The women who say Dr. Robert John Boyda Jr. sexually assaulted them after surgery may have had one thing in common. All 20 might have been under the influence of an anesthetic agent known -- in rare instances -- to cause sexual hallucinations.

The doctor's defense attorney has hinted he might make that argument if Dr. Boyda, of Mt. Lebanon, goes on trial for the assaults, which allegedly occurred at his Robinson and Scott offices from 2002 to 2007.

After closing arguments today in a protracted preliminary hearing, a district judge in Scott is expected to decide whether the oral surgeon should be held for trial on charges that he sexually assaulted and exposed himself to his patients while they were anesthetized .

Questioned by Deputy District Attorney Janet Necessary, many of the accusers said they remembered the sex abuse clearly, but most couldn't recall the details of being prepped for surgery. Details of the allegations against Dr. Boyda also are strikingly similar, coming from patients who ranged from 13 to 57 years old.

If Dr. Boyda used propofol, a drug widely used to sedate patients, or ketamine, also used to induce or maintain general anesthesia, or diazepam (generic for Valium), or a short duration sedative like midazolam (generic for Versed), his defense lawyer might try to cast doubts for the judge or jury by exploring the fine print on drug labels. He could argue that proper medical use of the drugs in proper dosages caused the victims to imagine vivid attacks amid the haze of anesthesia.

Attorney William H. Difenderfer hinted at this defense after 16 victims took the stand at one of Dr. Boyda's preliminary hearing dates in March. "When you're under this anesthesia, you do have hallucinations," he said.

The label on propofol mentions "amorous behavior," "abnormal dreams" and "hallucinations" as possible side effects of the drug. Journal articles say that under propofol any stimulation to the chest, like the removal of electrodes or the inflation and deflation of a blood pressure cuff, can trigger hallucinations of a sexual nature.

The fine print on midazolam warns about the possibility of "hallucination," "emergence delirium" and "dreaming during emergence." Diazepam notes "delusions (holding false beliefs that cannot be changed by facts)" and "changes in sexual desire." Ketamine includes a "special note" on its packaging, stating that in about 12 percent of patients "psychological manifestations vary in severity between pleasant dream-like states, vivid imaginary hallucinations, and emergence delirium." But it says "these reactions may be reduced if verbal, tactile and visual stimulation of the patient is minimized during the recovery period" or the doctor uses a low dosage.

In cases where doctors can verify that patients experienced hallucinations and no sexual contact occurred, the patients still have "a very vivid perception of what happened," said Dr. John Butterworth, chair of anesthesia at Indiana University School of Medicine, who has studied reports of sexual abuse during anesthesia. "They're not making it up. They do perceive it. It's a mistake to blame the victim, even if it's a hallucination."

The side effects are generally caused by drugs being used for conscious sedation (not full general anesthesia) or are "a consequence of misperceptions during induction or emergence from general anesthesia," Dr. Butterworth said.

Clarifying the blurry line between dreams and reality is tough in a legal proceeding. On the preventive end, doctors say the phenomenon does not easily lend itself to scientific study: You can't have a test group where you assault some patients and not others.

Dr. Erin Sullivan, who teaches anesthesiology and runs the cardiothoracic anesthesiology department at UPMC Presbyterian, said complaints regarding sexual encounters after anesthesia "are not a prominent concern" because doctors hear about them so rarely. She said, "I've not had patients tell me that this is what's happened to them. But it may just be rare because patients don't report it."

"I feel it would be inappropriate to ask" about those experiences, she said. "But if a patient volunteered the information, I would explain. We obtain informed consent. I don't go through all the side effects. That's not the standard of care."

Dr. Butterworth and several colleagues have documented in journal articles that "the problem of sexual ideation" goes back over 150 years, to the earliest cases of anesthetized patients.

Medical journals note that with modern medicine, this side effect, though uncommon, has resulted in false and career-crushing allegations against practitioners; it also has prompted some doctors to take advantage of drugs' "amnesic effects" to cover up actual assaults on patients. Doctors in several European countries have been convicted of sexual assaults under these circumstances, according to the British journal Anaesthesia.

Dr. Glen Gabbard, director of Baylor Psychiatry Clinic in Houston, stated in a 1991 American Health article that sexual exploitation of patients is "a very common phenomenon" and that the number of false accusations is "infinitely small" compared with cases of real abuse.

However, the same article notes that Dr. John Dundee, an anesthesiology professor at Queen's University Belfast in Northern Ireland, studied 35 cases of hallucinations, in mostly female patients who had been sedated with diazepam or midazolam: "Most, he found, had a 'disturbing sexual element' -- usually breast or genital fondling. Yet Dundee maintains that in most cases sexual contact would not have occurred -- because others were present or because the allegations were actually 'physically impossible.'"

More recently, a Warwick, R.I., anesthesiologist who used propofol was acquitted on charges that he touched a patient's breasts during knee surgery. He admitted he'd massaged the woman's shoulders, as she alleged. He lost his license to practice. An emergency room doctor in Ottawa, Canada, also lost his license, but was acquitted on criminal counts of forcing a patient to squeeze his penis -- he said he'd asked the patient to squeeze two fingers of his hand, to see how awake she was.

Overall in 2007, 10 doctors licensed in Pennsylvania were disciplined for sexual misconduct, according to the state licensing board. The state doesn't specify what the circumstances were in those cases.

In cases like Dr. Boyda's, it is difficult to prove what did or did not happen in a closed exam room, especially when one party is a trained physician and the other is undeniably, and by express permission, drugged.

"You don't want someone to be using anesthetic agents as a way to satisfy some sort of perverse agenda," said Dr. Butterworth. "On the other hand, no one wants to falsely accuse someone of this, because it's such a terrible thing." He conceded that in his review of the phenomenon, he'd never heard of a case involving as many victims as the Boyda case does.

He noted a case from the journal Anaesthesia involving an anesthesiologist who underwent surgery while her spouse and six or seven other physicians were present in the room. She had a vivid recollection of someone sexually assaulting her in the operating room. But everyone, including her husband, assured her it had not happened.

For that reason, Dr. Butterworth said, "when I use these drugs for sedation, I'm always surrounded by people." Having nurses or other "chaperones" in the room is essential. The British journal articles also suggest that doctors meticulously document patients' reports of sexual abuse under anesthesia.

Patients coming off sedation are "vulnerable and you don't want anything to happen to them," Dr. Butterworth said. At the same time, he said, "the reason they're vulnerable is also the reason they're not the ideal witness" if they have to testify at criminal proceedings.

The defense might lean heavily on the credibility of Dr. boyda's support staff to account for what happened in each case.

Wendy Murphy, a former prosecutor and adjunct professor at the New England School of Law who teaches about date rape drug prosecutions, said she'd be happy to prosecute a case like Dr. Boyda's.

Facing this kind of defense, she said, she might ask an expert witness whether the medical journal studies specified that patients' hallucinations involved sexual misconduct by the doctor.

"The good news is there's 20 victims. So no matter what the literature says about sexual hallucinations, there won't be a single anecdote, let alone a study, that says that many patients were assaulted.

"If I'm the prosecutor, I'm not going to fight over whether they were hallucinations. I'd say the nature of this drug causes the patient to be incapacitated. Medical patients were in no position to consent to any kind of contact," she said.

Dr. Butterworth, who has been deposed as an expert witness on this phenomenon, said, "The case reports indicate that [sexual abuse] does occur, which puts the justice system in a terrible quandary. In our system, criminal cases have to be proved beyond a reasonable doubt. When you know the drug can cause hallucinations, that's a very, very high barrier."