GEORGE C. PRATT, Circuit Judge:
Ludwig Wittgenstein wrote that “Philosophy is a battle against the bewitchment of our intelligence by.means of language.”
Philosophical Investigations
47 (1953). Dr. Naveed A. Siddiqi apparently found out — the hard way — that making a proper Medicare claim was a battle against the bewitchment of his intelligence by means of bureaucracy. His bewitchment resulted in five felony convictions.
This appeal revolves around the code number “96500”, which Siddiqi entered on his Medicare claim form to denote that he had provided a certain reimbursable medical service. Although no one — including us — seems to know exactly what “96500” means, the government assures us that whatever service “96500” stands for, Dr. Siddiqi did not provide it, because he was out of the country at the time. Consequently, the government argues, Siddiqi’s convictions for mail fraud, theft of government property, and false claims, should stand.
FACTS AND BACKGROUND
Naveed A. Siddiqi is an oncologist in Elmira, New York. In 1973, he became a Medicare-approved participating physician (“provider”). Medicare is a federally-funded health insurance program which provides basic insurance coverage for outpatient medical services provided to individuals over age 65, as well as to other qualified individuals. The Medicare program is overseen by the Health Care Financing Administration (HCFA) of the United States Department of Health and Human Services.
Medicare is subdivided into two parts, commonly known as part A and part B. Part A covers health care services rendered primarily by hospitals, while part B covers those services rendered by physicians and other, non-institutional, providers.
See generally New York City Health & Hosps. Corp. v. Perales,
954 F.2d 854, 856 (2d Cir.1992). HCFA contracts with private insurance companies all over the United States to process the claims submitted under each plan. In the Elmira area, Blue Shield of Western New York (Blue Shield) administers Part B of the Medicare program.
Participating providers submit claims, either by mail or by computer, to Blue Shield. Providers are given Blue Shield’s
Doctor’s Manual,
a three-ineh-thick loose-leaf binder containing various descriptions of medical services and five-digit numerical codes for use in claims for reimbursement. Each provider is required to select from the
Manual
the code which most accurately describes the services rendered. After processing, Medicare mails the check to the physician.
A grand jury in the Western District of New York charged Siddiqi with 77 counts of fraud; more specifically, the indictment charged Siddiqi with violations of 18 U.S.C. §§ 287 (false claims), 641 (theft of government property), 1001 (false statements), and 1341 (mail fraud). For our purposes, the pertinent part of the superseding indictment reads as follows. Only the underscored portions were disputed at trial.
[F]rom on or about March 11, 1988 and continuing thereafter up to and including April 27, 1989, within the Western District of New York, the defendant, Na-veed A. Siddiqi, did devise and intend to devise a scheme and artifice to defraud the Medicare program by attempting to obtain and obtaining the payment of Medicare insurance claims by means of false and fraudulent pretenses, statements and representations to [Blue Shield], the Medicare contract carrier, well knowing these said misrepresentations were false and fraudulent when made. The scheme and artifice to defraud being in substance as follows:
1. It was part of the scheme that Medicare beneficiaries who were patients of Naveed A. Siddiqi were scheduled for appointments at St. Joseph’s Hospital for the administration of chemotherapy treatment by hospital medical staff on an out-patient basis.
2. It was a further part of the scheme that the hospital medical staff would administer the chemotherapy upon Naveed A. Siddiqi’s oral instructions.
3. It was a further part of the scheme that no chemotherapy treatment was administered to patients of Naveed A. Sid-diqi at his office on the same day that the patient received chemotherapy treatment at the hospital.
4. It was a further part of the scheme that the chemotherapy provided during these scheduled appointments would be administered without the participation or supervision of Naveed A. Siddiqi.
5. It is a further part of the scheme that Naveed A. Siddiqi caused claims to be submitted electronically for chemotherapy administration to each patient on the same day that the patient received chemotherapy at St. Joseph’s Hospital.
6. It was a further part of the scheme that Naveed A. Siddiqi would direct each such claim to be billed as the administration of chemotherapy well knowing that chemotherapy was solely administered on that day to the patient at St. Joseph’s Hospital during which time Naveed A. Siddiqi was not participating in or supervising the administration of the chemotherapy.
7. It was a further part of the scheme that Naveed A. Siddiqi knew St. Joseph’s Hospital would file a claim to the Medicare program for the chemotherapy treatment performed on Naveed A. Siddi-qi’s patients.
8. It was further a part of the scheme that Naveed A. Siddiqi caused the duplicate claims for chemotherapy treatment knowing full well that the chemotherapy was administered only by the St. Joseph’s Hospital staff, in order to receive additional monies from Medicare.
9. It was a further part of the scheme that Naveed A. Siddiqi would receive checks from [Blue Shield] for payment of these purported services.
10. It was a further part of the scheme that Naveed A. Siddiqi did receive a substantial amount of money from Medicare through [Blue Shield] as a result of submitting false claims.
In essence, Siddiqi claimed to have billed Medicare for the “professional component” of the chemotherapy treatments given to his patients at St. Joseph’s Hospital.
The parties stipulated, pre-trial, “[t]hat between July 12, 1988 and July 28, 1988,
Naveed A. Siddiqi was on travel outside the United States and was not present in the United States between those two dates.” The five counts that resulted in convictions all related to claims submitted for services in relation to chemotherapy treatments given during the period in July 1988 when Siddiqi was out of the country.
In his opening statement, the assistant United States Attorney told the jury that Siddiqi’s crime was submitting claims to Medicare that stated,
in essence, “I performed that chemo. So I'm entitled to reimbursement for the cost of the administration of that chemo, for the cost of the drug and for the cost of the actual administration of the chemo.”
That’s what’s alleged. That’s all it is[.]
However, there was no proof that Siddiqi billed for the cost of the drug itself, and the case was ultimately submitted on a theory that Siddiqi had billed only for
services
that he did not perform.
The first witness called by the government was Adeline Adee, of Blue Shield. Through Adee, the government introduced the compendious
Doctor’s Manual,
which included this pertinent coding instruction:
CHEMOTHERAPY INJECTIONS
Procedures 96500-96549 are independent of the patient’s office visit. Either may occur independently from the other on any given day, or they may occur sequentially on the same day. Oncologists may see their patients at 2 to 4 week intervals with none to 5 chemotherapy procedures between visits.
96500 Chemotherapy injection, intravenous, single premixed agent, administered by qualified assistant
under supervision of physician
or by physician; by push technique.
(emphasis added).
Most of the government’s case, however, was presented through agent Joseph Niegsch of the U.S. Department of Health and Human Services. Niegsch explained that Siddiqi’s “scheme to defraud” consisted of billing Medicare roughly $40.00 under the 96500 code for services rendered, “when it was actually hospital staff and hospital personnel that administered the chemotherapy”.
All of the relevant chemotherapy treatments were administered at St. Joseph’s Hospital at Siddiqi’s express direction, and there was no dispute that the hospital, and not Siddiqi, “administered” the chemotherapy. Rather, Siddiqi claimed to have “supervised” the administration of the chemotherapy by writing the order, reviewing the patients’ charts on the days that chemotherapy was administered, and monitoring the patients’ progress in order to adjust the dosage or, if necessary, discontinue the treatment. In short, Siddiqi said, “the purpose of [the 96500] code was that we were responsible for that patient. We review their records; [and] if necessary we see them”.
It was undisputed that the term “supervision” is not defined in the
Doctor’s Manual.
Further, when agent Niegsch was asked what “supervision” in the 96500 code meant, he responded: “I really don’t have an opinion on what constitutes supervision.” Neither, for that matter, did any other government witness. However, Niegsch did testify that the government only charged Siddiqi for fraud on the days “when we believe he was not at the hospital.” This included the days between July 12 and July 28, when Siddiqi admittedly was out of the country.
Siddiqi’s theory of defense was that a physician provides a service (the “professional component”) compensable under Medicare when he sends his patients to the hospital for outpatient chemotherapy treatment, selects their treatment and dosage, monitors their progress through chart review, and remains available on short notice, if needed, during the hospital’s administration of the chemotherapy. The government’s witnesses never disputed that such a service would be reimbursable through the Medicare program; indeed, this is supported by the regulations,
see
42 C.F.R. § 410.20(a) (1990) (“Medicare Part B pays for physicians’ services, including
diagnosis,
therapy, surgery,
consultations, and
home, office, and
institutional calls.’’)
(emphasis added), as well as by a portion of the
Doctor’s Manual
not stressed by either party (defining “minimal level of service” as “a level of service supervised by a physician but not necessarily requiring his presence” and noting that “[e]ach level of service may be used by all physicians”).
Moreover, Siddiqi presented undisputed expert and lay testimony that there was widespread confusion regarding the proper billing code for the professional component of chemotherapy, that 96500 was the “standard practice” at the times charged in the indictment, and that even the American Society of Clinical Oncologists (ASCO) recommended the use of the 96500 code to bill for the professional component. Siddiqi testified that it was at an ASCO meeting that he first learned of the use of the 96500 code to bill this professional service.
As to the period that he was admittedly out of the country, Siddiqi testified that it was the normal practice to arrange for a substitute, or “covering” physician. From July 12 to July 28, Siddiqi testified that a Dr. Mufti was assigned to cover for him. Although Siddiqi generally cancels chemotherapy treatment for his patients when he is absent from the area, he testified, there were some patients — notably, for this case, Cecil Connelly and Arthur Kennedy, the subjects of the conviction counts of the indictment — who required continued outpatient hospital chemotherapy during his absence.
For those patients, Siddiqi testified that he personally selected the drug, dosage, and schedule for those patients, and made arrangements with Dr. Mufti to be available if needed during his absence, to check his patients’ charts, and to look for certain medical factors in his chart reviews for those patients who would continue to receive chemotherapy in his absence. Siddiqi did not, however, present any evidence to corroborate his “coverage” theory. Indeed, little other evidence was presented on the coverage issue, which is not surprising because the AUSA, in questioning one of Siddiqi’s witnesses, said “I don’t think anybody’s questioning that
coverage exists in this trial,
but my question relates to the method by which coverage is standardly reported in the hospital.” (emphasis added). In summation, however, the AUSA changed course and directly challenged Sid-diqi’s testimony, arguing to the jury that “there probably wasn’t any coverage” while Siddiqi was out of the country, and even if there was he never told the hospital about it.
Boiled down, the way the case evolved could be outlined as follows:
Government: Your claim was false because the hospital, not you, “administered” the chemotherapy. (See 116 of superseding indictment).
Sidpiqi: The 96500 code does not say that I administered. the chemotherapy. It says that I administered
or
supervised the chemotherapy.
Government: Perhaps. But you did not “supervise” the chemotherapy either.
Siddiqi: I diagnosed the need for chemotherapy, prescribed the dosage, monitored the results, and made myself available if needed. Isn’t that “supervision”?
Government.- I don’t know what “supervision” means, but for five of the 77 counts you claimed all of that when you were out of the country.
Siddiqi: When I was away, I arranged for Dr. Mufti to do those things I could not do in person.
Government: That’s doubtful, but in any event you never told the hospital about Dr. Mufti.
Siddiqi: Oh yes, I did.
In the jury’s mind, the only issue seems to have been whether Siddiqi told the truth when he said that he told the hospital that Dr. Mufti would be covering for him during his absence. With no corroboration of this last statement, however, the jury found against him — but only on the five counts relating to his time out of the country. On the other 72 counts, the jury found the government’s case lacking.
On July 10, 1991, the district court imposed a guideline sentence of no imprisonment, three years’ probation, 1,000 hours of community service as an emergency room physician, a fine of $2,000.00, $640.88
in restitution, and a $250.00 special assessment.
On July 15, 1991, Siddiqi filed a notice of appeal to this court, which we docketed as number 91-1425. However, on August 5, 1991, Siddiqi filed a motion for new trial in the district court,
see
Fed.R.Crim.P. 33, based on a newly-discovered document known as a “physicians away list”. This list, allegedly maintained by the hospital, consisted of the names of “away” physicians, the dates that they would be away, and the physicians who would be “covering” for the absent physicians. The “physicians away list” indicated that Siddiqi would be leaving on 7/13/88, returning on 8/1/88, and listed the name “Dr. M. Mufti” under “PHYSICIAN(S) COVERING”. In moving for a new trial, Siddiqi professed that this document was maintained by the hospital in the regular course of business, and was circulated to the hospital staff, including the staff responsible for administering the chemotherapy. Siddiqi represented that he discovered the document on July 3, 1991, and that it could not have been discovered earlier even exercising due diligence, because (1) during his pretrial investigation, Siddiqi asked the hospital for the memorandum slip he had prepared for the hospital indicating his vacation dates and covering physician; the hospital told him that they had destroyed the slip, but neglected to inform him of the existence of the physicians away list; and (2) Siddiqi did not appreciate the critical importance of such evidence until the government’s closing argument. The government opposed the motion, arguing (1) that Siddiqi had not made a sufficient showing of undiscovera-bility, and (2) that the evidence was “cumulative and would not have altered the outcome of the trial.”
In an unpublished order dated September 3, 1991, the district court denied Siddiqi’s motion, passing over the issue of whether the evidence could have been discovered before or during trial with the exercise of due diligence, but noting that
defense counsel amply presented his coverage theory to the jury, both through cross-examination of the various Government witnesses as well as through his own case-in-chief. Considering all of the evidence introduced at trial, I find that it is highly unlikely that the admission of this “physicians away list” would have led to a different result.
On September 11, 1991, Siddiqi filed a second notice of appeal under docket number 91-1539. By order of this court dated September 24, 1991, the two appeals were consolidated, and this appeal followed.
DISCUSSION
Siddiqi claims (1) that the evidence was insufficient to support his convictions on counts 31, 33, 40, 42 and 43; (2) that the district court erroneously failed to grant a new trial on the basis of the newly-discovered “physicians away list”; and (3) that the district court erred in failing to give two of Siddiqi’s proffered jury instructions. We reject Siddiqi’s challenges to the jury charge; we further conclude that there was, barely, sufficient evidence to support Siddiqi’s convictions; however, we also conclude that the district court clearly erred in rejecting the proffered new evidence as cumulative. If this evidence turns out, after further proceedings in the district court, to be new and admissible, we conclude that the jury probably would have acquitted Siddiqi on the five counts which are the subject of this appeal. We therefore remand for the district court to determine (1) whether this evidence could have been discovered prior to or during trial with the exercise of due diligence, and if not (2) whether the new evidence is admissible.
As we have noted, we do have grave concerns about the sufficiency of the government’s case. However, we are forced to reject Siddiqi’s challenge to the sufficiency of the evidence, principally because, on the five counts of conviction, a jury could have disbelieved Dr. Siddiqi’s self-serving “coverage” testimony and then rationally concluded that, without such coverage, the doctor, who was admittedly out of the country, did not provide a service reimbursable by Medicare. Indeed, that is the most likely explanation for the jury’s
verdict—it convicted Siddiqi of only five of 77 counts, and the five counts of conviction were the only counts in the indictment which corresponded to the two-week period Siddiqi was admittedly out of the country. As the government candidly acknowledges in its brief on appeal, “It was the defense of * * * ‘coverage by another physician’ that obviously was rejected by the jury.”
Rule 33 of the Federal Rules of Criminal Procedure states that the district court may grant a new trial “if required in the interest of justice”, and it specifically contemplates that such motions may be made based on newly-discovered evidence. The district court’s decision to deny a new trial on the ground of newly-discovered evidence is reviewed for abuse of discretion.
See, e.g., United States v. Diaz,
922 F.2d 998, 1006-07 (2d Cir.1990) (citing cases),
cert. denied,
— U.S. -, 111 S.Ct. 2035, 114 L.Ed.2d 119 (1991). Relief is justified under rule 33 if the defendant makes a showing that the evidence is in fact “new”,
i.e.,
it could not have been discovered, exercising due diligence, before or during trial, and that the evidence is so material and non-cumulative that its admission “would probably lead to an acquittal.”
United States v. Alessi,
638 F.2d 466, 479 (2d Cir.1980) (citing cases);
United States v. Costello,
255 F.2d 876, 879 (2d Cir.),
cert. denied,
357 U.S. 937, 78 S.Ct. 1385, 2 L.Ed.2d 1551 (1958);
Berry v. State,
10 Ga. 511, 527 (1851) (new evidence will not entitle the defendant to a new trial unless “it would probably produce a different verdict.”).
One reason that there was sparse evidence to support Siddiqi’s convictions was that the government’s presentation of this case was fluid, even amoebic. Its initial theory of prosecution was that Dr. Siddiqi did not actually administer the chemotherapy; but when Siddiqi pointed out that he did not have to administer the chemotherapy, only
supervise
it, the government had to face up to the irrefutable facts that the 96500 code stands for administration
or
supervision and that the indictment alleged that the chemotherapy was administered without Siddiqi’s “participation
or
supervision”. Of course, the only government witness asked point-blank what “supervision” meant replied that he had “no opinion”.
In spite of the government’s inability to provide any definition of “supervision”, it still had a chance to salvage five of the counts—the ones corresponding to Dr. Sid-diqi’s trip abroad—by arguing to the jury that, whatever the elusive term “supervision” means, a doctor who is in another country wasn’t providing it. When the government was met with Dr. Siddiqi’s un-contradicted testimony that Dr. Mufti was covering for him while he traveled, the government seemingly conceded that “coverage exists in this case”—a concession which may well have lulled or misled Siddi-qi as to the need for corroboration of his coverage testimony—until the evidence was closed. Then, the AUSA argued in summation (1) that “there probably wasn’t any coverage”, and (2) that even assuming coverage, it probably wasn’t sufficiently communicated to the hospital staff to constitute “supervision”, whatever that may be. With the evidence already closed, Sid-diqi could no longer present corroborative evidence to counter the government’s last-gasp argument, asserted in summation for the first time, that Dr. Mufti’s coverage had not been communicated to the hospital staff that actually administered the chemotherapy.
In the circumstances of this case, if the proffered document is new and admissible in evidence, the interests of justice would demand vacatur of Siddiqi’s convictions. While the district court felt that the hospital’s “physicians away list” was theoretically cumulative of Siddiqi’s own oral testimony that he employed a covering doctor while away (as the district court concluded), it is certainly much less impeachable than the defendant’s own, self-interested testimony. Rather than being merely cumulative, it was evidence of a different kind and came from an impartial source. Moreover, the list would have been the only evidence that Siddiqi’s absence, and Mufti’s coverage, were communicated to the hospital staff. Given the fluid way in which this trial evolved, we are satisfied that accept-
anee of the physicians away list would probably have led the jury to acquit Siddiqi on all five counts.
We are extremely mindful that Medicare and Medicaid fraud constitute a great drain on a limited source of social funding.
See, e.g., United States v. Weiss,
930 F.2d 185 (2d Cir.),
cert. denied,
— U.S. -, 112 S.Ct. 133, 116 L.Ed.2d 100 (1991);
United States v. Weiss,
914 F.2d 1514 (2d Cir.1990), ce
rt. denied,
— U.S. -, 111 S.Ct. 2888, 115 L.Ed.2d 1053 (1991);
United States v. Huber,
603 F.2d 387 (2d Cir.1979),
cert. denied,
445 U.S. 927, 100 S.Ct. 1312, 63 L.Ed.2d 759 (1980);
United States v. Precision Medical Labs., Inc.,
593 F.2d 434 (2d Cir.1978);
United States v. Halper,
590 F.2d 422, 426 (2d Cir.1978). Those who perpetrate such fraud deserve relentless prosecution and severe punishment, and nothing in our opinion should be read as allowing such despicable individuals to hide behind the ambiguities of bureaucratic regulations. However, neither can we allow the government to ambush a defendant with that same ambiguity.
CONCLUSION
In its brief, the government suggests, however maladroitly, that there may be some question as to the authenticity or admissibility of the physicians away list. To resolve that question we are remanding for further proceedings. On remand, the district court should determine (1) whether the document could have been discovered before or during trial with the exercise of due diligence, and (2) whether the document is admissible into evidence. If it could have been discovered, or if it is not admissible for lack of authenticity or other reasons, then the convictions shall stand, since we find Siddiqi’s challenges to the jury charge meritless. However, if the document is both new and admissible, then in light of our conclusion that it probably would have led to an acquittal if presented to the jury, the district court shall order a new trial on counts 31, 33, 40, 42, and 43 of the indictment.