On October 27, 1993, the Board of Registratian in Medicine (board) suspended the right of Dr. Peter L. Herridge (petitioner) to renew his license to practice medicine in the Commonwealth for three years. See Wang v. Board of Registration in Medicine, 405 Mass. 15, 17-21 (1989). The petitioner appealed from the board’s decision pursuant to G. L. c. 112, § 64 (1992 ed.), and a single justice of this court reported the case to the full court without decision.
The board ruled that, because of inappropriate conversatian of a sexual nature initiated by the petitioner with a patient during treatment, and a brief sexual relationship with this patient which occurred shortly after the physician-patient relationship terminated, the petitioner was (1) guilty of conduct which places into question his competence to practice medicine in violation of G. L. c. 112, § 5 (c) (1992 ed.), and 243 Code Mass. Regs. § 1.03 (5) (a) (3) (1993); (2) guilty of misconduct in the practice of medicine within the meaning of 243 Code Mass. Regs. § 1.03 (5) (a) (18) (1993); and (3) guilty of conduct which undermines public confidence in the integrity of the medical profession and shows a lack of good moral character. The petitioner does not deny engaging in sexual relations with the patient on two or three occasions after the termination of the physician-patient relationship. He argues, nonetheless, that the record lacks substantial evidence that he engaged in unethical conduct during the physician-patient relationship. The petitioner also argues that his right to due process was violated during the course of the proceedings before the board.
Although we reject the petitioner’s contentions concerning due process, we conclude that the case must be remanded to *156the board. One fundamental issue for the board and, before that, for the hearing officer appointed to hear the allegations, was whether to believe the patient or the petitioner. There were discrepancies in the patient’s testimony susceptible of interpretation as fabrication. In these circumstances, it was incumbent on the hearing officer, whose determinations of credibility, explanation for those determinations, and findings of fact were adopted verbatim by the board, to provide a thorough and reasoned explanation for the decision to credit the patient’s testimony.1 If the patient’s testimony is discounted, there remains insufficient evidence of misconduct in the record (despite the petitioner’s admission of a sexual relationship) to warrant the sanction imposed.
1. Due process claims. We consider first the petitioner’s contention that various rulings made by the hearing officer were not consistent with the requirements of due process and that the board’s decision must be set aside for this reason alone. See Arthurs v. Board of Registration in Medicine, 383 Mass. 299, 316 (1981).
a. Access to the patient’s psychiatric records. Prior to the hearing, the petitioner’s attorney requested that the prosecuting attorney produce certain psychiatric hospital records of the patient that had been released to the prosecuting attorney by the hospital. The records related to treatment received by the patient before and after she was treated by the *157petitioner.2 The prosecuting attorney declined to produce the records on the ground that they were privileged communicatians under G. L. c. 233, § 20B (1992 ed.), to which the patient had not agreed to waive confidentiality. After receiving notification that the prosecuting attorney would not rely on the records, the hearing officer declined to review the records in camera and denied the petitioner’s request for access to them. There was no error.
On the patient’s assertion of privilege, the records were protected from disclosure by G. L. c. 233, § 20B, which makes confidential communications between a patient and her psychotherapist. The petitioner relies on the principle applicable to criminal cases, “that in certain circumstances, a defendant must have access to privileged records so as not to undermine confidence in the outcome of the trial” (emphasis in original). Commonwealth v. Bishop, 416 Mass. 169, 176 (1993). This principle, however, rests on Federal and State constitutional guarantees of due process that have applicatian in criminal proceedings. See Commonwealth v. Stockhammer, 409 Mass. 867, 881-883 (1991). The power of the board to impose civil penalties for misconduct in the practice of medicine does not render the proceedings penal in nature. See Zora v. State Ethics Comm’n, 415 Mass. 640, 653 (1993); Arthurs v. Board of Registration in Medicine, 383 Mass. 299 (1981). The Bishop rules do not apply to proceedings before the board. Moreover, although the patient had been treated by mental health professionals, there was no evidence to suggest that she was delusional or had a disorder affecting her ability to perceive and to recall events. See Daniels v. Board of Registration in Medicine, 418 Mass. 380, 386-391 (1994). Denial of access to the patient’s psychiatric records did not violate the petitioner’s right to due process.
*158b. Recusal motion. The hearing officer informed the parties that she was acquainted with Dr. James C. Beck, who testified as an expert on psychiatric ethics for the board.3 On motion of the petitioner, she declined to recuse herself. A matter of recusal is generally left to the discretion of the adjudicator, “and an abuse of that discretion must be shown to reverse a decision not to allow recusal.” Haddad v. Gonzalez, 410 Mass. 855, 862 (1991). The hearing officer determined that her relationship with Dr. Beck would not compromise her ability impartially to assess the credibility and weight of his testimony. Her casual acquaintance with the witness, which ended some five years prior to commencement of the hearing before her, created no appearance of bias or partiality. “This is not a case in which it appears ‘that the probability of actual bias on the part of the . . . decisionmaker is too high to be constitutionally tolerable.’ ” Varga v. Board of Registration of Chiropractors, 411 Mass. 302, 307 (1991), quoting Withrow v. Larkin, 421 U.S. 35, 47 (1975).
2. Credible and substantial evidence of misconduct. We turn now to the petitioner’s contention that the board’s credibility determination is neither adequately explained nor supported by the record.4 We summarize the board’s findings of fact, supplemented by reference to exhibits and the testimony before the hearing officer, to the extent necessary for analysis of the legal issues before us. The patient first met the peti*159tioner in 1982 when she was working at McLean Hospital (hospital) as an intern in the process of obtaining her master’s degree in psychology. Her supervisor was Dr. George Dion. The petitioner completed a three-year residency training program in psychiatry at the hospital in 1984; thereafter, he was a staff psychiatrist at the hospital until he left Massachusetts for New Jersey in the spring of 1986. The petitioner’s specialty is in psychopharmacology, defined by the board as “the prescribing of medication to alleviate biological imbalances believed to cause psychiatric symptoms.” While he was on the staff at the hospital, he treated a number of people with bulimia (generally women) referred to him by a group of physicians who had pioneered treatment of this condition with antidepressant drugs. In 1983 and 1984, the patient and the petitioner worked on the same ward at the hospital. Discussions of a social nature took place among those working on the ward, and, on at least one occasion, the patient and the petitioner were among a group who met socially with other hospital personnel outside the hospital.
In December, 1984, the patient revealed to Dr. Dion that she was having problems with bulimia and depression. She sought from Dr. Dion a referral to a physician at the hospital who would treat her condition with medication. She was also concerned that her medical problems and any treatment remain confidential so that her employment prospects at the hospital would not be compromised. Dr. Dion mentioned several names, including that of the petitioner, and the patient decided to request treatment from the petitioner. The patient sought treatment by medication only. While she was under the petitioner’s care, she was also receiving treatment from a psychotherapist.
The patient’s first appointment with the petitioner, which lasted one hour and was in his office at the hospital, took place on January 4, 1985. As reflected in notes of the visit, he obtained the patient’s medical history, ordered blood tests, and prescribed the antidepressant Nortriptyline. The patient had a total of nine office visits with the petitioner, the last of which was on October 8, 1985. Appointments were scheduled *160most closely in time during the first two months of the physician-patient relationship, until the patient was stabilized on a medication which she tolerated well. Thereafter, the appointments were bimonthly. Appointments subsequent to the first appointment were approximately thirty minutes in duration.
The board credited the patient’s testimony that, beginning during the first or second appointment, the petitioner engaged in personal conversation that inappropriately interjected sexual topics into the physician-patient relationship. The patient testified that during her first or second appointment, the petitioner told her he found her attractive. During subsequent appointments, the petitioner discussed his personal life with her, including problems with his marriage, his wife’s overly religious nature, her “paranoia” about the possibility that he was having affairs and that he used a particular kind of soap so that she would not know he was having affairs. In addition, according to the patient, the petitioner told her that he found gratification in treating young, attractive bulimic women and confided his fear that he would never be satisfied sexually by one woman. The petitioner denied that these conversations (or an incident of physical intimacy in the office alleged by the patient) had occurred.
During the physician-patient relationship, there was discussion of formation of a therapy group for bulimic women for which the petitioner would act as medical advisor. In October, 1985, about three weeks after the patient’s last appointment, at the petitioner’s suggestion, he and the patient met for dinner to discuss the matter. During the first or sec- and week in November, 1985, at the petitioner’s suggestion, he and the patient arranged to meet at her apartment and to walk to a nearby restaurant for lunch. They did not go out to lunch. Instead, they remained at the patient’s apartment and engaged in sexual relations. This occurred on two additional occasions within the next two weeks.5 Shortly thereafter, the *161patient made contact with the petitioner by telephone and informed him that she did not want to continue their sexual relationship. The petitioner and the patient did not meet again.
“While the task of assessing the credibility of witnesses is one uniquely within an agency’s discretion (School Comm. of Wellesley v. Labor Relations Comm’n, 376 Mass. 112, 120 [1978]), this court may modify or set aside findings and conclusions that are arbitrary or unsupported by substantial evidence. See G. L. c. 30A, § 14 (7) [1992 ed.]; Embers of Salisbury, Inc. v. Alcoholic Beverages Control Comm’n, 401 Mass. 526, 529 (1988).” Bettencourt v. Board of Registratian in Medicine, 408 Mass. 221, 227 (1990).
In her initial communication to the board, made some three years after the events under discussion,6 the patient stated that it was not until nine to ten months into her treatment that “it became clear to [her] that [the petitioner] meant to pursue a personal relationship with [her].” Before the hearing officer, the patient testified that, during the first or second appointment, the petitioner told her he found her attractive, and “lengthy,” inappropriate conversations occurred regularly thereafter during the office visits. Although her initial statement to the board contained no allegation of physical intimacy during office visits, before the hearing officer, the patient testified that during an appointment on October 8, 1985, in his office, the petitioner kissed and caressed her for ten tó fifteen minutes until she put a stop to the physical intimacy.
The board’s finding that the physician-patient relationship ended on or around October 8, 1985 (the date of the last recorded office visit) required it to reject a considerable portian of the patient’s testimony bearing on the duration of the physician-patient relationship. The patient testified that the *162petitioner continued to consult with her by telephone about her use of antidepressant medication through February, 1986, that she had a final office appointment with him that month, and that he “continued to prescribe and monitor the antidepressant that [she] was taking through the spring of [1986] and continued to write prescriptions.” She testified that he told her in the spring of 1986, that he was leaving the area, and that she did not seek treatment from another physician, a Dr. Edward S. Schwartzreich, until the summer of 1986.
The petitioner, in contrast, testified that he began making plans to leave the Commonwealth in September, 1985, and that, in September and again in October, 1985, he spoke with Dr. James Hudson, another psychiatrist specializing in psychopharmacology at the hospital, about taking over the care of the patient (without identifying her by name). Dr. Hudson confirmed these conversations and also testified that the petitioner said in October, 1985, that he had just terminated his care of the patient.7 Dr. Schwartzreich testified that he was told in December, 1985, that the patient had terminated her care with her previous psychopharmacologist and needed to consult someone. Dr. Schwartzreich’s first appointment with the patient, recorded in his appointment book, was on March 21, 1986. There was evidence that, in the course of treating the patient, the petitioner had written prescriptions for antidepressants refillable for up to one year, but no evidence that any prescriptions were written after October, 1985. The patient also testified that the petitioner told her he would not have to charge her for every visit and that she was billed for only three appointments. The evidence, however, established that the patient was billed for every ap*163pointment, and paid for every appointment by personal check.8
From the outset, the duration of the physician-patient relationship was a disputed and crucial point. The patient’s testimany consistently pointed to a physician-patient relationship whose duration encompassed the brief sexual relationship between herself and the petitioner. On this point, without comment, the board declined to credit the patient’s testimony.9 In addition, the patient maintained that the physician-patient relationship had been tainted by improper comments and physical intimacy. There was circumstantial evidence on this point also tending to cast doubt on aspects of the patient’s testimony. The pattern of appointments, which dropped rapidly to bimonthly (and which were short), was not particularly compatible with use of the physician-patient relationship for purposes of seduction, and offered little scope for the “lengthy” personal discussions to which the patient testified. As to the patient’s allegation of physical intimacy in the petitioner’s office, the board noted that her original complaint made no mention of such an incident. The board “decline [d] to make a finding resolving [whether the contact occurred],” but stated that “[t]he absence of a finding on this issue does not reflect any judgment of credibility made in the [petitioner’s] favor.”
In passing on the credibility issue, the board could not properly ignore the fact that it was declining to credit significant portions of the patient’s testimony, some of which, as has been noted, was susceptible of an inference of fabrication *164aimed at establishing that sexual relations occurred during the physician-patient relationship. Not only did the board fail adequately to explain its reasons for crediting the patient,10 but its stated reasons for disbelieving the petitioner also lack support in the record. The board declined to credit his testimony because of “his denial that he ever found [the patient] attractive, [yet] he engaged in sexual relations with her on more than one occasion.” In fact, the petitioner testified that he did not think of the patient as attractive until they had dinner together in October, 1985, when their discussion moved from the professional to the personal.
The board also noted that the petitioner had written the patient’s home and work telephone numbers on his record of her September, 1985, appointment with him, and concluded that this supported the patient’s assertion “that the [petitioner] initiated the contacts leading to their October and November encounters outside the office.” During the hearing, no questions were asked about these telephone numbers scribbled in the margin of the petitioner’s notes. The board noted that the patient’s telephone numbers were not recorded anywhere else in her medical record. There was no evidence whether (as most physicians do) the petitioner generally recorded his patients’ telephone numbers. Any number of inferences may be drawn from the appearance of the telephone numbers in the patient’s medical record; the inference drawn by the board is speculative and, therefore, unsupported by substantial evidence in the record. See Bettencourt v. Board of Registration in Medicine, 408 Mass. 221, 228-229 (1990).
In the face of conflicting evidence, and inconsistencies in the patient’s testimony, the board could not choose to rely on portions of that testimony, reject other, significant portions of it, and fail to explain its reasons for doing so. The board was *165required to confront the problems in the patient’s testimony and to provide “an explicit analysis of credibility and the evidence bearing on it.” Morris v. Board of Registration in Medicine, 405 Mass. 103, 107, cert, denied, 493 U.S. 977 (1989). See also Bettencourt v. Board of Registration in Medicine, supra.
Despite the petitioner’s argument that the applicable ethical standards did not prohibit sexual relations between a psychiatrist and a former patient in 1985, the board expressed the opinion that “it was particularly egregious and unethical conduct for [the petitioner] to engage in a sexual relationship with [the patient] after termination of their professional relationship.” The board’s imposition of sanctions depended, nonetheless, on the patient’s testimony that the petitioner had, as the board put it, “sexualized the doctor-patient relationship through highly personal and inappropriate conversatians in the course of treatment,” before pursuing a sexual relationship. Even Dr. Beck, the expert witness on ethics called by the prosecuting attorney,11 was of the opinion that, under the standards applicable to the practice of psychiatry in 1985, sexual relations with a former patient would be unethical only if the physician-patient relationship had been deliberately used as an enticement to sexual contact. Expert witnesses called by the petitioner were even more emphatic that, absent misconduct during the physician-patient relationship, sexual relations with a former patient did not violate 1985 ethical standards governing the practice of psychia*166try.12 Thus, if the patient’s testimony is discounted, the record lacks substantial evidence of misconduct justifying suspension of the petitioner’s right to renew his license to practice medicine in the Commonwealth. See D’Amour v. Board of Registration in Dentistry, 409 Mass. 572, 581 (1991).
3. Disposition. If the petitioner engaged a patient under his care in personal, sexually suggestive conversation leading to sexual contact that exploited the professional relationship, the board was entitled to conclude that he had violated ethical standards governing the practice of psychiatry. Although in our view there were mitigating circumstances apparently not considered by the board,13 and the point is therefore close, we also conclude that the sanction imposed was not disproportionate to sanctions imposed in other cases of sexual misconduct by psychiatrists,14 and was not an abuse of the *167board’s discretion. See Waisbren v. Board of Registration in Medicine, 418 Mass. 756, 759 (1994).
We do not conclude, on this record, that a judgment or a board decision in favor of the physician is required. The board may wish to remand the matter to the hearing officer for an adequate explanation of the credibility determination. In view of the passage of time since the hearing was held, if the board chooses this route, it must ensure that the hearing officer retains a sufficient memory of the proceedings to make an informed decision on the issue of credibility. The board may choose instead to order a new hearing. There has been unnecessary delay in this matter attributable to the board. Whatever it chooses to do must be done promptly. See Weiner v. Board of Registration of Psychologists, 416 Mass. 675 (1993).
We vacate the decision of the board and remand the matter to the board for additional proceedings consistent with this opinion.
So ordered.
Although we refer in this decision to the credibility determinations made by the board we are, in effect, reviewing the credibility determinations of the hearing officer, see Palmer v. Board of Registration in Medicine, 415 Mass. 121, 124 (1993), whose decision on this point was controlling. See id.; Morris v. Board of Registration in Medicine, 405 Mass. 103, 106-114, cert, denied, 493 U.S. 977 (1989). References to the board, therefore, on issues of credibility are references to the findings initially made by the hearing officer. If the board is dissatisfied with a credibility determination made by a hearing officer, it has the option of remanding the case for a more detailed explanation of the determination, or it may order a new hearing. See Morris v. Board of Registration in Medicine, supra at 106.
Records related to the petitioner’s treatment of the patient formed part of the record of the proceeding. See G. L. c. 233, § 20B (1992 ed.). See also Morris v. Board of Registration in Medicine, supra at 107 n.2.
The extent of the relationship was as follows: The hearing officer’s husband and Dr. Beck’s wife were attorneys with a professional relationship from 1978-1980 and 1987-1988. The hearing officer and her husband had socialized with Dr. Beck and his wife from three to five times from 1978 through 1985. She had not seen Dr. Beck since 1985, and she had never had professional contact with him.
The petitioner styles this claim as a violation of his right to due process. We have treated the board’s obligations and our power of review as matters of administrative law, and do so here. See Bettencourt v. Board of Registration in Medicine, 408 Mass. 221 (1990); Morris v. Board of Registration in Medicine, supra at 312-313. We need not, and do not, consider whether, and if so, when, deficiencies in a credibility determination made by the board might amount to a violation of the constitutional due process rights of a physician appearing before it.
The petitioner testified that the patient initiated all of these meetings. He had a memory of only two occasions of physical intimacy at the patient’s apartment, although he thought it possible that there had been a third.
The patient’s three-year delay in reporting the alleged misconduct does not, in our view, bear on her credibility to any great degree. See Palmer v. Board of Registration in Medicine, supra at 123 (rejecting challenge to credibility based, in part, on ten-year delay in reporting physician’s sexual misconduct).
It was undisputed that the petitioner relocated to New Jersey in the spring of 1986. In anticipation of his departure, he terminated his relationships with all of his private patients between September, 1985, and March, 1986.
As a practicing psychotherapist, it may be inferred that the patient was aware that the provision of services without charge has been viewed as indicative of impropriety in a psychotherapeutic relationship. See Morris v. Board of Registration in Medicine, supra at 112; Matter of Ferrell, Adjudicatory Case No. 89-23-T (July 18, 1990).
Some of this testimony was of events as memorable and “unusual” as the testimony on which the board relied. For example, the patient testified that, while she was in bed with the petitioner, he asked her whether she would still feel comfortable consulting him in his capacity as a physician about her use of antidepressant medications. The board apparently did not credit this statement.
The board explained its decision to credit the patient rather than the petitioner in the following terms: “[The patient’s] ability to recall specific details of the rather unusual statements made by the [petitioner] and her candor in admitting that she reacted positively when the [petitioner] told her that he found her attractive buttress our finding that she was an extremely credible witness.”
Despite the petitioner’s argument to the contrary, Dr. Beck was qualified to testify as an expert on ethical rules applicable to the practice of psychiatry. Dr. Beck did not have to be engaged in the practice of psychopharmacology to testify on ethical constraints applicable to the field of psychiatry as a whole. See Letch v. Daniels, 401 Mass. 65, 68 (1987). There was a sufficient basis in the evidence for the hypothetical questions posed to Dr. Beck by the prosecuting attorney. See Commonwealth v. Merola, 405 Mass. 529, 545-546 (1989). Any omission of relevant facts in the hypothetical questions was appropriately tested on cross-examination. See P.J. Liacos, Massachusetts Evidence § 7.10.2, at 417 (6th ed. 1994).
Before the board, the petitioner argued that the ethical standard prohibiting sexual relations with a patient did not apply to a psychiatrist practicing psychopharmacology. The board concluded (correctly, we think) that the guidelines do not distinguish between psychiatric subspecialties, and are intended to apply to all psychiatrists. Although the petitioner continues to press the distinction between psychopharmacology and psychotherapy as a basis for contending that the penalty imposed by the board was excessive, he appears to have abandoned reliance on the argument that the relevant ethical standards did not apply to him.
On the question of mitigating circumstances, the board appears not to have considered the lapse of time since the events in question and the lack of other patient complaints in the interim. Cf. Palmer v. Board of Registration in Medicine, 415 Mass. 121, 125 (1993). There was no evidence adduced at the hearing that the relationship caused harm to the patient. See Matter of Ferrell, supra. Cf. Matter of Patel, Adjudicatory Case No. 91-11-DALA (Sept. 9, 1992); Matter of Goldberg, Adjudicatory Case No. 89-14-ST (Nov. 1, 1989). In view of the purpose of its sanctions, which is to protect the public, see Levy v. Board of Registration & Discipline in Medicine, 378 Mass. 519, 527-529 (1979), the board also might have considered that the petitioner’s current position does not focus on the treatment of women in the situation of the patient.
We have reviewed a number of the board’s decisions and consent decrees disciplining psychiatrists for sexual misconduct. It appears that in every case, the board has revoked or suspended temporarily the physician’s license to practice medicine.