Warning a Patient’s Intended Victim While Preserving Testimonial Privilege: A Statute to Regulate All Pychotherapists
In Guerrier v. State, 811 S o. 2d 852 (Fla. 5th DCA 2002), a case of first impression in Florida, a psychiatrist was required to testify that a patient told him he was going to kill his former girlfriend. Believing the woman was in danger, Dr. Donny Baskaran instructed his nurse—who also heard the statements—to warn the intended victim. During a trial for aggravated stalking, the judge allowed Baskaran to take the stand over the objection of the patient, Francois Guerrier, that communications between the two, including the threat, were covered by the psychotherapist-patient privilege. After being convicted, he appealed, but the Fifth District Court of Appeal affirmed.1
This article briefly explains and distinguishes confidentiality and privilege and asserts that an apparent misunderstanding of these concepts led to the holding in Guerrier that warning a patient’s intended victim effected a waiver of the privilege. This article discusses the opinion and demonstrates why this conclusion is wrong; and points out that Florida statutes provide different rules regarding confidentiality and privilege for a psychiatrist, psychologist, social worker, or other mental health counselor. The article suggests solutions to both the issue raised in Guerrier and the confusion caused by discrepancies in the various laws, and proposes a statute that should be adopted for all psychotherapists.
Confidentiality and Privilege
Although related, confidentiality and privilege are not the same and must be evaluated separately. Confidentiality requires that the psychotherapist never divulge what he or she is told during treatment while the statutory privilege permits the patient to prevent a therapist from testifying in a legal proceeding about these confidential communications.
• Confidentiality
Psychiatrists are obligated to keep confidences based on the patient’s right to privacy and the need for complete honesty to achieve the full benefits of therapy.2 Nevertheless, following Tarasoff v. Regents of the University of California, 551 P.2d 334 (Cal. 1976)—the landmark case that established a therapist’s duty to protect a patient’s intended victim3—Florida legislators recognized there might be instances where it becomes necessary to betray confidences. Consequently, they first added a dangerous patient exception to the confidentiality requirement in §456.059 for psychiatrists, and later to §490.0147 for psychologists, and §491.0147 for social workers and other mental health professionals. Surprisingly, despite the fact that the problem is the same for all these professionals, the statutory treatment is different.4
Because the therapist in Guerrier was a psychiatrist, the court turned to §456.059. This section gives the doctor who believes a patient will hurt someone within the near future discretion to disclose threats “to the extent necessary” to warn the targeted individual.5 Doing so obviously represents a breach of confidentiality, but Florida lawmakers determined the need to safeguard potential victims outweighed the benefits of not revealing these statements. The American Psychiatric Association agrees.6 Therefore, legally and ethically, the physician has the option of warning potential victims, even though by doing so the physician will be breaching a confidence.7
• Privilege
The patient has a privilege to prevent his psychotherapist from testifying in a legal proceeding8 about confidential communications made for the purpose of “diagnosis or treatment” of a “mental or emotional condition.”9 Privileges, which represent “an exception to the general rule that all relevant evidence is admissible,”10 are grounded in the determination that society benefits more from protecting the relationship than from compelling the testimony.11 Thus, in enacting §90.503, legislators recognized disclosing a patient’s confidences in a courtroom could be “a substantial factor in restricting the patient’s freedom in providing essential information to the doctor for treatment.”12 Accordingly, to encourage people to seek help, and to increase the efficacy of treatment, patients are granted the right to preclude their therapists from discussing their communications in any judicial, legislative, or administrative action.
However, similar to confidentiality, privilege is not absolute. Section 90.503, which creates the privilege, also specifies three circumstances to which it does not apply: 1) attempts to involuntarily hospitalize a person for mental illness,13 2) court-ordered examinations, or 3) lawsuits in which a patient relies on his or her mental or emotional condition as an element of a claim or defense.14 Obviously, warnings to potential victims do not fall within any of these enumerated exemptions and they should not be implied because, where lawmakers create specific statutory exceptions, judges “may. . . infer that ‘had the legislature intended to establish other exceptions it would have done so clearly and unequivocally.’”15 Thus, the patient should be able to stop a psychiatrist from discussing the patient’s threat in court.16
Guerrier v. State
The Guerrier court disagreed but its reasoning was flawed.
First, unlike confidentiality under §456.059, the statutory privilege in §90.503 does not include a dangerous patient exception that would let a physician testify without consent. Contrary to the holding in Guerrier, absent such an exception, a psychiatrist satisfies his or her obligations by informing either the intended victim or law enforcement personnel of the risk. The psychiatrist should not be compelled to appear in court.
Second, although the Fifth District ruled that in enacting §456.059 legislators “provided for an exception to the privilege,” the statute actually only grants psychiatrists discretion to breach confidentiality. Therefore, Baskaran’s testimony should have been excluded because, as the court acknowledged: 1) the privilege was not recognized at common law so it must be strictly construed; and 2) §456.059 does not “specifically provide that the psychiatrist may testify regarding the threat at any subsequent trial.”17 The panel’s assertion that it would be an “absurd result” if the physician who reports the statement were prevented from testifying about it is itself absurd. In improperly creating another situation to which the legislatively created privilege does not apply, the judges failed to appreciate the difference between confidentiality and privilege.
Third, according to Guerrier, the “goal of victim protection” under §456.059 “extends to eliciting from the psychiatrist relevant evidence (the threat) that will facilitate the prosecution” of the dangerous patient if the patient commits the crime.18 This conclusion is puzzling. It is not the psychiatrist’s job to make the state’s case; rather, the statute requires only that the doctor warn, rather than protect, the victim. This is an important distinction with which courts19 have wrestled since 1976 when, on rehearing in Tarasoff, the California Supreme Court expanded what it originally held was a duty to warn into a duty to protect.20 Indeed, it is difficult to imagine how a therapist could protect a patient’s intended victim—a person the therapist probably does not even know.
Despite the contrary language in Guerrier, Florida has not adopted a duty to protect.21 In fact, in Boynton v. Burglass, 590 So. 2d 446, 447 (Fla. 3d DCA 1991), decided prior to the effective date of §455.2415 (which later became §456.059), the Third District specifically rejected Tarasoff.22 Further, the court explained that even if the statute had applied, it would not have changed the result as it “is couched in permissive terms, and merely provides that a psychiatrist may disclose patient communications. . . . ”23 The Boynton majority refused to establish a duty to warn or protect because of the “inherent difficulties psychiatrists face in predicting a patient’s dangerousness.”24
Thus, until Guerrier, both Florida lawmakers and judges demonstrated a consistent unwillingness to embrace the Tarasoff duty to protect. As the court explained in Boynton, “To impose a duty to warn or protect third parties would require the psychiatrist to foresee a harm which may or may not be foreseeable, depending on the clarity of his crystal ball. . . [and] would not only be unreasonable and unworkable, it would also wreak havoc with the. . . relationship.”25 As such, the doctor’s obligation is satisfied upon telling either the individual or the police of the possible danger because these are the people in the best position to protect the victim by implementing safety measures.
Contrasting Other Professionals
The discussion, however, cannot end here. Section 90.503 establishes a psychotherapist-patient privilege for psychiatrists, psychologists, social workers, and other mental health counselors. And the Boynton court noted that, although the defendant was a psychiatrist, its decision “applie[d] equally to psychologists, psychotherapists, and other mental health practitioners.”26 While Guerrier also only involved psychiatrists, it did not mention the different types of therapists. But analysis of the statutes that regulate these various professionals reveals troubling inconsistencies concerning confidentiality and privilege when a patient threatens someone.
For example, §456.059, which does not even refer to privilege, permits a psychiatrist to warn if the psychiatrist believes “it is more likely than not. . . in the near future” a patient will harm an identifiable victim.27 But §490.0147, which governs psychologists, states “the privilege may be waived” if there is “a clear and immediate probability of physical harm to. . . other individuals.”28 Contrast §491.0147, which regulates social workers, marriage and family therapists, and mental health counselors and which says “secrecy may be waived” where “a clear and immediate probability of physical harm to. . . other individuals” exists.29
Another important distinction is that §456.059 provides immunity from civil or criminal liability to the psychiatrist who discloses confidential communications in warning about a threat; neither §490.0147 nor §491.0147 grants this essential protection to those licensed under these provisions. Also, §456.059 only deals with threats to harm third parties while both §§490.0147 and 491.0147 permit “waive[r]” when the patient is a danger to self.30 Further, psychiatrists can either warn the victim or law enforcement but psychologists and other mental health professionals may reveal the information to an “appropriate family member. . . or other appropriate authorities” as well.
Because these three statutory provisions all address the same subject—a therapist treating a dangerous patient—they should be consistent.31 According to the rules of statutory construction, if they are not, as is currently true, the differences must be given meaning.32 Thus, as privilege is not mentioned in §456.059, but is mentioned in §490.0147, legislators apparently only intended to abrogate the privilege for psychologists. This conclusion provides further support for the argument that the Guerrier court erred when it held Baskaran, who is a psychiatrist, must testify.
Moreover, despite the fact that the title of §491.0147 is “confidentiality and privileged communications,” the text does not refer to privilege but says warning an intended victim waives “secrecy.” There is no indication why the language changed from abrogating “privilege” for psychologists to “secrecy” for social workers and other mental health counselors. This choice is especially perplexing because, as confidentiality prevents the practitioner from revealing a patient’s communications, by definition notifying an intended victim destroys secrecy. “[C]ourts should assume that the legislature knew the plain and ordinary meaning of words when it chose to include them” in a statute.33 Therefore, as lawmakers understood “secrecy” when they selected the term, and knew about “privilege” because they used it in Ch. 490 for psychologists, the assumption is they intended to allow those licensed under Ch. 491 to breach confidentiality (warning waives secrecy) in these cases but not to require them to testify (no mention of privilege).
Similarly, contrasting §490.0147 (threats waive privilege) with §491.0147 (threats waive secrecy) suggests legislators meant to indicate professionals licensed under these two statutes should not be treated the same. This argument is particularly persuasive as §§490.0147 and 491.0147 were enacted together as part of the same bill.34 Consequently, it is likely the distinction was intentional.
What all this means is that psychologists are the only psychotherapists for whom privilege is waived by reporting a dangerous patient. Because no justification for this discrimination against psychologists—and those who consult them for help—exists, the statutes need to be amended to be consistent. Moreover, to resolve the problem created by Guerrier, a sentence should be added to all three statutes providing that warning an intended victim has no effect on the psychotherapist-patient privilege under §90.503.
In addition to the obvious benefit of fairness to the patients of the various professionals, such amendments also would be useful because judges appear confused by the question of which statute to apply. Kasdaglis v. Department of Health, 827 So. 2d 328 (Fla. 4th DCA 2002), which concerned a licensed clinical social worker who informed a divorced father that his minor child might hurt himself, is illustrative. The Fourth District cited §490.0147; but Ch. 490 regulates psychologists, not social workers. Luckily, the court reached the correct outcome anyway and held the practitioner appropriately disclosed the conclusion that the patient might be dangerous but did not reveal the underlying confidential communications.35 In fact, one of the reasons the panel reversed was that “[t]he final order ignore[d] this statutory distinction.”36
Although the issue was different, the analysis in Kasdaglis is helpful in rejecting the need to compel a psychiatrist to testify about a patient’s threat. Under §456.059, the physician has discretion to warn; but if the physician does choose to warn, he or she should only tell the victim or police the conclusion that the patient represents a danger to another. It would ordinarily be a mistake, and unnecessary, to reveal confidences used to formulate that decision. It would be even more inappropriate to require the physician to become a witness against the patient.
The Fourth District also applied the wrong statute in Tessler v. Tessler, 539 So. 2d 522 (Fla. 4th DCA 1989). The experts in this custody modification proceeding were psychologists but the court discussed the waiver provisions of §491.0147,37 which governs social workers and other mental health counselors.
Suggestions
Warning an intended victim should not waive the privilege. A licensed practitioner who exercises discretion and alerts a third person to the risk has taken appropriate steps to do what is necessary to keep the patient from harming an innocent person. “It is one thing for a psychotherapist to contact law enforcement or a potential victim to prevent a patient from carrying out dangerous, criminal intentions, and quite another to compel the therapist to testify to confidential conversations with the patient in a later criminal proceeding against the patient.”38 Forcing the therapist to take the stand is likely to destroy whatever chance the therapist might have had to reach the patient and may deter other troubled individuals from seeking treatment because they fear forced disclosure of their secrets. To avoid these unacceptable consequences, the patient should be permitted to prevent a therapist from being a witness against him or her.
Unfortunately, while certainly an important step, settling this issue would not completely resolve the problems. Legislators must also change the laws that govern those defined as psychotherapists in §90.503. The distinctions in §§456.059, 490.0147, and 491.0147 make no sense. These professionals are performing the same role, the privilege statute includes them all, and their ethics codes require confidentiality while providing for disclosure if needed to prevent harm to some identifiable individual.39
Thus, communications between psychotherapists and their patients should generally remain confidential. But, because therapists may be their patients’ only confidants and they are probably in the best position to assess the likelihood of the threat being carried out, statutes should provide an exception when necessary to warn a specific victim. Further, whether the practitioner is a psychiatrist, psychologist, social worker, or other mental health counselor, he or she must be able to warn the individual or law enforcement personnel without having to worry about civil or criminal liability. To accomplish this, the law should grant immunity for breaching confidentiality.
To qualify for this protection, however, the therapist must restrict disclosure to information necessary to notify the victim of the danger. Using the analysis in Kasdaglis as a guide, the professional should not divulge the statements that led to such conclusion. In other words, most of the time it will be sufficient to explain that the individual may be at risk; the practitioner should not reveal the substance of the patient’s communications. Moreover, even after the therapist breaches confidentiality, he or she should not be compelled to testify in any subsequent legal proceedings40 if the patient objects.
Conclusion
To resolve both the problem raised by Guerrier and the inconsistencies in the current statutes, legislators should amend the laws41 so they are all the same. Sections 456.059, 490.0147, and 491.0147 should read:
Communications confidential; exceptions.—
Communications between a patient and [insert type of licensed professional], as defined in [insert specific statutory provision], shall be held confidential and shall not be disclosed except upon the request of the patient or the patient’s legal representative. Provision of [insert type of licensed professional] records and reports shall be governed by [insert specific statutory provision]. Notwithstanding any other provision of this section where:
(1) A patient is engaged in a treatment relationship with a [insert type of licensed professional];
(2) Such patient has made an actual threat to physically harm an identifiable victim or victims; and
(3) The treating [insert type of licensed professional] makes a clinical judgment that the patient has the capability to commit such an act and it is probable that in the near future the patient will carry out that threat,
the [insert type of licensed professional] may disclose patient communications, but only to the extent necessary to warn any potential victim or to communicate the threat to a law enforcement agency. No civil or criminal action may be instituted, and there shall be no liability on account of disclosure of otherwise confidential communications by a [insert type of licensed professional] to the extent necessary to disclose a threat pursuant to this section. Disclosures made under this section do not affect the patient’s right to assert the psychotherapist-patient privilege under §90.503.
1 Guerrier, 811 So. 2d at 852–53.
2 Carolyn Peddy Courvile, Rationales for the Confidentiality of Psychotherapist-Patient Communications: Testimonial Privilege and the Constitution, 35 Houston L. Rev. 187, 210–13 (1998).
3 Tarasoff, of course, was only binding in California. However, since the decision, most jurisdictions have adopted some version of either the duty to protect or to warn. For discussion of the differences, see text accompanying notes 19–23. Only Florida and Virginia have specifically rejected the decision. Melissa L. Gilbert, Comment, “Time-Out” for Student Threats?: Imposing a Duty to Protect on School Officials, 49 UCLA L. Rev. 917, 927 (2002).
4 See text accompanying notes 25–34.
5 Fla. Stat. §456.069 (2003).
6 Principles of Medical Ethics with Annotations Especially Applicable to Psychiatry, §4(8), available at psych.org/apa_members/ethics_princip.cfm. (“at times” psychiatrists “may find it necessary” to reveal confidential information “to protect the patient or the community from imminent danger”). See also note 38 and accompanying text.
7 Fla. Stat. §456.059 (2002); see note 39 and accompanying text.
8 McCormick on Evidence (John W. Strong ed., 5th ed. 1999).
9 Fla. Stat. §90.503 (2002).
10 Guerrier, 811 So. 2d at 854.
11 George C. Harris, The Dangerous Patient Exception to the Psychotherapist-Patient Privilege: The Tarasoff Duty and the Jaffee Footnote, 74 Wash. L. Rev. 33, 45–46 (1999).
12 Attorney Ad Litem for D.K. v. Parents of D.K., 780 So. 2d 301, 305–06 (Fla. 4th D.C.A. 2001).
13 Although this means the physician may testify in a commitment hearing even over a patient’s objection, it is appropriate because the purpose is to help the patient. The balance shifts in a case like Guerrier, however, where disclosure is counter to the patient’s interests and could result in incarceration.
14 Fla. Stat. §90.503(4) (2002).
15 Bearelly v. State, 2002 WL 982429, *4 (Fla. Cir. Ct. 2002) (quoting Florida Legal Services, Inc. v. State of Florida, Dep’t of Labor and Employment Security, 381 So. 2d 1120, 1122 (Fla. 1st D.C.A. 1979)).
16 Cf. State v. Famiglietti, 817 So. 2d 901, 904 (Fla. 3d D.C.A. 2002) (where none of the statutory exceptions applied, psychiatric records were privileged).
17 It is probable that the error was harmless and Guerrier’s conviction would have been upheld even without the psychiatrist’s testimony. As may be true in many cases, because of the abundance of other evidence, the therapist’s statements were not necessary to prove guilt.
18 Guerrier, 811 So. 2d at 855–56.
19 Gilbert, supra note 3, at 927–29 (reviewing reaction to Tarasoff and how the “decision quickly spread through American law”).
20 Tarasoff, 551 P.2d at 345.
21 See generally Andrew Clifford Greenberg, Comment, Florida Rejects a Tarasoff Duty to Protect, 22 Stetson L. Rev. 239 (1992).
22 Boynton, 590 So. 2d at 451 (emphasis original).
23 Id. at 450.
24 Id.
25 Id. at 448 n.8.
27 Fla. Stat. §456.059 (2002).
28 Fla. Stat. §490.0147 (2002).
29 Fla. Stat. §491.0147 (2002).
30 These statutes also provide for waiver when the patient files a complaint against a therapist, or agrees in writing, or, in family therapy, when each family member agrees in writing. Fla. Stat. §§490.0147 and 491.0147 (2002).
31 Ronald Benton Brown & Sharon Jacobs Brown, Statutory Interpretation: The Search for Legislative Intent 45–46 (2002).
32 2A Norman J. Singer, Statutes and Statutory Construction §46.06, at 181–92 (2000 Rev.).
33 Hanker v. Yarian, 755 So. 2d 93, 96 (Fla. 2000).
34 Laws 1987, c. 87-252. Section 490.0147 was §13; §491.0147 was §15.
35 Kasdaglis, 827 So. 2d at 333. Interestingly, the court switches to ch. 491 when discussing patient records. Id. at 333–34.
36 Id. at 333.
37 Tessler, 539 So. 2d at 523.
38 Harris, supra note 11, at 35.
39 See, e.g., Code of Ethics of the National Association of Social Workers, §1.07(c) (1999), available at naswdc.org/pubs/code/code.asp; Ethical Principles of Psychologists and Code of Conduct 2002, §4.05(b)(3) (2002), available at apa.org/ethics/code2002.html; Principles of Medical Ethics with Annotations Especially Applicable to Psychiatry, §4(8), available at psych.org/apa_members/ethics_princip.cfm.
40 The only exception should be an action for involuntary hospitalization.
41 Famiglietti, 817 So. 2d at 908 (while “reasonable people can disagree over whether evidentiary privileges have value, and under what circumstances (if any) privileged communications may be invaded,” because they created them, if privileges are to be modified, the legislature is the appropriate body to do so).
Phyllis Coleman is a professor of law at Nova Southeastern University where she teaches bioethics, family law, sports law, and contracts. She received her B.S., M.Ed., and J.D., with high honors, from the University of Florida.
Ronald A. Shellow has been in the solo private practice of psychiatry since 1961. He received his B.S. and M.D. from the University of Illinois College of Medicine. Dr. Shellow is certified to practice psychiatry and forensic psychiatry by the American Board of Psychiatry and Neurology and is an examiner for that board.
The authors thank Ron Brown, Mark Dobson, and Doris Shellow for their helpful comments and suggestions.
