Over about four weeks in the spring of 1981, patient after patient died in the small intensive care unit of Community Hospital of the Valleys in Perris, a town in Riverside County, California. The deaths came at night, often in the early morning hours, and they followed a pattern that nurses had not seen before. Robert Rubane Diaz, a registered nurse on the night shift, was on duty for every one of them.
Diaz was charged with the murders of 12 patients, 11 at the Perris hospital and one at San Gorgonio Pass Hospital in Banning, where he worked briefly after the Perris unit closed. Prosecutors said he killed them with massive overdoses of lidocaine, a drug used to steady an irregular heartbeat. He gave up his right to a jury, and Riverside County Superior Court Judge John H. Barnard tried the case alone. The judge convicted him of 12 counts of first-degree murder and imposed a judgment of death. On August 31, 1992, the California Supreme Court affirmed the convictions and the death judgment, describing the use of lidocaine to kill patients in intensive care as "unprecedented."
His case left a second mark on American law. The 41-day preliminary hearing at which he was ordered to stand trial was closed to the press and public at his request, and a newspaper publisher's challenge to that closure reached the United States Supreme Court. In 1986 the Court ruled that the public has a qualified First Amendment right to attend such hearings.
The court records say little about Diaz's life before 1981. The California Supreme Court described him as a registered nurse and noted that his own testimony at trial showed he was a trained nurse who understood lidocaine's effects and its proper doses. Forensic scientists who wrote about the case in a 2003 paper for the American Academy of Forensic Sciences described him as a 46-year-old nurse who had worked at two California hospitals. His date and place of birth, his family and his schooling were not reported in the court opinions.
In March 1981 Diaz was working at Chino Community Hospital in San Bernardino County. On March 29, 1981, he began on the night shift in the combined cardiac care and intensive care unit at Community Hospital of the Valleys, which the court called CHOV.
The unit was small. It had six beds, each in a separate room with a large window for watching the patient, and for most of the period it was not full. Two to four nurses worked each night shift. No doctor was assigned to the unit, though one was always on duty elsewhere in the hospital to answer emergency "code blue" calls. Under the hospital's standing orders, registered nurses could give certain drugs on their own judgment, including lidocaine for specified heart-rhythm problems.
The prosecution's theory was that the patients were given lidocaine in amounts far beyond any treatment dose.
The California Supreme Court went through the deaths one by one, recording for each patient the age, the reason for admission, the time of death, what the nurses saw, and what the experts concluded.
Irene Graham, 89, was admitted to the unit on the evening Diaz started work at CHOV, March 29, 1981, with chest pains and congestive heart failure. During the night Diaz told another nurse that Graham had a racing heart rhythm and that he had given her two boluses of lidocaine. She was pronounced dead at 6:15 a.m. Experts testified that the lidocaine in her tissues was extremely high, and the pathologist gave lidocaine poisoning as the cause of death, with heart disease contributing.
Bernard Kean, 52, was admitted at about 1 a.m. on April 3 with chest pain, and doctors thought he might have had a heart attack. He died at 5:07 a.m. The defense said alcohol withdrawal explained his symptoms. The pathologist called the heart attack the primary cause but said a lidocaine overdose had shortened his life.
Beatrice Cline, 79, who had a history of heart trouble, was admitted with chest pain and was alert before the night's events. Diaz gave her morphine and atropine, and three minutes later her condition collapsed. She was pronounced dead at 2:32 a.m. on April 4. Experts found large amounts of lidocaine in her tissues.
Minnie Dempsey, 88, was admitted with chest pain, shortness of breath and congestive heart failure and was conscious on arrival. She died at 7:18 a.m. on April 5. That morning, at the change of shift, Diaz told the arriving nurse that Dempsey was going to "code," paced in visible agitation, and then broke off the briefing to call for emergency help.
Gertrude Bryant, 63, was being treated for high blood pressure and congestive heart failure. She was pronounced dead at 7:33 a.m. on April 7. Once again, the court recounted, Diaz told the same nurse that the patient was "going to code" and seemed excited. The defense pointed to an earlier stroke and to the blood pressure drug she had been given; the experts pointed to a nursing note of an abnormal rhythm and to high tissue levels of lidocaine.
John Rainwater, 95, had severe heart failure and an irregular heartbeat but was alert and stable earlier in the night. Diaz said he would give atropine for a drop in heart rate, then called out that he was giving another dose. Rainwater died at 4:45 a.m. on April 10. Experts concluded that he had received a massive overdose of lidocaine shortly before his death.
Two patients died on April 13. Kenneth Silvera, 82, had come in with chest pain and palpitations. Diaz gave him lidocaine and then increased the dose, and he died at 7:32 a.m. Earlier that night Diaz had gone to another floor to tell a nurse that a patient in the unit was "going sour." Marian Stewart, 80, had pneumonia and respiratory failure and was already receiving lidocaine for an irregular heartbeat. She told staff she was afraid of dying. Diaz told a colleague he was going to give her lidocaine, and about 90 minutes later her heart went into fibrillation. She died at 6:15 a.m.
The court grouped the next two deaths under the dates April 17 to 19. Henry Castro, 66, was admitted with a suspected heart attack. On his first night Diaz telephoned the respiratory staff to say a patient was having problems, and Castro's crisis that night was treated. Two nights later Diaz told a nurse that Castro looked "strange," and Castro's condition failed again. He died at 4:53 a.m. Virginia Bayless, 84, was in a coma after a stroke. Her chart recorded no lidocaine, yet blood drawn days after her death showed very high levels. She died at 1:38 a.m. The defense suggested she might have been given lidocaine in the emergency room; the Supreme Court called that "speculation," and noted that a defense expert agreed she had received a "massive" dose.
Bertha Boyce, 75, died on April 22 to 23. That night Diaz suggested to a colleague that they take their meal breaks early because "something was going to go wrong." After an emergency with another patient, Boyce's condition failed and she died some hours later. By then the hospital had begun requiring that every syringe used to inject a patient be dropped in a collection box. Among the syringes were two nearly empty boluses labeled with Boyce's name in Diaz's handwriting. The labels were for a weak solution, but the liquid left in them tested many times stronger. A pathologist testified that she died of lidocaine intoxication in association with hardening of the arteries.
The 12th patient was Clifford Swanson, 79, at San Gorgonio Pass Hospital. The opinion dates the events to the night of April 25 and the early morning of April 26. Diaz was assigned to the night shift in the hospital's intensive care unit. Nurses testified that they saw Diaz, not an emergency room doctor, at Swanson's bedside shortly before his crisis. A prosecution expert said the heart tracings were consistent with a massive overdose. The defense blamed low blood sugar, lack of oxygen and potassium given during the emergency, but the court credited the nurses.
The prosecution's case rested on several kinds of evidence, which the Supreme Court listed: the lidocaine found in the patients' tissues, far above what their legitimate doses could explain; the sudden seizures followed by heart and breathing failure that the nurses described, which they said resembled epileptic attacks; and heart tracings showing a widened beat, which the prosecution's experts linked to lidocaine overdose.
Beyond the medicine there was Diaz himself. One nurse testified that every night she worked with him he carried a box of lidocaine in his shirt pocket, and that when she asked, he said he wanted it on hand in case it was needed. Another saw a syringe and a vial in his pocket, and once saw him draw medicine from a vial into a syringe. Several nurses described how he seemed to know a patient's crisis was coming before anyone else did.
Then there was the pattern of shifts. The patients died on nights when Diaz was working. "Defendant was the only nurse who was on duty on all of those nights," the court wrote. "None of the other nurses was present on any more than five of the relevant nights."
According to the opinion, a confidential informant told the Riverside County District Attorney's office that 17 patients had died at CHOV between March 8 and April 14, 1981, in circumstances "identical in many respects." Marshall Tolford, an investigator for the district attorney, worked the case with a coroner's investigator, Michael Worthington. Health department records showed 24 deaths at the hospital between March 8 and April 20; Worthington said the usual rate was about three deaths a month. Most of the death certificates listed heart attacks, which made Tolford suspicious.
Six of the recently dead had not yet been buried or cremated, and autopsies were performed on them. Two showed lethal levels of lidocaine. Police seized the records of 24 patients from the hospital. Diaz left CHOV on April 23, and on April 24 the hospital closed the intensive care unit. He had already taken a job at San Gorgonio Pass Hospital, where, the opinion records, two patients died on the night of April 24 to 25 while he was on duty.
Tolford wrote a 10-page affidavit for a warrant to search Diaz's home. On May 1, 1981, police searched the house and seized syringes, an empty lidocaine box and a vial of lidocaine, all of which were later used at trial.
The defense later attacked the affidavit, and the courts found that it contained mistakes. It wrongly said four patients' records showed no lidocaine prescription, and it attributed to a licensed vocational nurse a remark, that she and Diaz were known as the "deadly duo," which the trial court found she had not made in her interviews with Tolford. The trial court ruled that the errors were not deliberate or reckless, and the California Supreme Court upheld the search.
That vocational nurse also told investigators about a syringe Diaz had asked her to give Castro. The manufacturer later reported that this syringe and two others had been tampered with. About seven months after the CHOV unit closed, a construction worker tearing down drywall in the unit found a preloaded syringe in a bed. Its label was for a weak solution; its contents tested many times stronger. It bore the vocational nurse's fingerprint. The prosecution argued that Diaz had planted it to turn suspicion toward her. The syringe was later dropped and broken while in police custody, though the Supreme Court found no showing that any evidence had been lost.
Most of the patients had been embalmed and buried, so no blood was available for 10 of the 12. Their bodies were exhumed. Bayless and Boyce had not been embalmed and were autopsied directly. Toxicologists at the University of Utah's Center for Human Toxicology measured lidocaine in tissue samples, then worked from those figures to estimates of the total amount in each body. They concluded that each patient had received a massive overdose shortly before death.
Diaz was arrested and charged with murdering 12 patients. On December 23, 1981, the state filed a complaint in Riverside County Municipal Court charging him with 12 counts of murder and seeking the death penalty.
The preliminary hearing began on July 6, 1982. Diaz moved to exclude the public under a California law allowing closure of such hearings, and the magistrate, faced with no opposition, granted the motion. The magistrate cited the national publicity the case had drawn and the risk that "only one side may get reported in the media." Most of the prosecution's evidence over the 41 days of hearings was medical and scientific; the rest came from staff who had worked the shifts on which the patients died. Diaz called no witnesses, though his lawyers cross-examined most of the state's. At the end he was held to answer on all charges. The magistrate refused a request from the Press-Enterprise Co., a newspaper publisher, for the transcript, and sealed it. In a dissent Justice Stevens later described that record as 4,239 pages in 47 volumes.
On January 21, 1983, the prosecution asked the Superior Court to release the transcript, and the newspaper joined the request. Diaz opposed it. The judge called the material "as factual as it could be" but found a "reasonable likelihood" that releasing it would prejudice his right to a fair trial, and kept it sealed. While the newspaper's appeal was pending, Diaz waived a jury and the transcript was released.
On December 31, 1984, the California Supreme Court, in an opinion by Justice Broussard, upheld the closure rule. It held that a magistrate must close a preliminary hearing on finding a reasonable likelihood of substantial prejudice to the defendant, and that the First Amendment gave the public no right to attend. Justice Lucas disagreed with the standard and argued that the defendant should have to show at least a substantial probability of prejudice. By then, the court noted, the trial had been completed.
The United States Supreme Court reversed on June 30, 1986, in Press-Enterprise Co. v. Superior Court. Writing for seven justices, Chief Justice Burger held that a qualified First Amendment right of access applies to preliminary hearings as they were conducted in California. A hearing could be closed to protect a defendant's fair trial only on specific findings, made on the record, of a "substantial probability" of prejudice that reasonable alternatives could not prevent. The California standard, the Court said, was too lenient. Justice Stevens dissented, arguing that preliminary hearings, like grand juries, had no tradition of openness.
Diaz's trial was heard by Judge John H. Barnard in Riverside County Superior Court, case number CR-19889, without a jury at either the guilt or the penalty stage. On appeal his lawyers argued that the waiver showed poor representation. The Supreme Court disagreed. The record did not show counsel's reasons, it said, but there were plausible ones: the lawyers may have believed the judge would be receptive, and the waiver was made on the condition that he hear the case; they may have thought a jury would be less open to complex expert testimony across 12 counts; or Diaz may simply have preferred a judge. The defense also argued that a court trial would be shorter, at a time when the chief trial lawyer had heavy administrative duties.
The prosecution's experts tied each death to lidocaine. The defense called its own pharmacologist and cardiologists, who testified that the tissue levels were consistent with the therapeutic doses most of the patients had been given, once their weakened ability to clear the drug and the effects of resuscitation were taken into account. They offered other explanations for the seizures and heart tracings, said some seizures had not been witnessed, and questioned the reliability of the hospital charts. The defense position was that some patients died of their illnesses and others of reactions to properly given medicine.
Diaz testified. He denied poisoning any of the 12 patients and described in detail what he had observed of their symptoms.
The judge also heard evidence about a death with which Diaz was not charged. Estel Jones, 62, had been admitted to Chino Community Hospital after a possible heart attack on March 22, 1981, about a week before Diaz moved to CHOV. On the night of March 24 Diaz was on duty with another nurse, and at 11 p.m. he predicted they "would be having trouble" with Jones. Later he said he would give Jones medication for chest pains, and the other nurse left him alone in the room for about five minutes. When she returned, Jones's condition collapsed. Tissue tests showed a lidocaine level that experts read as a massive overdose, and prosecution experts attributed his death to lidocaine toxicity. The defense did not object to the evidence, and the Supreme Court held that it was admissible to show identity and a distinctive method.
Barnard found Diaz guilty of 12 counts of first-degree murder and found 12 multiple-murder special-circumstance allegations true. He imposed a judgment of death.
The prosecution never put forward a motive. Many of the patients were not regarded as terminally ill, the Supreme Court noted, and the state did not argue that the deaths were mercy killings or suggest any other reason for them. The 2003 forensic paper offered its authors' own view, that Diaz wanted to appear to have a physician's insight into how sick patients were and to predict when they would die. The paper also said he may have been responsible for about 50 more deaths; no court made such a finding.
Because the judgment was death, the case went directly to the California Supreme Court. Justice Kennard wrote the lead opinion, joined by Chief Justice Lucas and Justice Baxter, and the court affirmed on August 31, 1992. Justice Panelli wrote separately, joined by Justices Arabian and George, and Justice Mosk wrote a concurring and dissenting opinion.
Diaz challenged the evidence for each patient. The court rejected every challenge. It found substantial evidence that "an overdose of lidocaine caused the death of each of the 12 patients" and that Diaz gave each fatal dose. It pointed to the Boyce syringes, the syringe found in the bed, the lidocaine he carried, the lidocaine found at his home, and his presence on every one of the relevant nights. "The evidence pointed unerringly to defendant as the killer," the court concluded. It held that murder by poison is first-degree murder without separate proof of premeditation.
The court rejected his other claims. He had waived any challenge to the scientific methods under the Kelly-Frye test for new scientific techniques by not raising it at trial. The charging document did not need to spell out a poisoning theory, since the preliminary hearing had given him notice of it. The errors in the search warrant affidavit did not require suppression. The court did agree on one point: a single multiple-murder special circumstance was enough, so it ordered 11 of the 12 findings stricken as duplicates and otherwise affirmed the judgment.
Diaz kept contesting the science. In the early 1990s he went to court in New York to obtain records held by the Nassau County Medical Examiner. His target was an unpublished 1981 doctoral dissertation, "Lidocaine Toxicity," by W. Christopher Long, then a toxicologist in that office, which measured lidocaine in brain tissue from more than 140 people autopsied between 1976 and 1981. Diaz argued that 24 of them had levels that the prosecution in his case had treated as signs of a lethal dose. Either those people were also poisoned, he contended, or the prosecution's expert testimony was badly flawed.
His petition was dismissed, and the Appellate Division required him to show a "direct and personal interest" in the records. On November 16, 1993, the New York Court of Appeals, in a unanimous opinion by Chief Judge Kaye, reversed. The law required only a "substantial interest," the court said, and Diaz had one in at least some of the records as a matter of law, because the Long study appeared to be the only large study of lidocaine in human tissue. The records "may enable Diaz to show that he was convicted and sentenced to death for murders that did not occur," the court wrote, describing his aim rather than endorsing it. It did not decide whether the material could be used in any later proceeding, and it sent the case back to a Nassau County court to balance privacy concerns against his interest.
As of that November 1993 ruling, Diaz remained under sentence of death in California and maintained that the deaths were not murders.
March 22 to 24, 1981: Estel Jones, 62, is admitted to Chino Community Hospital and dies on a night Diaz is working; Diaz is not charged, but the death is later heard as evidence at his trial.
March 29, 1981: Diaz begins work on the night shift of the intensive care unit at Community Hospital of the Valleys in Perris.
March 29 to 30, 1981: Irene Graham, 89, dies.
April 3 to 13, 1981: Bernard Kean, Beatrice Cline, Minnie Dempsey, Gertrude Bryant, John Rainwater, Kenneth Silvera and Marian Stewart die.
April 17 to 19, 1981: Henry Castro and Virginia Bayless die.
April 22 to 23, 1981: Bertha Boyce dies; two syringes labeled with her name are found to hold the strong solution.
April 23, 1981: Diaz leaves the Perris hospital; the next day it closes its intensive care unit.
Late April 1981: Clifford Swanson, 79, dies at San Gorgonio Pass Hospital in Banning.
May 1, 1981: Police search Diaz's home and seize syringes, an empty lidocaine box and lidocaine.
December 23, 1981: A complaint charging 12 counts of murder is filed in Riverside County Municipal Court.
July 6, 1982: The preliminary hearing opens; it runs 41 days behind closed doors, and Diaz is held to answer.
January 21, 1983: The Superior Court refuses to release the hearing transcript.
By December 31, 1984: Diaz has waived a jury and been tried before Judge John H. Barnard, convicted of 12 counts of first-degree murder and sentenced to death; the California Supreme Court upholds the closure of the preliminary hearing.
June 30, 1986: The United States Supreme Court reverses in Press-Enterprise Co. v. Superior Court.
August 31, 1992: The California Supreme Court affirms the convictions and death judgment and strikes 11 duplicate special-circumstance findings.
November 16, 1993: The New York Court of Appeals rules that Diaz has a substantial interest in records of a lidocaine study held by the Nassau County Medical Examiner.