generated more than $2 million dollars in fraudulent billings in the nineteen-month period between January 1993 and August 1994. Cf. Allibhai 939 F.2d at 253 (finding that money laundering scheme that yielded only $1 million dollars in laundered money over a three-year period was otherwise extensive). Gifford recruited numerous office employees to provide billing and collection support for his fraudulent practices. Of equal importance, Gifford’s far-reaching fraud could not have succeeded without the unwitting participation of his vulnerable patients and the unknowing assistance of employees in the many insurance companies that received Gifford’s fraudulent billings. The district court’s finding that Gifford played an aggravating role in an extensive conspiracy to defraud health insurance companies and government programs is not clearly erroneous and will be affirmed. IV. Adjustment Because Gifford’s Crimes Impacted Vulnerable Victims Gifford fifth issue challenges the district court’s finding that Gifford’s offenses impacted vulnerable victims, which resulted in the addition of two points to his base offense level. See U.S.S.G. § 3Al.l(b). The district court’s imposition of the two-level increase was based upon its judgment that Gifford’s patients were unusually vulnerable to criminal activity. Gifford argues that the increase was inappropriate because his patients were not the victims of his offense. Eather, Gifford maintains that the primary victims of his criminal conduct were the fraudulently billed insurers and government programs. We have previously recognized that a physician’s patients can be victimized by a fraudulent billing scheme directed at insurers or other health care providers. See United States v. Bachynsky, 949 F.2d 722 (5th Cir.1991). In Bachynsky we recognized that “the deep pockets” paying phony insurance claims are not the only victims when a doctor’s unwitting patients are made the instrumentalities of a fraudulent billing scheme. Id. at 735; see also United States v. Kuban, 94 F.3d 971, 974 (5th Cir.1996) (discussing who can be a victim for purposes of guideline 3A1.1), cert. denied, — U.S. -, 117 S.Ct. 716, 136 L.Ed.2d 635 (1997). The rationale applicable in Bachynsky is equally applicable here. Gifford’s patients were often debilitated by pain or depression, and easily became addicted to the treatment proffered by Gifford to support his fraud. The record supports the conclusion that Gifford preyed upon vulnerable patients by addicting them to morphine in order to support his fraudulent billing scheme. Gifford’s patients were therefore victims of that scheme and the district court’s imposition of a two-level adjustment was not clearly erroneous. V. Adjustment Based Upon Gifford’s Abuse of Position of Trust or Special Skill Gifford’s final issue challenges the district court’s finding that Gifford abused a position of trust, which resulted in the addition of two points to his base offense level. See U.S.S.G. § 3B1.3. Gifford correctly argues that the adjustment must stand or fall on the issue of whether he abused a position of trust. Guideline 3B1.3 does not allow both an adjustment for an aggravating role in the offense, which we have already affirmed, and an adjustment based “solely on *656 the use of a special skill.” U.S.S.G. § 3B1.3. The district court did not, however, base Gifford’s adjustment “solely on the use of a special skill.” To the contrary, the PSR justified the adjustment with evidence that Gifford abused his position of trust with his patients. In the sentencing hearing, the government further defended the adjustment by arguing that Gifford abused his position of trust with the insurers and government programs that provided reimbursement with respect to the fraudulent billings. Having reviewed the record, we are convinced that compromising his patients’ trust was a necessary component of Gifford’s lucrative scheme to maximize his earnings. Gifford’s abuse of his patients’ trust “significantly facilitated the commission” of the offense. U.S.S.G. § 3B1.3. The district court’s two-level adjustment under § 3B1.3 is not clearly erroneous and will be affirmed. CONCLUSION Sidhu’s conviction is supported by adequate evidence with respect to all counts. Sidhu’s claim for ineffective assistance of counsel was not developed in the district court, and is therefore inappropriate for appellate review at this time. Sidhu’s relatively less culpable role in the fraudulent scheme was developed at trial and accounted for by the district court in the sentencing hearing. The district court reduced Sidhu’s sentence to account for his relatively less culpable role. For that reason, the district court did not err by attributing the total amount of reasonably foreseeable loss to Sidhu. Accordingly Sidhu’s convictions and sentence are AFFIRMED. Gifford’s multiple arguments challenging the findings used by the district court to determine his base offense level do not present clear error. Accordingly, Gifford’s sentence is AFFIRMED. Footnotes [1] Dr. Gifford is variously referred to in the record as Dr. Loren Gifford or Dr. Loring Gifford. This opinion will use Dr. Gifford's name as specified in his appellate brief. [2] Gifford’s wife, who worked as a chiropractor in his office, was also indicted but the charges against her were dismissed. Gifford’s remaining office staff were not indicted. [3] The maximum statutory sentence that could be imposed was 120 months, which applied to Gif-ford’s convictions for violation of 18 U.S.C. § 1512 and 18 U.S.C. § 1957. Gifford has not challenged the validity of those convictions on appeal. Gifford’s guideline range using a base offense level of 30 and a criminal history category of II was 108-135 months. The guideline range applicable to a base offense level of 31 and a criminal history category of II is 121-151 months, a period in excess of the statutory maximum of 120 months. [4] Gifford attempts to support this argument with commentary note 7 to sentencing guideline 2F1.1, which states that the court may substitute "intended loss” when it is greater than the "actual loss.” U.S.S.G. § 2F1.1 comment, (n. 7). The commentary cited by Gifford permits the district court to increase the amount of loss attributable to an offense by including losses that the defendant intended to inflict. Whatever more subtle meaning the commentary may be trying to convey, we seriously doubt that the provision can be read to require that the district court define an intended loss in fraud cases, and then substitute that intended loss for more reliable and less subjective estimates of loss, such as the face amount of fraudulent claims. See United States v. Lghodaro, 967 F.2d 1028, 1031 (5th Cir.1992) (total amount of fraudulent claims, rather than amount paid by the insurer, established loss); see also United States v. Sowels, 998 F.2d 249, 252 (5th Cir.1993) (combined credit limit of stolen credit cards represented intended loss regardless of the actual charges made).