Skip to main content


Reset A A Font size: Print

United States Court of Appeals, First Circuit.

UNITED STATES of America, Appellee, v. Anibal PAGÁN-ROMERO, Defendant, Appellant.

No. 16-1396

Decided: July 09, 2018

Before Lynch and Lipez, Circuit Judges, and Ponsor, District Judge.* Paul M. Glickman and Glickman Turley LLP, Boston, MA, were on brief, for appellant. Francisco A. Besosa-Martínez, with whom Rosa Emilia Rodríguez-Vélez, United States Attorney, and Mariana E. Bauzá-Almonte, Assistant United States Attorney, Chief, Appellate Division, were on brief, for appellee.

For many years, Appellant, Dr. Anibal Pagán-Romero, operated a medical clinic in Quebradillas, Puerto Rico. On October 5, 2015, a jury found him guilty of conspiracy to commit mail fraud and substantive mail fraud, based upon his certification of false injury claims submitted to the American Family Life Insurance Company (“AFLAC”).

This appeal arises from the district court’s decision to grant the jury’s oral request, made during deliberations, for a dictionary. As will be seen below, this decision, made over defendant’s objection and with no discussion on the record, was improper. A review of the record, however, reveals that the trial judge took thorough, effective action to investigate the impact of the error and properly concluded that Appellant suffered no prejudice. We therefore conclude that the judge did not abuse his discretion in denying Appellant’s motion for a new trial. Based on this, we will affirm.

I. Background

Appellant owned the Policlínica Familiar Shalom, a medical clinic and pharmacy in Quebradillas, Puerto Rico, where he also practiced medicine. On May 8, 2014, Appellant was charged, along with thirty-five co-defendants, with twenty-one counts of conspiracy to commit mail fraud in violation of 18 U.S.C. §§ 1349 and 1341, and sixty-one counts of mail fraud in violation of 18 U.S.C. §§ 2 and 1341. The indictment alleged that between January 2004 and November 2009, Appellant conspired with others to defraud AFLAC by filing false claims under its accidental injury policies.

On August 31, 2015, the case went to trial before a jury. The government’s theory was straightforward: before paying a claim, AFLAC required certification from a physician that he had, in fact, provided treatment on a specific date for a particular medical condition. Appellant, the government contended, had falsely certified claims over many years without ever treating, often without even seeing, the claimants. Former employees of the clinic testified pursuant to plea agreements and confirmed the extent of the fraud, admitting that they had overseen the submission of false claims and had even submitted bogus claims, certified by Appellant, on behalf of themselves and family members.

Eventually, the scope of the fraud reached such proportions that some of Appellant’s staff became uncomfortable, and Appellant directed AFLAC claimants to come through a side entrance of the clinic and work directly with co-conspirators closer to the heart of the fraudulent operation. Paperwork related to the AFLAC claimants was filed separately by Appellant and his co-conspirators; the claimants’ files routinely contained no progress notes or other routine medical documentation, only the claim forms. Certification by Appellant of treatment supposedly given to these claimants was sometimes noted as occurring on dates when Appellant was out of the country, or on Saturdays and Sundays when the clinic was closed. Testimony confirmed that Appellant received $10 to $20 for each falsified claim. The government’s evidence included over 270 exhibits, including audio recordings in which Appellant was heard speaking to two undercover FBI agents about claim forms he certified using false information.

At trial, Appellant’s defense was that he was unaware of the fraudulent scheme, which he contended was perpetrated without his knowledge by employees who stole his signature. Appellant’s nephew Noel Pagán-Rivera testified that he had filed fraudulent claims at his uncle’s clinic without the latter’s knowledge. Appellant himself testified, denying any wrongdoing, asserting that some of the fraudulent claim forms had been filled out by a person or persons unknown to him, and asserting that he did not knowingly participate in any scheme to defraud.

On September 30, 2015, counsel rested. The following day, the jury heard closing argument and instructions from the court and began deliberations.

The jury instructions made clear that an essential element of mail fraud was that “Anibal Pagán-Romero knowingly and willingly participated in this scheme with the intent to defraud.” 1 The instructions expanded on this point by stating that “Anibal Pagán-Romero acted knowingly if he was conscious and aware of his actions, realized what he was doing or what was happening around him and did not act because of ignorance, mistake, or accident.” To determine Appellant’s state of mind, the instructions stated:

[Y]ou may consider any statement made or acts done or omitted by him and all other facts and circumstances received in evidence that may aid you in determination of Anibal Pagán-Romero’s knowledge or intent. ․ You may infer, but you are certainly not required to infer, that a person intends the natural and probable consequences of his acts knowingly done or knowingly admitted.

These instructions were perfectly correct, and Appellant does not argue otherwise. The trial judge also instructed the jurors, correctly, that they were not to do any outside research of their own over the course of deliberations. Significantly, the trial judge included in his instructions the standard admonition that communications with the judge or his staff needed to be put in writing.

The day after the jury began deliberating, October 2, 2015, the trial judge advised the jurors that he needed to be absent for one week and allowed them to choose whether to continue their deliberations during his absence with another judge supervising, or suspend until he came back. The jurors chose to continue their deliberations in the trial judge’s absence, and another judge made himself available to supervise the deliberations as needed.

Deliberations went on for two more days, and on the third day, October 5, the jury convened to continue its work at approximately 9:45 a.m. The record of what happened next is not clear. At some point, an oral request somehow emerged from the jury for an English-English dictionary. The record does not identify from whom the request came (the foreperson or some other member of the jury) or precisely to whom the request was directed or the time. The docket merely states: “Parties were informed of oral communication request from the jury with CSO asking for an English-English dictionary and a laptop, objection from the defense, as to the dictionary, was denied.” The request was not in writing, contrary to the trial judge’s instruction, and no transcript reciting exactly how the objection was articulated, or how it was denied, appears on the record. Whatever the process, the upshot was that some time before 12:55 p.m. on October 5, 2015, an English-English dictionary made its way into the deliberation room. At 12:55 p.m., a note was received from the jury to the effect that it had reached a verdict. Appellant was found guilty on all counts.

Following the conviction, Appellant moved for a new trial, citing the improper provision of the dictionary. On February 23, 2016, the original trial judge convened the first evidentiary hearing on the motion. The court at this time questioned the foreperson, who recalled that one juror had used the dictionary, although he could not remember which juror this had been or what the juror had used the dictionary for. He did recall that the requesting juror had been a woman. He could not recollect whether anyone read aloud from the dictionary. Asked whether the dictionary affected the jury’s deliberations, the foreperson answered: “I don’t think so.”

On March 14, 2016, the court held a second hearing on the possible impact of the dictionary. This time, the four female members of the jury appeared, and the judge questioned them. The juror who requested the dictionary reported that she had used it during deliberations to look up the word “knowingly.” She stated that the dictionary had not influenced her deliberations, and that the entire panel of jurors had discussed the dictionary’s definition. The second juror recalled that the dictionary definition of “knowingly” was read aloud to the entire jury from the dictionary. She stated, however, that it did not influence the deliberations. The third juror stated that the definition had “helped us out.” She added: “It was just a few persons that were confused with ․ what was the meaning of knowingly.” This juror indicated that the jury had made its decision based solely on the evidence and the jury instructions, not on the dictionary definition. A fourth juror recalled using the dictionary to look up the word “knowingly,” but stated that this did not influence jury deliberations.

On March 21, 2016, the district court explored the dictionary issue a third time, on this occasion questioning all twelve trial jurors individually. Based on this questioning, the court found that the jurors’ answers were consistent as to the following: (1) a member of the jury had requested a dictionary; (2) the dictionary was used to look up the word “knowingly”; (3) the dictionary definition of the word “knowingly” did not influence jury deliberations; (4) the jury followed and relied on the district court’s jury instructions; and (5) the verdict was based solely on the evidence and the district court’s jury instructions. Relying on these conclusions, the court ruled that “the use of the dictionary to look up the word ‘knowingly’ in no way affected or brought in any extraneous evidence or information to the jury which would affect their deliberations.” Further, the dictionary definition of “knowingly” was not inconsistent with, and in no way undermined, the definition set forth in the jury instructions. In fact, the trial judge observed that the dictionary definition, which defined “knowingly” both as “having knowledge” and as being “shrewd, clever, implying a secret understanding,” would, if considered by the jury, have imposed a burden on the government “that [went] even beyond the jury instruction.” The dictionary’s definition, the court concluded, was not to Appellant’s detriment, and, if anything, would have benefitted him. Based on this, the court concluded that, even assuming the provision of the dictionary was an error, the mistake was harmless. The motion for new trial was denied.

Ultimately, Appellant received concurrent sentences of 120 months custody of the Bureau of Prisons and five years of supervised release on each count, as well as restitution in the amount of $2,056,303.

II. Discussion

In response to a nonfrivolous claim that a jury might have been influenced by improper exposure to extrinsic material, a district court must conduct an inquiry into whether the exposure in fact occurred and, if so, whether it was prejudicial. United States v. Camacho-Santiago, 851 F.3d 81, 89 (1st Cir. 2017), cert. denied, ––– U.S. ––––, 138 S.Ct. 223, 199 L.Ed.2d 144 (2017), reh’g denied, ––– U.S. ––––, 138 S.Ct. 466, 199 L.Ed.2d 349 (2017).

In this case, the undisputed facts make the first step in the analysis simple: the jury’s exposure to material not properly offered during trial—i.e., the dictionary—obviously occurred.

The supervising judge’s decision to grant the jury’s request, made during its deliberations, to use the dictionary was error. Previously, we have declined to opine as to whether a juror’s use of a dictionary during deliberations—unknown to the judge and first disclosed after the verdict—constituted misconduct on the part of the juror. United States v. Rogers, 121 F.3d 12, 17, n.5 (1st Cir. 1997). The case now before us goes beyond Rogers. Here, the court itself blessed the use of the dictionary by approving the jurors’ request.

Provision of a dictionary to a jury by a judge after the close of the evidence and the instructions—except perhaps in extraordinary circumstances and after thorough discussion with counsel on the record—should not happen. At best, an extrinsic resource of this sort is superfluous. Proper definitions of key terms should be included in the instructions themselves, as they were here. At worst, dictionary definitions will conflict with definitions set forth in the instructions and create confusion, or even mislead a jury. Other courts have reached the same conclusion. See United States v. Lawson, 677 F.3d 629, 645 (4th Cir. 2012); United States v. Aguirre, 108 F.3d 1284, 1288 (10th Cir. 1997); United States v. Gillespie, 61 F.3d 457, 459 (6th Cir. 1995); United States v. Steele, 785 F.2d 743, 749 (9th Cir. 1986).

As already noted, the substantive mistake here was compounded by procedural errors in the way the request from the jury was handled. The proper process for managing a jury communication during deliberations has been set forth in deeply engraved authority: (1) the request from the jury should be in writing; (2) the writing should be marked as an exhibit; (3) the writing should be shown, or read verbatim, to counsel; and (4) counsel should be given an opportunity to be heard as to a proper response. United States v. Maraj, 947 F.2d 520, 525 (1st Cir. 1991). These steps should ordinarily be traced in open court on the record, so that a transcript of the pertinent discussion (including any objection) is available on review. None of these steps occurred here.

Moving on from the fact that a mistake occurred and that, as a result, the jury here was exposed improperly to extrinsic material, the analysis must proceed to the question of prejudice. In scrutinizing the trial judge’s decision to deny Appellant’s motion for new trial, we generally apply an abuse-of-discretion standard. The facts of this case do not, as Appellant contends, require that we presume prejudice.

It is true that older Supreme Court authority seemed to suggest that a jury’s exposure to any extrinsic material should be deemed presumptively prejudicial. Remmer v. United States, 347 U.S. 227, 229, 74 S.Ct. 450, 98 L.Ed. 654 (1954). But see United States v. Bristol-Martir, 570 F.3d 29, 41 n.5 (1st Cir. 2009) (questioning the “continued vitality” of Remmer’s holding, citing United States v. Bradshaw, 281 F.3d 278, 287-88 (1st Cir. 2002) ). It is now well established that less serious instances of potential taint should be addressed using the abuse-of-discretion standard, with the presumption of prejudice being reserved for more serious instances. Camacho-Santiago, 851 F.3d at 89; United States v. Zimny, 846 F.3d 458, 464 (1st Cir. 2017); United States v. Morosco, 822 F.3d 1, 13 (1st Cir. 2016); United States v. Rodriguez, 675 F.3d 48, 58 (1st Cir. 2012); Bristol-Martir, 570 F.3d at 41; Bradshaw, 281 F.3d at 291; United States v. Gomes, 177 F.3d 76, 80 (1st Cir. 1999). The less stringent standard applies where exposure to outside material is “inadvertent,” where the “trial judge responds to the claim of contamination by conducting an inquiry and employing remedial measures,” and where “egregious circumstances” are absent. United States v. Ofray-Campos, 534 F.3d 1, 21 (1st Cir. 2008) (quoting Bradshaw, 281 F.3d at 288).

It is true that in this case the exposure to the dictionary was not inadvertent, but rather was affirmatively approved by the supervising judge. This fact, while troubling, is insufficient to trigger a presumption of prejudice, where the trial judge’s response was energetic and probing, and the mistake, while clear, cannot fairly be described as “egregious.”

This is not a case like Ofray-Campos or United States v. Santana, 175 F.3d 57 (1st Cir. 1999), where the improper material significantly enhanced the evidentiary support for the government’s case, justifying the heavier presumption of prejudice standard. In general, the use of a dictionary will pose a qualitatively less serious risk of harm. See United States v. Cheyenne, 855 F.2d 566, 568 (8th Cir. 1988) (holding, in a case where a juror improperly consulted a dictionary, that while exposure to actual “evidence” would be presumed prejudicial, exposure to the “definition of a legal term” would not).

Of course, exceptions to this general approach may arise, in cases where, for example, the dictionary definition was contrary to, or confusingly inconsistent with, the instructions, where the jurors confirmed that they had actually relied on the misleading definition, or where the court made an inadequate effort to inquire into the impact of the taint. But none of these circumstances adheres in this case.

Identifying the standard of proof, of course, does not end the inquiry. Even where the abuse-of-discretion standard applies, situations may arise where the decision to deny a motion for new trial would demand reversal. One such situation would be where the trial judge failed to make an adequate inquiry into whether the extraneous material actually influenced the jury, as we found in Bristol-Martir.

Here, we discern no such problem. It is well established that in examining a trial court’s response to a claim of jury taint, we “abjure imposition of a rigid set of rules” for the conduct of the inquiry. United States v. Boylan, 898 F.2d 230, 258 (1st Cir. 1990). Flexible guidelines, however, as to what a “methodologically sound” inquiry entails set forth a reasonably clear path. Id. at 259. For instance, one such response included the following elements:

The court engaged counsel for both sides in an ongoing dialogue about the most appropriate way in which to handle the needed investigation, examined each juror twice, and pursued no fewer than eight lines of questioning proposed by defense counsel. The court’s probing into the extent of the jurors’ exposure to the extraneous information and its potential impact on their ability to render an impartial verdict was thorough and incisive. The court gave the jury clear and emphatic curative instructions. Last—but surely not least—the court made explicit findings that are amply rooted in the transcript of the two rounds of voir dire examinations and that make considerable sense when scrutinized against the record of the trial as a whole.

Bradshaw, 281 F.3d at 291–92.

The inquiry in this case followed similar lines. In considering Appellant’s motion for a new trial, the district court convened an evidentiary hearing to inquire into the use, if any, of the dictionary by the jury. At that hearing, the foreperson was questioned as to the jurors’ reliance on the dictionary and gave his opinion that the definition “did not really influence the deliberations.” A subsequent evidentiary hearing followed, in which four jurors were questioned. Finally, in a third proceeding, all twelve jurors were questioned individually prior to sentencing. The district court’s conclusion that Appellant suffered no prejudice from the provision of the dictionary was firmly anchored in the jurors’ testimony disclaiming reliance on the dictionary in reaching a verdict.

No sound reason suggests that the trial judge abused his discretion in reaching this conclusion. This is not a case like Bristol-Martir, where the trial judge failed even to inquire whether the jury’s decision had been affected by the extraneous material. The judge’s inquiry here explicitly probed that very issue with every single juror. It is true that—inevitably, given the twelve individual examinations—some inconsistencies in the responses appeared, but the trial judge was in the best position to weigh the significance of any ambiguities. It is well established that in conducting inquiries of this kind the district court has “wide discretion,” Bradshaw, 281 F.3d at 291, and absent circumstances not present here, we will defer to its findings. Camacho-Santiago, 851 F.3d at 89.

In sum, the record of this inquiry is more than sufficient to support the conclusion that the district court conducted a thorough and meticulous inquiry into the impact of the use of the dictionary and supportably concluded that it had no impact on the ultimate verdict.

Two other factors buttress our conclusion here. First, the evidence against Appellant was strong. The dictionary’s extraneous influence carried no significant risk of tipping a less than robust case in the government’s favor. Second, the dictionary offered no alternate definition of “knowingly” that was less favorable to Appellant, or more favorable to the government, than the definition contained in the instructions. In other words, even if the jurors had used the dictionary’s definition of “knowingly,” Appellant would have been no worse off.2

III. Conclusion

For the reasons set forth above, we conclude that the denial of Appellant’s motion for new trial constituted no abuse of discretion. We therefore hereby affirm the decision of the district court.


1.   The record offered with this appeal contains no transcript of the jury instructions, but their text is undisputed.

2.   Appellant’s argument that the dictionary’s alternate definition of “knowing” as “shrewd, clever, or implying a secret understanding” may have led the jury down an errant path makes no sense. This definition, as the trial judge found, would have increased the burden on the government, since the jury instruction’s definition required the government only to prove that Appellant was “conscious and aware of his actions, [and] realized what he was doing.”

PONSOR, District Judge.

Copied to clipboard