NAVEED A. SIDDIQI, Petitioner-Appellant, v. UNITED STATES OF AMERICA, Respondent-Appellee.

United States Court of Appeals for the Second Circuit

98 F.3d 1427; 1996 U.S. App. LEXIS 28125

January 23, 1996, Argued; October 31, 1996, Decided.

PRIOR HISTORY: Appeal from a denial by the United States District Court for the Western District of New York (Michael A. Telesca, Judge), of appellant's Section 2255 petition to vacate his conviction for the submission of fraudulent claims for Medicare reimbursement. We hold that Siddiqi was afforded the effective assistance of counsel at trial. However, we vacate his conviction because the government's shifting theories of guilt have caused a miscarriage of justice.

COUNSEL: Timothy K. Ford, MacDonald, Hoague & Bayless, Seattle, Washington (David Rothenberg, Geiger & Rothenberg, Rochester, New York, of counsel), for Petitioner-Appellant.

Christopher V. Taffe, Assistant United States Attorney, Western District of New York, Rochester, New York (Patrick H. NeMoyer, United States Attorney, of counsel), for Respondent-Appellee.



WINTER, Circuit Judge:

Naveed Siddiqi, M.D., appeals from Judge Telesca's denial of his petition under 28 U.S.C. @ 2255 to vacate his conviction and his sentence for Medicare billing fraud. Siddiqi asserts that he was convicted because of constitutionally ineffective assistance of counsel at trial. We find no merit in his ineffective assistance claim, but we nevertheless vacate his conviction and sentence. The government has, throughout this prosecution, adopted shifting theories of guilt. This inconstancy of position impeded Siddiqi's defense at trial and has severely hampered judicial consideration of this matter. At this final stage, in order to rebut a claim of ineffective assistance, the government now embraces a theory that is legally insufficient. A miscarriage of justice having occurred, we vacate the conviction.

1. Background

Siddiqi is an oncologist who, at pertinent times, practiced medicine in Elmira, New York. He was board-certified in internal medicine, oncology, and hematology. In 1973, he was approved to participate as a physician in Medicare, the federally funded health insurance program that provides basic medical coverage to individuals over age 65 and to other qualified individuals.

The Medicare program is overseen by the Health Care Financing Administration ("HCFA"), part of the United States Department of Health and Human Services ("HHS"). Medicare is divided 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 services rendered by physicians and other noninstitutional providers. HCFA contracts with private insurance companies across the country to process the claims submitted under Medicare. In the Elmira area, Blue Shield of Western New York ("Blue Shield") administers Part B of the Medicare program.

Participating physicians and other Medicare providers submit claims by mail or computer to Blue Shield. In submitting claims for reimbursement, providers use Blue Shield's Doctor's Manual, a three-inch-thick looseleaf binder containing various descriptions of medical services and five-digit numerical codes. Each provider selects from the Manual the code that most accurately describes the services rendered. After processing the claim, Medicare mails the check to the physician. A deliberately misleading use of a particular code would, of course, support a criminal fraud charge under various federal statutes. See, e.g., 18 U.S.C. @ 1341; 18 U.S.C. @ 287; 18 U.S.C. @ 1001.

2. The Indictment and Pretrial Proceedings

In 1990, a grand jury in the Western District of New York returned an indictment charging Siddiqi with 77 counts of fraud related to his Medicare billing practices between March 11, 1988 and April 27, 1989. Each count of the indictment involved one of three separate allegedly fraudulent schemes. The jury acquitted Siddiqi on every count related to two of the schemes. It convicted on five counts involving a scheme in which Siddiqi submitted allegedly false Medicare claims for chemotherapy treatments. These counts were for mail fraud, in violation of 18 U.S.C. @ 1341; theft of government property, in violation of 18 U.S.C. @ 641; and false claims, in violation of 18 U.S.C. @ 287.

The five counts of conviction alleged that Siddiqi defrauded the Medicare program by submitting claims for outpatient chemotherapy treatments that he prescribed for his patients and that were administered by the medical staff at St. Joseph's Hospital. The indictment stated that "no chemotherapy treatment was administered to [Siddiqi's patients] at his office on the same day that the patient received chemotherapy treatment at the hospital." It further stated that the hospital chemotherapy was "administered without [Siddiqi's] participation or supervision" and that Siddiqi "caused [**5] claims to be submitted . . . for 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 [Siddiqi] was not participating in or supervising the administration of the chemotherapy." Finally, the indictment charged that Siddiqi "knew St. Joseph's Hospital would file a claim to the Medicare program for the chemotherapy" and that Siddiqi thus "caused the duplicate claims for chemotherapy treatment . . . in order to receive additional monies from Medicare."

These counts all related to claims submitted for services in connection with chemotherapy treatments administered to patients between July 12, 1988 and July 28, 1988. The parties stipulated that Siddiqi, a Moslem, was on a pilgrimage to Mecca during that time. We will, therefore, style these five counts the "Mecca counts."

3. The Trial

In his opening statement, the prosecutor advanced a simple theory as to the Mecca counts: "This case is really all about Medicare fraud; phony billing, billing for services not rendered . . . . the doctor submitted a claim to Medicare saying, 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 administering the chemo.'" Siddiqi, the government asserted, was "claiming for services rendered by somebody else."

The prosecution's opening statement thus presented a simple theory of double billing: St. Joseph's billed Medicare for the cost of the drug and of administering it, while Siddiqi billed Medicare for the very same things. If proven, of course, Siddiqi would indeed be guilty of fraud no matter where he was at the time the chemotherapy was administered. Indeed, the prosecutor's opening never referred to Siddiqi's whereabouts at times pertinent to the Mecca counts.

However, the opening statement was a harbinger of the way in which this prosecution was to be conducted. The double billing theory was grand advocacy but had no basis in the evidence.

First, the Mecca counts involved Siddiqi's use of code 96500 in the Blue Shield Doctor's Manual to bill for the chemotherapy treatments in question. To convict Siddiqi on any of the Mecca counts, the government had to show knowing falsity in using that code. However, code 96500 does not state that it covers the cost of the chemotherapy drug itself. Siddiqi's use of that code was not a statement, therefore, that he had provided the drug and was seeking reimbursement for its cost. Moreover, there was no proof that Siddiqi billed for the cost of the chemotherapy drugs, each bill in the Mecca counts amounting to roughly $ 40.00, or roughly 10% of the bills submitted by the hospital for the drugs and their administration. In retrospect, it is extremely doubtful that the government ever had any evidence justifying the prosecutor's assertion that Siddiqi billed for the cost of the drugs.

Second, the opening statement was incorrect in asserting that Siddiqi claimed to have administered the drug. The 96500 code covers "chemotherapy injection, intravenous, single premixed agent, administered by qualified assistant under supervision of physician or by physician; by push technique." Because code 96500 covers "supervision" of the administration of chemotherapy by a qualified assistant, Siddiqi's designation of code 96500 was not necessarily a statement that he actually administered the drug.

To prove falsity under code 96500, therefore, the government had to prove that Siddiqi knowingly [**8] did nothing that might amount to supervision of the administration of the chemotherapy by the staff at St. Joseph's. The government's main witness in this regard was Joseph Niegsch, an agent with the Health & Human Services Inspector General's office. He testified that "the government alleges in [the Mecca counts] that the hospital actually performed the chemotherapy and not Dr. Siddiqi." The fraud "relates to the submission of Medicare claims for reimbursement by Dr. Siddiqi in which he listed as his claims or services that he provided patients -- he listed chemotherapy as the service he provided. All of the claims related to . . . services that were provided in the hospital. Now, Dr. Siddiqi on his claims states that he did the chemotherapy." Niegsch's testimony was of course wrong. Siddiqi's "statement" was the designation of code 96500, which includes "supervision" of the administration of chemotherapy by an assistant. Use of that code was not, therefore, a claim of having actually administered the drugs.

Niegsch impliedly filled this hole in the government's case when he later testified that Siddiqi could not have supervised or administered services for the Mecca claims because Siddiqi was out of the country at the time. He stated, "the government alleges they were fraudulent claims in that the services were actually provided in the hospital, but they were not provided by Dr. Siddiqi, and in many cases he was not in the hospital when the services were actually provided." Niegsch identified for the jury the dates on which Siddiqi charged Medicare under code 96500 and then read for the jury the stipulation that Siddiqi was out of the country from July 12 through July 28, 1988. Siddiqi's whereabouts thus became relevant by implication as evidence that he could not have supervised the administration of chemotherapy on the dates in question.

On cross-examination, Niegsch stated that Siddiqi "billed for services that were actually performed by someone else." Siddiqi's attorney asked Niegsch what the term "under supervision of physician" meant in the Doctor's Manual code 96500. Niegsch said he had no opinion.

At some point, perhaps by stipulation, "outpatient documents" relating to patients who received chemotherapy while Siddiqi was in Mecca were introduced. One of these related to a patient named Arthur Kenny. It bore a notation, "Call Dr. Siddiqi if any problems." There was no testimony during the government's main case concerning Arthur Kenny's outpatient document or other such documents, their relevance seemingly being simply to prove the administration of chemotherapy by the hospital while Siddiqi was in Mecca.

Oncology nurse Mary Ann Baker also testified as a government witness. She was in charge of the outpatient oncology service of St. Joseph's Hospital. Baker harbored no great friendship for Siddiqi and had declined to speak with his attorney before trial. She testified that Siddiqi did not attend the chemotherapy injection sessions. However, she also stated that it was a common practice for oncologists not to attend such sessions. Nevertheless, the doctors were available if needed, and she could reach them, including Siddiqi, by pager if necessary. Siddiqi's lawyer elicited from Baker on cross-examination that it is also a common practice for "physician A to have physician B cover for him." To the question, "And Dr. Siddiqi like others would on occasion as necessary have someone cover for him . . . in his absence?", she answered "yes." There was no redirect examination by the government.

Baker's testimony and cross-examination introduced issues that have since plagued this matter. First, she stated that the oncologist's absence during the administration of chemotherapy was common but that the doctor might yet be involved as being on-call if needed. Arguably, such availability might amount to "supervision" under code 96500.

Second, Baker stated that it was a common practice for another doctor to "cover" for one absent on vacation or otherwise. Whether or not the significance of coverage to the five Mecca counts was apparent at this point in the trial of the 77 counts, the government certainly gave no weight to it in bringing charges against Siddiqi. Baker had cooperated with the FBI and testified before the grand jury. In each instance, she had stated that Siddiqi usually had a Dr. Mufti cover for him when Siddiqi was away.

On the first appeal, we noted that Siddiqi presented "undisputed expert and lay testimony" that 96500 was the standard billing code for the "professional service" component of the administration of chemotherapy at the times charged in the Mecca counts and that this billing practice was recommended by the American Society of Clinical Oncologists. United States v. Siddiqi, 959 F.2d 1167, 1171 (2d Cir. 1992). A Dr. Weissman testified as a defense expert. He stated that considerable confusion existed over how to bill Medicare for chemotherapy and quoted one Medicare official as describing the area as "chaos." Weissman also stated that oncologists were rarely present at the actual administration of chemotherapy. He testified on direct that, after prescribing drugs and dosages, an oncologist might bill for being available for consultation and for observing charts that monitored a patient's reaction to the drugs even though there would be no contact with the patient. On cross-examination, he was asked a series of questions concerning whether an oncologist might bill Medicare for outpatient visits even though the physician performed no services. His answers boiled down to the following: If the doctor was available but performed no services, opinions differed among doctors as to whether a bill could properly be rendered; if the doctor was away, and not covered, no bill could be properly rendered. On redirect, he was asked about coverage, but the government successfully prevented this testimony on hearsay grounds.

Siddiqi testified as follows regarding the billing for services rendered while in Mecca:

Q. Now, there are certain services that are shown in three or four counts in the indictment that involve the period of time when you were out of the country and for which some services were provided in the hospital and office setting. Would you tell the jury the arrangements you made as far as coverage on your behalf? In other words, substitute coverage or the like.

A. Yes. I -- the people who were going to the hospital, I arranged with the oncology nurses for them to receive their treatment. So I arrange for their doses. I arrange for their schedule. I arranged for their modification of the doses if it becomes necessary. This was arranged before I left. And actually, when I left on Tuesday most of those people I actually saw on Monday. So that week basically was seen as I had seen them anyway. And after I arrange for their chemotherapy in the outpatient setting with heir drug dosage, the schedule, the number of drugs that they were going to get, I arranged for coverage with another physician, and I went ahead and I told him that there were these people who were going to be receiving treatments and they were going to get this type of treatment; there's going to be potentially this kind of side-effect. So that what I gave him was instructions. The patients would go say, for example, on a Monday and would get their blood count, and if their blood count was below four thousand white count or the platelet count was less than a hundred thousand, any of the oncology nurses would then contact him. Then he would arrange to adjust the dose that he had with him how to adjust the dose. [sic] If the patient got too sick he was supposed to see them. So all this arrangement was made with another physician for both outpatient and for the office.

Q. All right. Now, in the office context for the coverage did you inform your personnel who the physician was that would be covering?

A. Yes. I mean, obviously there are people that are getting -- you're gone for twenty days. There are patients who are going to get a heart attack and the patient is going to start bleeding. The nurse has to know who to contact, sure.

Q. And in the in-hospital setting, the physician -- and the physician was who?

A. Dr. Mufti.

Q. Dr. Mufti to the best of your knowledge, did he also have hospital privileges at St. Joe's?

A. Yes.

Q. And did he make daily rounds as you have described the rounds you made so far as you're aware?

A. Yes.

Q. Was that at least your intention and your belief before you left?

A. Yes.

Siddiqi was not cross-examined by the government on this phase of his testimony. Siddiqi did not present any other evidence of Mufti's covering for him. Mufti came to the trial and was ready to testify, but Siddiqi's trial counsel did not call him. As we noted on the first appeal, this was "not surprising," because the prosecutor, while cross-examining defense witness Martin Neltner, an accountant who was a medical billing consultant, apparently conceded that Siddiqi had arranged for Mufti to cover his patients in his absence. The prosecutor said, "With respect to coverage -- I don't think anybody is questioning that coverage exists in this trial, but my question relates to the method by which coverage is standardly reported in the hospital." Having stated that no one was "questioning that coverage exists in this trial" -- a remark entirely consistent with the information available to the government and to the defense -- the prosecutor next pursued hypotheticals concerning how the identity of a covering doctor might be made known to the nurse who was to administer the chemotherapy. The following colloquy ensued:

Q. Isn't it fair to say that whenever a doctor covers for another doctor, that doctor that is covering reports in at some time to the nurse oncologist?

A. When you say "reports in" --

Q. Has some sort of contact.

A. Has some contact. Could be by phone, could be in person.

Q. Anything.

A. That's correct.

Q. Some contact.

A. A lot of times the physician will rely on the nurse to call him --

Q. Sure.

A. -- when there is a problem.

Q. Sure. So if a particular doctor - let's do it as a hypothetical. If a particular doctor is standing in for Dr. Siddiqi, and in that particular situation the nurse is supposed to reach out to the covering doctor, one would expect to see the notes on the bottom of the hospital records "If a problem call covering doctor," correct?

A. Yes.

Q. Okay.

A. I would like to see that.Q. And it wouldn't be --

A. I would like to see that.

Q. You would like to see that?

A. Yes.

Q. If a problem call Dr. Siddiqi, but it would -- if a problem call the covering doctor, correct?

A. Yes.

Q. Okay. And in that light the professional component presumes that there's a service being rendered --

A. Yes.

Q. -- isn't that correct?

A. Yes.

Q. That's a given.

A. Yes.

Q. That we're not billing for services not rendered.

A. That is correct.

The questions concerning notations as to the identity of a covering doctor, posed as hypotheticals to an accountant, were the sole inquiries by the prosecutor regarding coverage during trial. However, they were sandwiched between a concession that no one was "questioning that coverage exists in this trial" and testimony at the end that seemed to confirm compensability where coverage was provided. The prosecutor's reference to a hypothetical note, "If a problem, call Dr. Siddiqi," must, in the context of what turned out to be a highly misleading concession and of otherwise helpful testimony, have seemed unimportant, although it was later the basis for a crucial part of the prosecution's summation concerning Arthur Kenny's outpatient document, as set out below.

The government's case on the Mecca counts at the close of evidence was at best confusing. Double billing had not been shown. Niegsch had no opinion on what "supervision" meant. Baker had testified that oncologists did not attend the administration of prescribed chemotherapy but were available if needed, or were covered by another doctor. Weissman had testified to disagreement over whether being available was compensable. With regard to whether Siddiqi had actually arranged for coverage, Baker had stated that Siddiqi usually had a doctor cover when he was away. Siddiqi testified that Mufti covered for him at the time of the Mecca counts, and the government seemingly conceded that coverage in cross-examining the accountant. Nevertheless, the cross-examination of the accountant, viewed in hindsight, indicates that the prosecutor had arrived at a strategy to salvage the case.

In summation, the prosecutor expressly repudiated the double-billing theory that had loomed so large in his opening. He stated: "There's no double billing. We're not alleging a double billing. We're alleging billing for services not rendered. Pure and sweet, that's it." However, he also reversed the government's position that no one was "questioning that coverage exists":

The coverage issue. The doctor's out of the country. Dr. Mufti's going to look after my patients. There's no reference to Dr. Mufti anywhere. Don't trust me. Take a look at it. Specifically take a look at Arthur Kenny. [Apparently, Arthur Kenny's outpatient document.] Arthur Kenny is in the indictment July 27th, '88, July 29th, '88. The reference is not "Call Dr. Mufti," and if you recall, I asked Mr. Neltner, "Shouldn't it say 'Call Dr. Mufti'?" The coverage -- if the covering doctor's there you should say "Call Dr. Mufti," and if it didn't there probably wasn't any coverage. The outpatient document says call Dr. Siddiqi if any problems. Well, old Arthur Kenny isn't going to get too quick a response if the doctor is out of the country at the time, is he? Take a look at it.

The summation was a success. As we previously noted, see United States v. Siddiqi, 959 F.2d 1167, 1171 (2d Cir. 1992), the issue in the jury's mind was evidently whether Siddiqi told the truth when he said he told the hospital that Mufti would be covering for him while he was away. It determined that he did not and convicted on the Mecca counts.

4. Post-Trial Proceedings

The defense filed a post-trial motion for a judgment of acquittal notwithstanding the verdict, see Fed. R. Crim. P. 29(c), which the district court denied. Siddiqi was sentenced to three years' probation, 1000 hours of community service as an emergency room physician, a fine of $2000.00, $ 640.88 in restitution, and a $ 250.00 special assessment. On July 15, 1991, Siddiqi

filed a notice of appeal.

On August 5, he filed a motion in the district court for a new trial based upon newly discovered evidence. See Fed. R. Crim. P. 33. The evidence was a document from St. Joseph's Hospital known as the "physicians away list." This list consisted of the names of physicians who would be out of town on certain dates and the names of other physicians who would be covering for them. It reflected that Dr. Mufti was covering for Dr. Siddiqi on the dates involved in the Mecca counts.

In response, the government argued that "the 'covering physician' [Mufti] was nowhere to be seen or heard from [at trial]" and ventured that "what was newly discovered was not the list itself but the potential relevance of the list to [Siddiqi's] defense." However, the government muddied the waters by stating that its consistent position was that "the defendant should not have billed for services he did not provide" and "billed for services he allegedly performed while he was outside of the United States on travel." These latter remarks strongly suggested that, regardless of coverage by Mufti, Siddiqi was guilty because he did not personally perform services regarding the administration of chemotherapy while he was in Mecca. The district court denied Siddiqi's motion, concluding it was "highly unlikely" that the introduction of the list at trial would have led to a different result. Siddiqi filed a second notice of appeal. The appeals from the conviction and from the denial of the Rule 33 motion were consolidated.

5. The First Appeal

Siddiqi raised three claims on the first appeal: (i) that the evidence was insufficient to support his conviction on the five Mecca counts; (ii) that the district court erroneously failed to grant a new trial on the basis of the newly discovered physicians away list; and (iii) that the district court erred in failing to give two of Siddiqi's proffered jury instructions. The arguments supporting each of these claims emphasized the shifting nature of the government's theories of the case.

The government's brief denied any inconsistency in position but hardly settled upon a single, clear theory of Siddiqi's guilt. Although the conviction on the Mecca counts was undoubtedly based on the government's argument -- and the jury's conclusion -- that Siddiqi had not arranged with Mufti for coverage, the government's array of theories on appeal barely alluded to that factual issue. For example, the government's brief summarized the relevant charges as alleging that Siddiqi "billed Medicare as though he administered the chemotherapy." The focus on the actual administration of the injections -- as opposed to the "supervision" of them -- was seemingly a resurrection of the asserted-in-the-opening, abandoned-in-the-summation double-billing theory. The brief also claimed the evidence at trial "showed that the appellant billed Medicare for chemotherapy administered to two of his patients during a two-week period of time in which appellant was out of the country." It further asserted: "The verdict of the jury reflects their [sic] collective opinion that the appellant billed Medicare for chemotherapy he claimed he administered at a time when it is undisputed he did not." However, the government also alluded to other theories: "the decision of the jury as to the five counts of conviction was an implicit rejection of the dual defense claims that he was entitled to bill for Saint Joseph's chemotherapy since he was ultimately responsible and that another doctor covered for him while he was away."

The only substantive discussion in the government's brief of the factual issue as to whether Siddiqi had arranged for coverage by Mufti was a single paragraph in the discussion of the sufficiency of the evidence. It stated: The appellant further stresses . . . that Mary Anne Baker clearly established that the appellant was available at the hospital while his patients were receiving chemotherapy and that he provided coverage for his patients at the hospital while he was away. Nowhere in the record of Ms. Baker's testimony is there an acknowledgement that Ms. Baker could have contacted the appellant while he was on vacation or that he provided coverage while he was out of the country. To the contrary, based upon Ms. Baker's own notation, she was operating under the obviously incorrect assumption that while Arthur Kenny was receiving chemotherapy on July 25, 1988, she could have contacted the appellant if any problems occurred. GA 34. [A cite to the Arthur Kenny outpatient document]

This was a misleading description of the evidence in that Baker did in fact testify that Siddiqi was either on call or provided coverage "as necessary." Moreover, there was no evidence at trial or elsewhere that she authored the notation "Call Dr. Siddiqi if any problems" on the Kenny outpatient document. We now know that she did not.

At oral argument, the prosecutor again offered various theories. He stated, "The doctor did not administer the chemotherapy . . . . He did not perform the service, he billed Medicare for the administration of the chemotherapy." The crime, he said, was "billing for something he did not do." These statements again appeared to invoke the double-billing theory or at least a theory under which Mufti's coverage or non-coverage and Siddiqi's whereabouts were irrelevant. Later, the prosecutor emphasized the need for Siddiqi's personal involvement in the administration of the drugs: "if [Siddiqi] wanted to get credit and bill for that service . . . he should have supervised the administration of the chemo." Still later, however, he asserted "there is no evidence that Mufti came in and was covering the doctor." After that, a new theory emerged: "Dr. Mufti may have come at some particular time later on in the day after the administration was done and looked at the chart, the hospital didn't know a Dr. Mufti" was covering for Dr. Siddiqi. This theory, which played no role at trial, seemed to posit that even if Mufti was available, coverage was irrelevant if it was not communicated to the hospital. It was also quite at odds with the government's position that the physicians away list would not have affected the verdict even though it showed that notice of Mufti's coverage was given. In response to questions by the court concerning whether Siddiqi had arranged for coverage by Mufti, the government emphasized the notation on the Kenny outpatient chart, "call Dr. Siddiqi if any problems."

In our decision, we rejected the jury instruction claim and concluded that "there was, barely, sufficient evidence to support Siddiqi's convictions." This was a reference to the Kenny outpatient document, which, as interpreted by the government, tended to show that Siddiqi had not arranged for coverage. However, we held that the district court erred in rejecting the newly discovered evidence claim, United States v. Siddiqi, 959 F.2d 1167, 1172 (2d Cir. 1992), and remanded for proceedings to determine whether the physicians away list could have been discovered prior to or during trial through the exercise of due diligence and, if not, whether the new evidence was admissible. Id. The question regarding admissibility was in response to the government's questioning of the authenticity of the list and of its admissibility under the hearsay rule. Fed. R. Evid. @@ 801 & 802. We noted that, if available, the list would "probably" have caused the jury to acquit. Our decision contained a critical ambiguity. In stating that the list would "probably" have led to an acquittal, we hedged because the government was challenging the authenticity of the list, creating the possibility that, even if admitted, the list might have been found bogus by the jury. Left unsaid was our view that if Siddiqi had arranged for coverage, then he could not have been validly convicted. n1

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n1 1. The author of this opinion is the only member of the original panel now serving as a judge. It is his belief now, and was his understanding then, that the decision of the panel upholding the sufficiency of the evidence was based entirely on a belief that a genuine factual dispute existed over whether Siddiqi had arranged for coverage by Mufti. Absent such a dispute -- that is, if no one had disputed that there was coverage -- we would have reversed.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

6. Proceedings After Remand

A hearing was held in the district court on remand. At that hearing, Baker was called by Siddiqi. As noted, Baker had declined even to talk with defense counsel before trial. After the conviction, however, she had a change of heart and informed defense counsel that Siddiqi had told her that he planned to go to Mecca and that Mufti would cover for him during that time. More significantly, another nurse, Cheryl Hollenbeck, testified that she was the author of the "call Dr. Siddiqi if any problems" notation on Arthur Kenny's outpatient document. As recited above, the prosecutor had read this notation in summation to the jury as proof that there was no coverage by Dr. Mufti and stated in his brief to us that it was written by Baker. Hollenbeck testified, however, that the notation was made by her after the chemotherapy had been administered and when Kenny was being discharged. This was on July 29, after Siddiqi's return from Mecca. The notation was not, according to Hollenbeck, an instruction to the oncology nurse but to Kenny, who was to call Dr. Siddiqi if he had problems. Finally, the physician's away list was authenticated. In short, the factual basis for Siddiqi's conviction -- that he had not arranged for coverage by Mufti -- evaporated.

Staying within the confines of our remand, the district judge interpreted our decision as directing inquiry into whether the physicians away list "could" -- not "should" -- have been discovered before trial. He concluded that it could have been produced in response to a properly framed subpoena and therefore denied the motion for a new trial. Siddiqi again appealed to this court. Reviewing the district court's ruling for abuse of discretion, we affirmed the denial of a new trial because the lower court found, as a matter of fact, that Siddiqi had failed to meet the due diligence requirement. United States v. Siddiqi,986 F.2d 500 (2d Cir. 1992) (unpublished summary order).

We note, however, that the government's brief on this second appeal stated "the theory of the Government's case" as follows: The prosecution in this case alleged that the defendant would personally claim for chemotherapy administered by the staff of St. Joseph's Hospital in Elmira, New York. The evidence at trial showed that the appellant claimed for services performed by the staff at the Oncology Department.

The failure to arrange coverage theory having been shredded, the double-billing theory was alive and well.

7. The Section 2255 Petition

Siddiqi thereafter retained new counsel and on July 2, 1994 filed the instant proceeding to vacate his conviction pursuant to 28 U.S.C. @ 2255. He claimed that his trial counsel was constitutionally ineffective in failing to discover and introduce the quite substantial evidence available to support the claim that Mufti had covered for him.

In response, the government asserted that "even if trial counsel [discovered the physicians away list and elicited witness testimony about coverage], it would not have altered the outcome." In explaining this conclusion, the government entirely abandoned the theory on which the jury convicted: that Siddiqi never arranged for Mufti to cover for him. The government's papers stated: The defendant was convicted for billing Medicare for services that were, in fact, never performed. . . . The government alleged that the defendant's 96500 billings were fraudulent, because the defendant had not performed any services in connection with the chemotherapy sessions, since all of the injections referenced in the indictment were administered by the Hospital outside of the defendant's presence.

In refuting Siddiqi's assertion that he performed a cognitive professional service in diagnosing, prescribing, and monitoring the progress of chemotherapy treatments, the government contended: The defendant did not personally perform professional cognitive services with respect to the injections at issue in the five counts of the indictment for which he was ultimately convicted. Those counts arose out of 96500 billings submitted for . . . injections given to patients while the defendant was out of the country.

Now conceding that Siddiqi arranged for Mufti to cover for him while he was in Mecca, the government introduced an entirely new argument: Although Dr. Mufti agreed to be available in the event that problems arose during treatments, no problems actually arose during treatments. . . . Thus, Dr. Mufti did not perform any services whatsoever in connection with the chemotherapy injections that occurred during the defendant's absence.

The new theory thus conceded the fact of coverage and the hospital's awareness of it but argued that Mufti's coverage was irrelevant because Mufti never performed services. Admitting that "the Hospital's knowledge of coverage arguable [sic] became an issue any [sic] closing argument when . . . the issue was raised by the government for the first time," the government contended that "the Hospital's awareness of Mufti's availability does not alter the simple fact that Dr. Mufti did not perform any services and, consequently, the defendant's bills were fraudulent." The government's brief concluded: No physician performed a professional service in conjunction with the second and subsequent chemotherapy injections administered to the defendant's patients while he was out of the country. . . . Dr. Mufti was merely "on call" in the event an emergency occurred. Simply being available, if needed, is not a professional service reimbursable by Medicare. . . . The defendant billed Medicare for services that were, in fact, never performed. That is a crime for which he was properly convicted.

We believe it fair to say that this theory of guilt was never before put to the jury, to the district court, or to this court.

The district court denied Siddiqi's Section 2255 petition, holding that Siddiqi failed to demonstrate constitutionally ineffective assistance of counsel because he did not establish that his lawyer's actions "fell below the prevailing professional norms," and that "a reasonable probability" existed that "absent [counsel's] unprofessional performance, the outcome would have been different." Siddiqi v. United States, No. 94-CV-6323T (W.D.N.Y. Feb. 21, 1995) (quoting United States v. Vegas, 27 F.3d 773, 777 (2d Cir.), cert. denied, 130 L. Ed. 2d 200, 115 S. Ct. 284 (1994) (citing Strickland v. Washington, 466 U.S. 668, 80 L. Ed. 2d 674, 104 S. Ct. 2052 (1984)).

With regard to the conduct of defense counsel, the district court found that it was reasonable for him to believe that further evidence corroborating Siddiqi's testimony as to coverage by Mufti was not necessary. Cross-examination of Baker regarding coverage was self-evidently dangerous in light of her refusal even to talk with defense counsel and his consequent lack of knowledge as to what answers she might give. The court found that it was reasonable for counsel to believe that the government had conceded the existence of coverage in light of the prosecutor's remark "nobody is questioning that coverage exists." Finally, the court noted counsel's concerns over the facts that Mufti had performed no services and that, had such services been necessary, Mufti would have billed Medicare himself.

With regard to whether the evidence in question -- the physicians away list and testimony by the nurses -- would have affected the outcome, the district court found that it would not have.

On this appeal the government's brief largely restates the contentions and theories it set forth in the district court, stressing that "simply being available, if needed, is not a professional service reimbursable by Medicare." However, at no time has the government ever provided a citation to any rule, regulation, case, or statute stating that being available was not compensable at pertinent times.

8. Ineffective Assistance of Counsel

Siddiqi argues that he received constitutionally ineffective assistance of counsel at trial. In Siddiqi's view, our prior decisions dictate that his conviction be vacated on this ground.

We noted in our first decision that introduction of the physicians away list at trial "would probably have led the jury to acquit Siddiqi on all five counts." Siddiqi, 959 F.2d at 1174. Then we agreed with the denial of a new trial on the ground that the physicians away list and other evidence of coverage could have been found prior to or during trial if the defense had exercised due diligence. United States v. Siddiqi, 986 F.2d 500 (2d Cir. 1992) (unpublished summary order). Because of the lack of due diligence, Siddiqi reasons, evidence that would "probably" have led to an acquittal was not obtained and introduced. In his view, his counsel's conduct fulfills both prongs of the test for ineffective assistance of counsel. It "fell below the prevailing professional norms" and prejudiced him because there is "a reasonable probability that absent counsel's unprofessional performance" he would have been acquitted on the five counts of conviction. United States v. Vegas, 27 F.3d 773, 777 (2d Cir.) (citing Strickland v. Washington, 466 U.S. 668, 80 L. Ed. 2d 674, 104 S. Ct. 2052 (1984)), cert. denied, 115 S. Ct. 284, 130 L. Ed. 2d 200 (1994). For reasons stated below, we have no doubt that Siddiqi meets the prejudice prong of the Strickland test.

However, we find it difficult to fault defense counsel's performance, but for reasons somewhat different from those relied upon by the district court. The government's theory of criminal conduct has been a target that moves opportunistically when confronted by contrary evidence or telling argument. The government has proffered at least four basic theories:

(A) Designation of code 96500 was a statement by Siddiqi that he administered (and, sometimes, also provided) the chemotherapy; because the hospital administered and provided the drugs and billed Medicare, Siddiqi is guilty of fraud.

(B) Designation of code 96500 was a statement that Siddiqi personally "supervised" administration of the chemotherapy; because he was in Mecca at the time, the statement is a fraud.

(C) Designation of code 96500 was a statement that Siddiqi supervised administration of the chemotherapy by arranging for Mufti to cover; because Siddiqi never arranged for Mufti to cover, Siddiqi is guilty of fraud. n2 (This was the theory on which Siddiqi was convicted.)

(D) Designation of code 96500 was a fraud because, even though Mufti actuallycovered for Siddiqi, being on call is not compensable unless actual services were needed and provided. (This is the government's present theory.)

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n2 There is a subtheory here: Siddiqi may have arranged for coverage by Mufti but neither one told the hospital. Although this theory made a cameo appearance on the first appeal, it did not, and would not, benefit from elaboration.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

As to the Mecca counts, we can in retrospect say that the trial was an ambush. The defense's selection and presentation of evidence was governed by the two government theories pursued prior to summation, (A) and (B). Under (A), coverage was irrelevant. Under (B), coverage by a doctor available if needed seemed a plausible defense, particularly after the government conceded that coverage existed. The government's summation, of course, ignored that concession by adopting (C), the theory on which the jury convicted. Theory (C) now having been entirely undermined, the government now pursues theory (D). This final theory, however, is legally inadequate for reasons stated below.

Until the summation, defense counsel had no reason to think that the government harbored any doubts about whether Siddiqi had arranged for Mufti to cover. Baker, who was anything but a friendly witness to Siddiqi, had testified that Siddiqi arranged for coverage "as necessary." The government showed no sign of disputing her testimony. The government also did nothing in its cross-examination of Siddiqi that indicated a dispute over whether Siddiqi had arranged with Mufti for coverage. Nor did the government call Mufti. The cross-examination of Neltner -an accountant -- touched upon whether someone administering the chemotherapy should know the identity of the covering doctor. The question related to whether "one would expect to see the notes on the bottom of the hospital records 'If a problem, call covering doctor,'correct?" Another, "If a problem, call Dr. Siddiqi, but it would -- if a problem call the covering doctor, correct?" -- set up the summation's reference to Arthur Kenny's outpatient document. However, at the time the significance of these questions was anything but self-evident. Lots of documents were in evidence and the notation on the Kenny outpatient document had not been explored. Moreover, the questions were posed as hypotheticals rather than as specific references to a particular exhibit or particular patient. Most critically, these questions were prefaced by the concession that no one was "questioning that coverage exists in this trial." Finally, the cross-examination of Neltner ended on an ambiguous note seeming to suggest that where coverage exists, billing is proper. We fail to see how it was unreasonable for defense counsel not to have anticipated the single time that the government pursued theory (C).

9. Miscarriage of Justice

That is not the end of the matter, however. Siddiqi argues that he is caught in a Catch-22. We agree. He was convicted on a then newly-minted, now-abandoned, theory that he defrauded the government because he had not arranged coverage by Mufti. When, after the conviction, he presented evidence that proved coverage sufficiently convincing for the government now to concede it, he was denied relief on the ground that the evidence would have been available before trial had due diligence been exercised. In response to his Section 2255 petition claiming that the lack of due diligence constituted ineffective assistance of counsel, the government argues that he received effective assistance because, although there was coverage by Mufti, coverage is irrelevant. It takes no great flash of genius to conclude that something is wrong somewhere. We find the ointment's fly in the failure of the government to settle upon a single theory of guilt and, then, at this final stage, in seizing upon a theory that is inadequate as a matter of law.

Collateral review is available under Section 2255 where "the alleged error constituted a fundamental defect which inherently results in a complete miscarriage of justice." Reed v. Farley, 512 U.S. 339, 114 S. Ct. 2291, 2300, 129 L. Ed. 2d 277 (1994) (internal quotation marks omitted). We are firmly convinced that Siddiqi's trial and post-trial proceedings did not meet rudimentary demands of fair procedure, and that the defects in his conviction are so fundamental as to constitute a miscarriage of justice. We therefore order that his conviction be vacated.

The government's theories have varied as necessary to fill the gaps in its own case, to deflect the ripostes of the defense, and to provide answers to questions from the bench. The shifting nature of the government's case -- which we criticized in our first opinion -- materially impeded the effective presentation of a defense during the trial by misleading defense counsel as to the government's theory of guilt. Siddiqi, 959 F.2d at 1172. This enabled the government to obtain a conviction from the jury on a factual theory that was put forth for the first time in summation and had no foundation in fact.

We too have had our consideration of this matter impaired by the shifting theories of the government. On the first appeal, the government's brief relied principally upon theories (A) and (B) and alluded only briefly to theory (C), which was the basis of the jury's conviction. The allusion was, as described above, rather misleading as to Baker's testimony and the authorship of the Kenny outpatient document. The downplaying of theory (C) avoided, of course, the danger (to the government) of magnifying the importance of the physicians away list.

In our decision on the first appeal, we concluded that the conviction was based on the jury's factual determination that Siddiqi had not arranged for coverage by Mufti. We viewed that determination as based on "barely sufficient" evidence, id. at 1172, and stated that the jury "probably" would have acquitted had the physicians away list been in evidence, Siddiqi, 959 F.2d at 1174. Our statement was hedged because the government was at that time questioning the authenticity and admissibility of the physicians away list, and we were unsure whether it was admissible or genuine. Id. We believed in short, that a genuine dispute of fact existed over coverage by Mufti. We had no idea that the government never had any basis for disputing the fact of coverage.

Unsaid, except by implication, was our view that, if Siddiqi had in fact arranged for coverage, the government's evidence would have been held to be insufficient as a matter of law. This implication would have been better spelled out. Instead, an ambiguity was left suggesting that the jury might or might not have acquitted even though it found there was coverage and that either result was legally sustainable. That was not intended.

As a result, Siddiqi's second appeal was narrowly focused on the due diligence issue, and we again had no reason to perceive the total lack of any evidentiary foundation for the theory on which the conviction had been obtained. The affirmance of the finding as to the lack of due diligence led, of course, to the present ineffective-assistance claim. At this stage, the government does not claim that Siddiqi had not arranged for coverage. The physicians away list and Baker's testimony show that he had. Hollenbeck's explanation of the Kenny outpatient document further disposed of any doubt. Now, faced with a need to justify the failure of defense counsel to present the virtually irrebuttable evidence of coverage and to show that such evidence could not have affected the outcome, the government has turned to theory (D) -- coverage existed but is irrelevant because being available without more is not billable under code 96500. Had we been faced with this position on the first appeal, we would have reversed.

We now have no choice but to hold the government to theory (D). Theory (A) had no basis in any of the evidence and should never have been proffered. Theory (C) is now known also to be baseless. Theory (B) lives on, but only if Theory (D) is correct in asserting that simply being available, whether covered or not, is not compensable.

Theory (D) being inadequate as a matter of law, we order that the conviction be vacated. We emphasize that this is not a civil billing case; it is a criminal fraud case. Each of the Mecca counts required proof that Siddiqi used code 96500 with a dishonest intent. Based on the present record, inference of such an intent cannot be drawn from use of the code. As noted, code 96500 allows billing for "supervision," a term that is, on this record, unclear. The government's principal expert on this issue was unable to provide a definition of supervision, and the government cannot be allowed to prevail on the claim that it is fraud for Siddiqi not to have anticipated the definition embodied in theory (D).

Professionals, including lawyers, sometimes bill for simply being available if needed. One may disagree with the practice because the professional may be doing other billable work at the time. However, billing is nevertheless arguably appropriate because to be available, the professional must forego the opportunity to go fishing or worship in Mecca at the same time. Absent some affirmative reason to believe that use of code 96500 does not cover being available, billing under that code is at worst an attempt to bill at the outer limits permitted, not fraud. See United States v. D'Amato, 39 F.3d 1249, 1261 (2d Cir. 1994).

The government has been unable to cite to us any authority whatsoever that would have alerted someone in Siddiqi's position that being available if needed is excluded as a compensable service under code 96500. Indeed, no one who has read the trial record in this case could draw that conclusion. It may well be that arranging for coverage -- a common practice among physicians -- is billable under code 96500 so long as the covering doctor also did not bill for being available -- and Mufti did not. We cannot say on this record, therefore, that an arrangement in which the primary doctor bills Medicare for an available covering doctor was fraud at the pertinent times. This practice may not have met Judge Cardozo's standard of "the punctilio of an honor the most sensitive," Meinhard v. Salmon, 249 N.Y. 458, 464, 164 N.E. 545 (1928), but it does not constitute fraud.

We note as to the coverage issue that the government bears the burden of proof on all elements of the crime. Some of the arguments made by the government seem to suggest that Siddiqi had the burden of proving coverage. He did not. As noted, coverage is a common practice among physicians and other professionals. The government had to prove fraud. With respect to the Mecca counts, it had to prove that Siddiqi did nothing, including failing to arrange for coverage, and yet billed under code 96500.

We are, therefore, faced with a situation in which a conviction was obtained on a theory now shown to lack a factual basis and presently defended by a theory that is legally insufficient. We are firmly convinced, moreover, that the conviction would not have occurred but for the government's shifting of theories that impaired both Siddiqi's defense and our consideration and disposition of this matter. Indeed, had the government's present theory been candidly stated on the first appeal, we would have reversed outright. See Note 1, supra.

Had Siddiqi been convicted on a legal theory that was invalidated by a subsequent decision of the Supreme Court, we would vacate it under Section 2255. See Ingber v. Enzor, 841 F.2d 450, 453-54 (2d Cir. 1988). Here we face a situation in which subsequent events compellingly demonstrate that a conviction had no legitimate factual or legal basis and that, but for the conduct of the prosecution in adopting shifting and at times misleading positions, no conviction would have been obtained or successfully defended on appeal. n3 We see no reason why a petitioner in Siddiqi's position should fare worse than one whose conviction was based on a legal theory invalidated by later caselaw.

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

n3 In view of this conclusion, we need not decide whether the conduct in question was deliberate.

- - - - - - - - - - - - - - - - -End Footnotes- - - - - - - - - - - - - - - - -

We are firmly convinced that Siddiqi's trial and post-trial proceedings did not meet rudimentary demands of fair procedure, see Reed v. Farley, 512 U.S. 339, 114 S. Ct. 2291, 2299-300, 129 L. Ed. 2d 277 (1994), and that the defects in his conviction are so fundamental as to constitute a miscarriage of justice, id. We therefore order that his conviction be vacated.