COMMONWEALTH vs. RICHARD SHUMAN.

COMMONWEALTH vs. RICHARD SHUMAN.

445 Mass. 268

September 9, 2005 - November 8, 2005

Norfolk County

Present: MARSHALL, C.J., IRELAND, SPINA, COWIN, & CORDY, JJ.

Homicide. Practice, Criminal, Capital case, Assistance of counsel, New trial. Evidence, Insanity.

A Superior Court judge's denial of a criminal defendant's motion for new trial based on newly discovered evidence did not create a substantial risk of a miscarriage of justice, where the evidence said to be new was available prior to trial, and where the opinion of the defense expert submitted in support of the motion was not substantially different from the defense offered at trial [271-276]; further, the judge did not abuse her discretion in deciding the motion without an evidentiary hearing [278-279].

The defendant in a murder trial was not denied effective assistance of counsel in the presentation of an insanity defense based on the defendant's medication-affected behavior. [276-278]

INDICTMENTS found and returned in the Superior Court Department on October 9, 1997.

The cases were heard by Margot Botsford, J., and a motion for a new trial, filed on November 19, 2002, was heard by her.

Donald A. Harwood for the defendant.

Varsha Kukafka, Assistant District Attorney (Susan Corcoran, Assistant District Attorney, with her) for the Commonwealth.

CORDY, J. On October 26, 1999, a Superior Court jury found Richard Shuman guilty of the premeditated murder of two of his business associates. Shuman appealed and filed a motion for a new trial in this court on November 19, 2002, on the grounds that newly discovered evidence cast substantial doubt on the justice of his convictions or, alternatively, that he was denied the effective assistance of counsel. We remanded the motion to the Superior Court for disposition and stayed appellate proceedings. After reviewing the affidavits and materials filed in support of the motion and without holding an evidentiary hearing, the motion judge, who was also the trial judge, denied

Page 269

the motion. The direct appeal from Shuman's convictions and the denial of his motion for a new trial have been consolidated in this court.

Shuman's defense at trial was lack of criminal responsibility due to a mental illness (depression reaching psychotic dimensions), exacerbated by other ailments and the side effects of medications that left him unable to conform his conduct to the requirements of the law. On appeal, he argues that evidence regarding the connection between the prescription medication Zoloft and a state of agitation known as akathisia was discovered after his trial and casts substantial doubt on the justice of his convictions; or in the alternative, his counsel was ineffective in failing to present evidence concerning the connection between Zoloft and akathisia at trial. He also argues that the motion judge erred in denying him an evidentiary hearing before acting on his motion for a new trial. Finally, Shuman asks the court to exercise its power under G. L. c. 278, § 33E, to reverse his convictions and grant him another trial so that the "critical evidence" he proffers regarding Zoloft may be heard and weighed by a jury. After undertaking a complete review of the trial record, we affirm the convictions, affirm the order denying Shuman's motion for a new trial, and decline to grant relief under G. L. c. 278 § 33E.

1. Background. We summarize the evidence in its light most favorable to the Commonwealth, reserving certain details for discussion in conjunction with the issues raised. The essential facts are not in dispute. On the afternoon of August 5, 1997, Shuman used a nine millimeter Beretta semiautomatic pistol to kill Jack Badler and Howard Librot. Shuman first went to Badler's office at Cabot Place in Stoughton. Badler called his human resources manager, Helen Anderson, to his office door so she might see that Shuman had a gun on his lap. He told her to dial 911 if Shuman did not put it away. Shuman told Anderson, "Don't worry. I won't hurt you," and put the gun away. As Anderson was leaving, Shuman told Badler that he had ruined his business and his life. He then shot Badler in the upper chest, eye, neck, and thumb, and left the office, telling two nearby employees, "Don't worry. I'm not going to shoot you two. I'm not upset with you."

Page 270

Shuman then drove to Librot's office, also in Stoughton. He entered the office and shot Librot in the head, neck, and chest. With his gun in hand, Shuman left the building and drove away. Both victims died as a result of their wounds. A few minutes later, Shuman arrived at his parents' house and told his mother that he had killed two people. While there, Shuman raised a gun to his head. His mother begged him to stop. Shuman put the gun away, and his parents drove him to the police station.

The shootings were directly linked to a worsening business relationship between Shuman and the victims. That relationship began when Shuman's printing business (Foremost Printers) purchased Web Corp. from Librot in 1989. Librot and his wife continued to run Web Corp. under Foremost's ownership. In 1991, Foremost Printers (Foremost) hired Badler to handle all of its finances and maintain control of the company books. Badler was a friend of Librot. By 1996, after a series of poor business decisions, Foremost was experiencing serious financial difficulties. In the face of these financial difficulties, tensions and disputes between Shuman, Librot, and Badler mounted. Badler and Shuman argued frequently about Badler's control of the company books. Foremost was unable to pay Badler for managing its books, and the company got behind in its financial obligations to the Librots. Librot instructed his employees not to permit Shuman on Web Corp. property and not to assist Shuman with any printing work. On August 1, 1997, Foremost's assets were sold. In connection with the sale, Shuman was required to sign a document that released Badler from any liability related to his work for Foremost. Shuman would not sign the release. On August 5, Badler processed the payroll for Web Corp., but not for Foremost. Badler refused to process Foremost's payroll until Shuman signed the release. That morning, there was a loud argument between Badler and Shuman in Badler's office. Shuman left, but returned later in the day with the gun, and the shootings followed.

There was also evidence that, starting in January of 1997, friends and family began to notice changes in Shuman's personality. He could not sleep, seemed depressed, lost weight, and on one occasion, held a gun to his head. Dr. George Gardos, a psychiatrist, saw Shuman on July 29, 1997, one week

Page 271

before the shootings, and diagnosed a major depression but did not believe Shuman to be a risk of harm to others. Dr. Gardos increased Shuman's preexisting Elavil prescription, to ease depression symptoms, and also prescribed Zoloft, a "mood elevator." After the killings, psychiatrist and defense expert Dr. Harold Bursztajn, of Harvard Medical School and the Massachusetts Mental Health Center, examined Shuman. Dr. Bursztajn opined that Shuman suffered from "a major depression, with anxiety which reached psychotic dimensions." He also testified that Shuman's diabetes exacerbated the depression, and that the antidepressants that had been prescribed to Shuman shortly before the killings magnified its symptoms. In particular, Dr. Bursztajn noted that Elavil, combined with Zoloft, can give rise to agitation, a need to act, and compulsive behavior that may be organized but is irrational. He also testified that these troubling side effects can come "right away," particularly in patients suffering from diabetes.

2. Discussion. a. Newly discovered evidence. Shuman argues that his motion for a new trial should have been allowed on the ground of newly discovered evidence linking Zoloft to a state of violent urges and agitation, known as akathisia, and proffers the following evidence to support his claim: (1) a posttrial study, published in 2000 by Dr. David Healy, linking Zoloft for the first time to violent urges in individuals who had previously suffered no mental illness; (2) expert testimony given in 2001 in civil litigation involving Pfizer, Inc., which publicly linked Zoloft to akathisia; (3) Dr. Joseph Glenmullen's posttrial examination of Shuman and his medical records, leading Dr. Glenmullen to conclude that Shuman was in a drug-induced state of akathisia when the murders occurred; and (4) an advisory released by the United States Food and Drug Administration (FDA) in 2004, warning of the possible side effects of Zoloft, including akathisia and agitation, especially at the beginning of drug therapy.

A defendant seeking a new trial on the ground of newly discovered evidence must establish that the evidence was unknown to the defendant or trial counsel and not reasonably discoverable at the time of trial. See Commonwealth v. Jones, 432 Mass. 623 , 633 n.6 (2000); Commonwealth v. Pike,

Page 272

 431 Mass. 212 , 218 (2000). He must also establish that the evidence "casts real doubt on the justice of the conviction." Commonwealth v. Salvati, 420 Mass. 499 , 506 (1995), quoting Commonwealth v. Scanlon, 412 Mass. 664 , 679-680 (1992). The evidence said to be new not only must be material and credible, see Commonwealth v. Brown, 378 Mass. 165 , 172 (1979), but also must "carry a measure of strength in support of the defendant's position." Commonwealth v. Pike, supra, quoting Commonwealth v. Grace, 397 Mass. 303 , 305 (1986).

The decision to deny or grant a motion for a new trial based on newly discovered evidence is committed to the sound discretion of the motion judge. Commonwealth v. Cintron, 435 Mass. 509 , 517 (2001). In deciding whether the new evidence would have been a significant factor in the jury's deliberations, had it been admitted at trial, we accord special deference to the action of a motion judge who was also the trial judge. Commonwealth v. Grace, supra at 307, citing Commonwealth v. DeChristoforo, 360 Mass. 531 , 543 (1971). The judge here found that the evidence submitted in support of Shuman's motion was not "new" and that there was "no substantial risk the jury would have reached a different conclusion had the evidence been admitted at trial." Commonwealth v. Grace, supra at 306.

Our examination of the record leads us to agree that Shuman's motion fails on two broad grounds. [Note 1] First, the evidence does not meet the test for "newly discovered" evidence because it was available prior to the trial. Second, the opinion of the defense expert submitted in support of the motion is not substantially different from the defense offered at trial.

The new evidence -- Dr. Healy's study, the expert testimony in the Pfizer litigation, Shuman's examination by Dr. Glenmullen, and the FDA advisory -- are new only in the sense that they had not occurred prior to trial. The link, however, between selective serotonin reuptake inhibitors (SSRIs), such as Zoloft, to violence and akathisia, was known in the scientific, medical,

Page 273

and academic communities prior to trial and was widely reported. [Note 2] Shuman claims, however, that the information available at the time of his trial principally linked akathisia to drugs

Page 274

such as Prozac, but not to the drug he was prescribed, Zoloft. As the motion judge found, there is "no suggestion in Dr. Glenmullen's affidavit that Zoloft is qualitatively different than other SSRIs, such as Prozac, such that the medical and other literature about Prozac that was published before October of 1999 would not be understood as relevant to Zoloft." [Note 3] The proposition that reasonable pretrial diligence would not have revealed the connection between SSRIs such as Zoloft and violent behavior in these circumstances is unconvincing. Moreover, although trial counsel has filed an affidavit stating he was "unaware of the phenomena of Zoloft poisoning and/or the now-established link between Zoloft and violence," he did question one of his own medical experts and the Commonwealth's expert regarding the connection between Zoloft and "[e]nhanced aggression," and the existence of "studies of whether or not Zoloft has contributed to killings." [Note 4] Additionally, in his opening statement, trial counsel referred to Shuman as someone "in a state of depression" who had been "crank[ed] . . . up on Zoloft" by his psychiatrist. There was also no evidence presented in connection with the motion for a new trial that Dr. Bursztajn, Shuman's highly credentialed psychiatric expert at trial (and a colleague of Dr. Glenmullen at Harvard Medical School), was unaware of the link between SSRIs and akathisia when he examined Shuman and formed his opinion, or that his opinion would be significantly different today from what it was at trial. In light of the extensive literature on the subject that existed in 1999, the judge found that a conclusion that Dr. Bursztajn was unaware of this link would be "very difficult to accept." She

Page 275

also noted that his testimony at the trial suggested otherwise. We agree.

The new evidence proffered here is merely a broadening of the research regarding SSRIs and violence already present in legal and scientific circles. The mere addition of further information to the preexisting debate does not amount to "newly discovered evidence" for the purposes of a new trial motion: "Undoubtedly, recent research has broadened the scientific community's understanding [but] we have concluded that expert testimony may not be considered newly discovered for purposes of a new trial motion simply because recent studies may lend more credibility to expert testimony that was or could have been presented at trial. To hold otherwise would provide convicted defendants with a new trial whenever they could find a credible expert with new research results supporting claims that the defendant made or could have made at trial." Commonwealth v. LeFave, 430 Mass. 169 , 181 (1999).

Even if we concluded that the evidence might marginally meet the "newness" requirement of a motion for a new trial, the proffered testimony of Dr. Glenmullen does not differ qualitatively from the testimony offered by Shuman's expert at trial. Dr. Glenmullen's posttrial conclusion that Shuman suffered from Zoloft-induced akathisia at the time of the killings is consistent with the testimony of Dr. Bursztajn, who testified that Elavil and Zoloft combined can make people more irritable and can give them more energy, "so they feel they have to do something." Dr. Bursztajn found Shuman "depressed, panicky and delusional." Dr. John Daignault, another defense expert who interviewed Shuman within hours of the shootings, found that Shuman was exhibiting signs of paranoia, depression, and that he was "[r]obotic." Shuman's delusion was that "he had to protect his family at any cost, at any price. And the only way to do so was to go ahead and shoot the two people that he felt had threatened his family." These descriptions of Shuman's symptoms are very close to Dr. Glenmullen's description of akathisia as a state of "anxiety, tension, irritability, and impatience," coupled with "panic reactions."

In essence, Shuman offers the same evidence of his symptoms simply cast under a different name, akathisia. Other than providing

Page 276

a specific term for those symptoms, Dr. Glenmullen's observations do little to change or add to the trial expert's testimony. At best, the distinction between the two opinions is that Dr. Glenmullen identifies Zoloft and the side effect of akathisia as the cause of Shuman's allegedly panicked, suicidal, and violent thoughts and actions, while Dr. Bursztajn attributed those same thoughts and actions to overwhelming depression exacerbated by Shuman's diabetes and side effects from the group of medications he was taking, including Zoloft. In these circumstances the motion judge did not abuse her discretion in finding that this distinction was not significant enough to create a substantial risk that the jury would have reached a different conclusion had the evidence been admitted at trial. See Commonwealth v. LeFave, supra at 176.

The denial of Shuman's motion either for its failure to establish newness or for its failure to cast doubt on the justice of the convictions, or on both grounds, did not create a substantial likelihood of a miscarriage of justice.

b. Ineffective assistance of counsel. "In evaluating a claim of ineffective assistance of counsel in a case of murder in the first degree, we begin by determining whether there was a serious failure by trial counsel." Commonwealth v. Harbin, 435 Mass. 654 , 656 (2002). We have described serious failure as "serious incompetency, inefficiency, or inattention of counsel -- behavior falling measurably below that which might be expected from an ordinary, fallible lawyer." Commonwealth v. Saferian, 366 Mass. 89 , 96 (1974). If serious failure is found on the part of counsel, the court then determines whether the failure resulted in a substantial likelihood of a miscarriage of justice, as required by G. L. c. 278, § 33E. Commonwealth v. Mitchell, 428 Mass. 852 , 854 (1999). Shuman argues that he was deprived of effective assistance of counsel because trial counsel failed fully to investigate and raise the defense of Zoloft-induced akathisia.

The record of the trial reflects a vigorous, well-prepared, and well-presented insanity defense, with substantial expert and medical testimony to support it. As the motion judge noted, counsel's cross-examination of the Commonwealth's rebuttal psychiatric witness was done "with skill, and specifically [included] ask[ing] that witness about the possible link between

Page 277

Zoloft and heightened aggression." Shuman, on the other hand, emphasizes the affidavit of his trial counsel to the effect that counsel was unaware of Zoloft poisoning or the link between Zoloft and violence, his expert did not inform him of Zoloft-induced violence, and counsel was unaware of the impact Zoloft had on Shuman's mental state.

Shuman's claims are unconvincing. Even if trial counsel did not use the term "akathisia" in his questioning of the witnesses, he did, as a part of the insanity defense, raise the link between Zoloft and heightened aggression. The Zoloft defense that Shuman now offers, and the insanity defense that trial counsel presented, are very close indeed; both aver that drugs, in combination with Shuman's preexisting depression, contributed to his aggressive behavior. Trial counsel argued in his opening statement that the defendant was "crank [ed] . . . up on Zoloft" by his doctor, and was a man "in a state of depression, with his medical condition, on Zoloft," who had been driven to the psychotic state that resulted in his pulling the trigger. Counsel also repeatedly argued in his closing that Shuman was affected by a "toxic soup" of medication, and had been "jack[ed]" up on "meds," including Zoloft. Application of the term "akathisia" to the behavior rather than "heightened aggression" is of little consequence.

In questioning trial counsel's use of the insanity defense, instead of the "Zoloft defense," Shuman challenges the tactical approach of his counsel and his principal psychiatric witness. This court gives some deference to trial counsel's tactical decisions, see Breese v. Commonwealth, 415 Mass. 249 , 251 (1993), and unless such decisions were "manifestly unreasonable when made," we will not find ineffectiveness. Commonwealth v. LaCava, 438 Mass. 708 , 713 (2003), quoting Commonwealth v. Martin, 427 Mass. 816 , 822 (1998). Shuman essentially asserts that his insanity defense may have been more effective if it had emphasized the alleged toxic effects of Zoloft, rather than the effect of depression worsened by a "toxic soup" of medication. We do not find the choice of emphasis in a defense to be manifestly unreasonable. See Commonwealth v. Lucien, 440 Mass. 658 , 670-671 (2004) ("Where the claim of ineffectiveness involves a tactical decision of defense counsel, we inquire

Page 278

whether the decision was 'manifestly unreasonable' when made").

This case is not akin to those in which we have found that the failure to investigate an insanity defense falls below the level of competence demanded of an "ordinary, fallible lawyer," where the relevant facts were known by or accessible to the attorney. See Commonwealth v. Doucette, 391 Mass. 443 , 458-459 (1984); Commonwealth v. Saferian, supra at 96. Presenting a similar but slightly different version of an insanity defense does not fall below the ordinary lawyer's level of competence. Nor is this a case where defense counsel "failed to pursue adequately . . . the only realistic defense." Commonwealth v. Licata, 412 Mass. 654 , 661 (1992). Shuman cannot claim that he was deprived of an "otherwise available, substantial ground of defence." Commonwealth v. Saferian, supra. Trial counsel's effective presentation of a well-prepared insanity defense, raising the link between Zoloft and aggression, rather than a pure "Zoloft defense," was not a serious failure on his part, nor is any difference between the two sufficient to create a substantial likelihood of a miscarriage of justice.

c. Denial of motion without an evidentiary hearing. It is within the judge's sound discretion to determine whether an evidentiary hearing is necessary. "In exercising the discretion to hold an evidentiary hearing, the judge must decide whether a substantial issue necessitating a hearing has been raised. In doing so, the judge looks not only to the seriousness of the claim presented, but also at the adequacy of the defendant's factual showing." Commonwealth v. Trung Chi Truong, 34 Mass. App. Ct. 668 , 674 (1993), citing Commonwealth v. Stewart, 383 Mass. 253 , 257-258 (1981). Shuman argues that the motion judge should have at least held an evidentiary hearing before deciding the new trial motion, especially because she discounted trial counsel's affidavit stating he was unaware of Zoloft's link to violence. We review the judge's decision for any abuse of discretion. Commonwealth v. Martinez, 437 Mass. 84 , 96 (2002).

The judge conducted a nonevidentiary hearing, thoroughly reviewed the extensive affidavits and materials filed by Shuman, as well as the trial testimony, and issued comprehensive findings. She found that the motion supported by Dr. Glenmullen's

Page 279

and trial counsel's affidavits did not present a sufficient basis for a new trial. We agree. "[E]very dispute among experts [does not require] an evidentiary hearing." Commonwealth v. Meggs, 30 Mass. 111 , 114 (1991). It is of no consequence that the judge discounted trial counsel's affidavit stating he was unaware of Zoloft's link to violence. She had ample reason to do so based on, among other things, trial counsel's arguments and examinations at trial. A judge may properly discount any portion of defense counsel's affidavit. See Commonwealth v. Goodreau, 442 Mass. 341 , 351 n.6 (2004); Commonwealth v. Savage, 51 Mass. App. Ct. 500 , 505 (2001).

d. Review under G. L. c. 278, § 33E. We have reviewed the entire record as required by G. L. c. 278, § 33E. The evidence of premeditation in the murders of Badler and Librot was strong. There is no basis for ordering a new trial or directing the entry of a verdict of a lesser degree of guilt. The defendant received a fair trial, and the jury's verdicts are consistent with the evidence.

Judgments affirmed.

Order denying motion for a new trial affirmed.

FOOTNOTES

[Note 1] We review the judge's decision under the substantial likelihood of a miscarriage of justice standard, as required by G. L. c. 278, § 33E. This is the standard that is applied when the appeal from the denial of a motion for a new trial is considered in conjunction with a direct appeal from a conviction of murder in the first degree. See Commonwealth v. Hill, 432 Mass. 704 , 710 n.14 (2000); Commonwealth v. Nieves, 429 Mass. 763 , 770 (1999).

[Note 2] At the time of the defendant's trial, a number of studies that examined the connection between violent behavior and SSRIs were available. For example, the following articles, which detail such studies and the effects of various kinds of SSRIs, are referenced in Dr. Glenumullen's affidavit. See, e.g., Fava, Four-Year Outcome for Cognitive Behavioral Treatment of Residual Symptoms in Major Depression, Am. J. Psychiatry 153 (1996); Hamilton, Akathisia, Suicidality, and Fluoxetine [Prozac], J. Clinical Psychiatry 53 (1992); Lane, SSRI-Induced Extrapyramidal [Involuntary Motor] Side-Effects and Akathisia: Implications for Treatment, J. of Psychopharmacology 12 (1998); R.E. Hales, S.C. Yudofsky, & J.A. Talbott, Psychiatry 1398 (3d ed. 1999); A.F. Schatzberg & C.B. Nemeroff, Psychopharmacology 939 (2d ed. 1998); Masand, Suicidal Ideation Related to Fluoxetine [Prozac] Treatment, New Eng. J. Med. 324 (1991); Dasgupta, Additional Cases of Suicidal Ideation Associated with Fluoxetine [Prozac], Am. J. Psychiatry 147 (1990); Papp, Suicidal Preoccupation During Fluoxetine [Prozac] Treatment, Am. J. Psychiatry 147 (1990); Mann, The Emergence of Suicidal Ideation and Behavior During Antidepressant Pharmacotherapy, Archives of Gen. Psychiatry 48 (1991); Koizumi, Fluoxetine [Prozac] and Suicidal Ideation, J. Am. Acad. of Child and Adolescent Psychiatry 30 (1991); King, Emergence of Self-Destructive Phenomena in Children and Adolescents During Fluoxetine [Prozac] Treatment, J. Am. Acad. of Child and Adolescent Psychiatry 30 (1991); Creaney, Antidepressant [Prozac and Luvox]-Induced Suicidal Ideation, Human Psychopharmacology 6 (1991); Dewan, Prozac and Suicide, J. Fam. Prac. 33 (1991); 5-HT Blockers [SSRIs] and All That, Lancet 345 (Aug. 11, 1990); Rothschild, Re-exposure to Fluoxetine [Prozac] After Serious Suicide Attempts by Three Patients: The Role of Akathisia, J. Clinical Psychiatry 52 (1991); Ichikawa, Effect of Antidepressants on Striatal and Accumbens Extracellular Dopamine Levels, Eur. J. Pharmacology 281 (1995); Dewey, Serotonergic Modulation of Striatal [involuntary motor system] Dopamine Measured with Positron Emission Tomography (Pet) and In Vivo Microdialysis, J. Neuroscience 15 (1995); DiRocco, Sertraline [Zoloft] Induced Parkinsonism. A Case Report and In Vivo Study of the Effect of Sertraline [Zoloft] on Dopamine Metabolism, J. Neural Transmission 105 (1998); Wirshing, Fluoxetine [Prozac], Akathisia and Suicidality: Is There a Causal Connection?, Archives of Gen. Psychiatry 49 (1992); Teicher, Emergence of Intense Suicidal Preoccupation During Fluoxetine [Prozac] Treatment, Am. J. Psychiatry 147 (1990); Pfizer, Inc., Safety Evaluation and Epidemiology, Suicide-Related Behavior in Children and Adolescents in the Sertraline [Zoloft] OCD Clinical Development Program (May 23, 1996); Beasely, Fluoxetine [Prozac] and Suicide: A Meta-Analysis of Controlled Trials of Treatment for Depression, Brit. Med. J. 303 (1991).

This information was also available in the legal community, as civil litigation and criminal prosecutions concerning the link between SSRIs and violence had been reported before the defendant's trial. See, e.g., Prozac Defense Rejected by Santa Clara Judge, The Recorder (Cal.) Nov. 13, 1991 (defense built around Prozac rejected); Brancaccio v. State, 698 So. 2d 597 (Fla. Dist. Ct. App. 1997) (defendant entitled to instruction on involuntary intoxication by Zoloft, side effects of which included aggression, difficulty sitting still, and loss of impulse control).

[Note 3] At trial, trial defense counsel repeatedly elicited testimony from the medical witnesses and the Commonwealth's expert that Zoloft was an SSRI.

[Note 4] Trial counsel also advised the judge at sidebar that he intended to question one of the medical witnesses about pharmaceutical company litigation alleging a connection between Zoloft and enhanced aggression, including litigation concerning an alleged Zoloft-induced murder. The judge refused to allow it. Additionally, he elicited testimony from the Commonwealth's expert that the Physicians' Desk Reference contains a warning of "[e]nhanced aggression" with respect to the use of Zoloft.