The Defence of Innocence 1990

Article published in Edward L. Greenspan, Q.C. ed., Counsel for the Defence (Toronto: Irwin Law, 2005)
Author: Austin Cooper, Q.C.

Between January 11 and March 22, 1981, four infants who were patients in the cardiac ward at the Hospital for Sick Children in Toronto died as a result of the deliberate administration of overdoses of a heart drug called Digoxin. Although, at the outset, there was some speculation that “mercy killing” was the reason for the deaths, it soon became clear that there was no motive for the killings. In fact, some of the babies could have recovered from their illnesses and led reasonably normal lives.

The usual reasons for one person killing another, such as matrimonial discord, jealousy, revenge, self-defence, drunkenness, or provocation, were all irrelevant. There was no apparent motive. Accordingly, these crimes were strange indeed, and, as the lawyer retained to defend the person charged, it appeared plan that the killer who snuffed out the lives of the babies had to be a strange person. Any normal, right-thinking, healthy person would not extinguish the lives of four helpless infants in their cribs for no apparent reason. It was therefore with some curiosity that I anticipated the first meeting with my client, Susan Nelles, at the Metro Detention Centre in March, 1981.

What sort of creature would I see? The first thing that struck me on that cold Sunday afternoon in the little room assigned for lawyers to meet their clients was how small she was. She appeared barely five feet tall. Apart from that, she certainly was not typical of the young women I had seen within the confines of prisons. There was a freshness about her appearance and a polish about her manner that set her apart from other inmates I had met.

As we sat at the table provided for us and discussed her pending application for bail, I noticed that she was clear eyed, reserved, polite, considerate, and above all, apparently quite sane; at least to my non-medical eye.

Later, while interviewing the witnesses to be called at her application for bail before Mr. Justice Steele, I was impressed by the quality of her friends. Her best friend and confidante with whom she shared an apartment was Alison Woodbury, a law student and a person whose wholesome manner and integrity were immediately apparent. She is now a lawyer with a large firm.  The evidence she gave at the bail hearing was impressive and it reinforced my initial reaction to this case: it was highly unlikely that the strange person expected to have committed such senseless and bizarre crimes would have friendships with people of Alison Woodbury’s quality.

If one assumed that the person who had committed these unusual crimes had to be an aberrant personality; a twisted, warped individual; the staff in our office quickly concluded that Susan Nelles did not appear to fit the role. For that reason alone (and, later, there were others), we began to speculate that perhaps the police had arrested the wrong person.

To test our thesis, we decided to have her examined by a psychiatrist of some eminence and by a senior psychologist on the staff of a major hospital. We wanted professionals whose qualifications, expertise, and impartiality would be unquestioned, so we turned to a psychiatrist and a psychologist who, although perhaps not well known around the courts, were the chiefs of their departments at teaching hospitals in Toronto. The chosen psychologist was Dr. Leonard Goldsmith at the Toronto General Hospital. He administered a battery of psychological tests to Ms. Nelles and gave us a written report that concluded she was the most well-adjusted person it had ever been his privilege to examine. We felt good about that.

We also arranged to have her examined by Dr. Stanley Greben, who was chief of the medical staff and head of the Department of Psychiatry at Mount Sinai Hospital, in Toronto, a former present of the Canadian Psychiatric Association, and a professor of psychiatry at the University of Toronto. He had an impressive background of scholarship and experience and had written extensively in his specialty. In his written report, Dr. Greben concluded that Susan Nelles was sane, socially well adjusted, and a person to whom the killing of other humans, particularly infants, was an abhorrent concept. She was a nurse, and her instincts and outlook were focused on comforting and healing, not killing. These reports further confirmed our first impression that Miss Nelles did not fit the role cast for her by the prosecution. She simply was not the type of person who could be expected to murder helpless infants for no apparent reason.

What did the evidence show? Did it bear out our thesis that the wrong person was accused of four murders of the first degree? What contact, if any, could the prosecution prove Susan Nelles had with Janice Estrella, who died on January 11, 1081; Kevin Pacsai, whose death occurred on March 11; Alana Miller, who died on March 21; and Justin Cook, who died a day later on March 22?

Before the preliminary hearing, we conducted a detailed analysis of the hospital records. It demonstrated that Miss Nelles was one of the five-member nursing team on ward 4A, the team on duty when all four children died. Further, she had been assigned to care for babies Pacsai, Miller and Cook during the long night shifts when each of them died; she had access to each of those three children during the seven to eight hours before their deaths. However, the records also showed that at various times during those shifts other nurses had relieved her and, therefore, had access to those three infants.

Our analysis of the nursing notes and hospital records relative to the fourth child, Janice Estrella, showed that Susan Nelles was not on duty during the night shift for the seven or eight hours before her death. We searched the records for the preceding twelve-hour day shift and found that although Miss Nelles was on duty as part of the nursing team, two other nurses had been assigned to care for the baby under circumstances of constant nursing care, meaning one of those two nurses always had to be present and tending to the child. Whenever one nurse took a break for coffee or lunch or to use the washroom, the other had to be there. In other words, the records indicated that Susan Nelles had no contact with Janice Estrella during the twenty-hour period preceding her death. Accordingly, our preliminary analysis of hospital and nursing records tended to confirm our impression that perhaps the police had the wrong person. We began to think that the prosecution was in trouble, and we had a reasonable chance of having the charges tossed out at the preliminary hearing. We hoped the oral evidence at the hearing would bear that out.

The preliminary hearing commenced on January 4, 1982, before Judge Vanek of the Provincial Court. As the evidence unfolded relative to Janice Estrella, it was established that the child, who had been prescribed Digoxin, had fatal levels of the drug in her blood at the time of her death. It was confirmed that Susan Nelles was off duty during the night shift when the baby died, that the two nurses assigned to the baby on the preceding twelve-hour day shift had watched her continuously, and that Susan Nelles had not been near her. We felt that the prosecution was truly in deep water. We also began to wonder who had killed baby Estrella. It was never determined who administered the fatal dose of Digoxin to Janice Estrella, but it became clear that it could not have been Susan Nelles. She simply had no access to the child in the twenty hours during which the fatal dose of Digoxin had been administered.

From the outset and throughout the proceedings we contended that the same person had to be responsible for all four killings. The crimes all followed the same pattern:

  • all involved infants on the cardiology ward’
  • all were killed by Digoxin overdose;
  • there was no motive for any of the murders;
  • all died early in the morning and while members of the same nursing team were on duty;
  • all had to have been killed by an unusual assailant;
  • one could not expect, in the light of common human experience, that people operating independently of each other could have had the strange need to snuff out infants in this way; and
  • there was absolutely no evidence of a conspiracy among two or more persons to murder in this manner.

Ultimately, the court accepted this thesis, as did the Crown in its argument. We felt that the facts proved in relation to baby Estrella should put an end to the case against Susan Nelles. However, the prosecution turned that argument around and submitted that, because the crimes could have been committed by only one person, if there was some inference that Susan Nelles had access to and perhaps killed Miller, Pacsai, and Cook, the Court could infer somehow she had obtained access to Estrella and murdered that baby too.

Fortunately, a month or so into the preliminary hearing the defence received further support when we learned that, following an order of the attorney general, the body of a forth infant, named Stephanie Lombardo, had been exhumed. Baby Lombardo had been a patient on the cardiac ward for five days during the latter part of December 1980. She died suddenly on the ward while recovering from an apparently successful heart operation. Lombardo had not been prescribed Digoxin, yet, when her body was exhumed in December 1981, a year after her death, an analysis demonstrated that her tissues were laced with Digoxin. The expert evidence confirmed that baby Lombardo very likely died as a result of a Digoxin overdose. The hospital employment records and the nursing records for December 1980 showed that Susan Nelles was on holiday during the whole period in which Lombardo was in the hospital. This fact was confirmed at the preliminary hearing by the oral evidence, we felt that the case against Susan Nelles had all but disintegrated and that we had an innocent client. Ultimately, Judge Vanek agreed in his ruling on May 21, 1982, when he gave his reason for discharging Susan Nelles: Since admittedly the same person must have killed all four children, the evidence relating to Estrella is positive disproof that [Nelles] is the person who poisoned Cook, Pacsai and Miller. Moreover, the Crown’s theory that Susan Nelles is the person who killed all four babies is inconsistent with the circumstances surrounding the death of Stephanie Lombardo.

During this enormously complicated but immensely interesting case, we faced several tactical and evidentiary problems. Here is now we tried to deal with them. We realized from the outset that the legal test by which the judge presiding at the preliminary hearing had to guide himself in determining whether to send the case to trial was a narrow one. He did not have to be convinced beyond a reasonable doubt before he could commit for trial. The test that Judge Vanek was required to apply was enunciated by the Supreme Court of Canada in, U.S.A. v. Sheppard was there evidence on which a reasonable jury, properly instructed, could convict the accused? During the three-and-a-half-month preliminary hearing, we had to be sensitive as to whether some evidence might be led against Susan Nelles on which a reasonable jury could convict. We realized, on the authority of Sheppard, it did not have to be very much evidence for Judge Vanek to conclude that the case ought to be sent to trial.

Some areas of the evidence caused us concern in this regard, such as the evidence of Dr. Rodney Fowler. He was the staff cardiologist and chief of the cardiac ward during March 1981, and he testified that he saw Miss Nelles in the early morning of March 22, 1981, shortly after Justin Cook had died. Cook was the last of the four children to expire. As Dr. Fowler was leaving the hospital, he saw Miss Nelles sitting at one of the desks in the nursing station apparently writing up the final report in Justin Cook’s medical chart. (She was the nurse assigned to that child during the evening shift on which he died.) Dr. Fowler gave evidence that he knew she had been previously involved with baby Pacsai and had given the prescribed Digoxin to that child before his death. By that time, it had been confirmed that Pacsai had died of a Digoxin overdose because an inquest had been made into the death. Dr. Fowler was anxious to see what Miss Nelles looked like in the early morning after Cook’s death. He said he glanced in her direction and saw a very strange expression on her face, one showing no sign of grief. He thought it was unusual for her to have this appearance after such a terrible thing had happened. We appreciated that this evidence was tenuous but felt it could prove troublesome and might be the straw that would convince the judge to let a jury decide the case in a trial. As a result, we prepared Dr. Fowler’s cross-examination very carefully. The cross-examination proceeded this way:

  • Now, as to Susan Nelles, was she a close friend of yours?
  • No.
  • Did you ever visit her house?
  • Never.
  • Or her apartment?
  • Never.
  • Has she ever visited your house?
  • Never.
  • I think you said you knew her for one year?
  • Just because she was on the ward.
  • On the ward?
  • Her brother is a resident on the ward and I knew her father.
  • Have you ever had lunch with her?
  • Never.
  • Or coffee?
  • Never.
  • Or dinner?
  • Never.
  • Ever had a sort of meaningful conversation with her about anything other than whether a baby has or hasn’t had its medication or has turned blue?
  • Never.
  • Okay, so you never worked a twelve-hour shift in her company?
  • Never.
  • Did you ever see her since her arrest on March 25 other than around the court or whatever?
  • No, I’ve never seen her since then.
  • Did you ever see her grieving after a relative had died?
  • I have never seen her grieving because I don’t know her. I’ve never seen.
  • Did you ever see her upset, ever?
  • Never.
  • Ever see her cry?
  • No.
  • Did you ever see her angry?
  • No.
  • Did you ever see her depressed or elated?
  • No.
  • Did you ever see her sad or shocked?
  • No.
  • Well, I’m going to suggest that you really don’t know much about the lady’s emotional range. You’ll agree with that?
  • I’ve seen many nurses who are looking after sick patients, and her reaction, again, was very unusual.
  • Well, I didn’t ask you that. I asked you, I suggest that you don’t know much about her emotional range?
  • No.
  • You just don’t know much about it?
  • No, except that it’s unusual, under the circumstances, for a person to have that reaction.
  • Okay. Well, a number of witnesses here have described Susan Nelles as being a cheerful person. Do you know her well enough to even know about that?
  • No.
  • She’s easy to get along with; Do you know her well enough to even know that?
  • No. I don’t work with her.
  • You don’t work with her. She’s conscientious and eager. You don’t know that, I guess?
  • No, I don’t know that. I presume that she does her job well or she wouldn’t have been working on the ward.
  • Do you think those people, whether nurses or doctors, who have worked on shift with her, would be in a better position than you to describe her emotional states. Someone who knew her well and worked with her on shift.  Do you think?
  • I’m not talking about her emotional state, I’m talking about her appearance under a situation, and I explained what she looked like. I don’t know anything about her reactions, her emotional reactions.
  • Well, my question was, do you think those who have worked on shift with her over a period of weeks or a year, twelve hours a day, would be in a better position than you to describe Susan Nelles’ emotional state?
  • Well, I’m not sure about that.
  • You’re not sure about that.
  • Because they haven’t seen her in exactly the same situation that I saw.
  • Oh, I see. Well, we’ve had some witnesses here who spoke about that. There was a Mrs. Ober who gave evidence in this courtroom, and I’ve got it as exactly as I can that “After Justin Cook’s death, Susan Nelles looked tired and strained and pretty upset. Is that the type of reaction you would have expected from a nurse after a death?
  • Yes.
  • We have Mr. Cook himself, Justin Cook’s father, who gave evidence here just a couple of weeks ago. He said, ” There were tears in Susan Nelles’ eyes and she said she was sorry. Is that the type of reaction you would have expected from a nurse after a death of an infant that she was looking after.
  • Yes, that’s what I would have expected.
  • You would have expected that. Another nurse, Susan Reaper, said in this courtroom under oath, after the death of Justin Cook “Susan Nelles was upset. She was smoking a cigarette.” Is that the type of reaction you would have expected?
  • I’m not sure. I don’t know.
  • Mrs. Lyons, Evonne Lyons, one of the nurses on 4B, gave evidence in this courtroom that “Susan Nelles, after the death of Justin Cook, seemed to be in shock”. Is that the type of reaction you might expect to see after the death of a child who you were looking after?
  • I don’t know what that term means, I’m not sure.
  • Well, you said she looked strange. Do you know what that term means?
  • Yes.
  • Phyllis Trayner, who is the same lady Mr. McGee asked you about and whom you said you knew by sight anyway, the head nurse?
  • I just know her by sight, just the way I know Miss Nelles.
  • She swore in this courtroom, just a couple of weeks ago, that, “After Justin Cook’s death, Susan Nelles was very upset. There were tears in her eyes.” That’s the type of reaction you would have expected?
  • That’s what I would have expected, but there was no evidence of that at all when I witnessed her.
  • When you witnessed her. Bertha Bell, a nurse on 4B swore in this courtroom some short time ago said that “After the death of Justin Cook, Susan Nelles was upset. She was very quiet. I know Susan Nelles well, I know when she was upset. That’s the sort of thing you’d expect after the death of a child you were looking after?
  • Yes.
  • And Mrs. Whittingham, who was also on duty, I believe on 4B, said, “Everybody was upset and feeling badly after Justin Cook’s death, including Susan Nelles.” Again, that’s how you’d expect the nurses to react?
  • Yes.
  • Right. Now you swore that you didn’t see Susan Nelles, according to my notes, until you were leaving the ward?
  • That’s correct.
  • After the death of Justin Cook?
  • Yes.
  • So you didn’t see her during the arrest or right afterwards, did you?
  • No.
  • You might not have had the same opportunity as these other people whose evidence I’ve outlined for you?
  • Yes.
  • Right. You took a glance at her, I think you said?
  • Yes.
  • She was seated at a desk, writing out a description of what happened?
  • Well, I don’t know what she was writing, but she was writing something.
  • I thought you said she was writing a report?
  • No, I don’t lnow what she was doing. I presume it was a description of the final events, but she was writing something at the desk.
  • Well, it would be her duty to write out such a description, wouldn’t it?
  • Yes.
  • Yes. She was looking down at a piece of paper while she wrote?
  • Yes.
  • Apparently concentrating?
  • Yes.
  • And you felt from your glance at her as you left the ward that her expression was strange?
  • Yes, it was not in keeping with a person who had just had a patent die under her care.
  • So it was strange in that there was no appearance of grief?
  • Of grief, right.
  • That’s why it was strange, is that correct?
  • Correct.
  • Are you really in any position to tell this Court how she was really feeling at that time. Are you?
  • All I can do is tell the Court what I saw.
  • What you saw. You have no idea what she was feeling, do you?
  • That’s an inference that you’ll have to make yourself.
  • Well, you just don’t know, do you, how she was feeling?
  • No, I just know what she looked like, and I’ve seen many, many people who were in that situation who don’t appear like she did. In other words, she had no signs of grief.
  • She had no signs of grief and that was strange to you?
  • Yes.
  • Right. But you don’t know how she was feeling. I mean, are you prepared to agree with that or not?
  • Yes, I can’t say that.
  • You can’t say. And you certainly don’t know how she was feeling during the arrest or when the baby was pronounced dead or immediately afterwards when talking to her nurses on the floor, do you?
  • No. I didn’t speak to her at all during that period.

Judge Vanek concluded his reasons for discharging Ms. Nelles with these words: “I am unable to find any evidence of guilt from what a doctor thought from a passing glance was a ‘strange expression’.”

It was also an area of concern for us that, apparently after the arrest of Ms. Nelles on March 25, 1981, there were no deaths of infants that could be attributed to a Digoxin overdose. We were concerned that Judge Vanek might conclude that this fact led to an inference of complicity by Miss Nelles, enough to compel him to send the case on for trial. However, we were able to elicit some facts in cross-examination that we felt were able to neutralize any such inference. First, after the death of Justin Cook, all Digoxin was locked up in a narcotics cabinet. Before that, it had been readily available to anyone who sought it. As well, after Cook’s death, all Digoxin had to be signed for by two nurses on an inventory sheet, and the inventory control was tightened considerably. Since Digoxin was a prescription drug, it became very difficult for any person to obtain access to it, certainly without considerable risk of exposure.

Second, we argued that, with the arrest of Miss Nelles, any further killings by Digoxin would attract attention to the real offender still at large. That person might be expected to keep a low profile.

Third, the evidence revealed some bizarre happenings at the hospital after the arrest of Susan Nelles. They tended to support our contention that someone with a twisted personality, the type of person who might have caused the senseless killings charged to Susan Nelles, might still be on the loose in the ward. We were able to adduce through cross-examination of members of the nursing staff that a series of ominous and threatening telephone calls had been made to nurses and members of the hospital staff and their families after the arrest of my client. Someone, a female, who was familiar with the deaths of the babies and with the nurses on duty, was calling people at the hospital for no apparent reason and was making death threats to two particular nurses on the ward. In addition, we were able to show that a series of strange marks in lipstick was made on the locker doors of the same two nurses on the cardiac ward, on the door of the apartment of one of them, and on the rear window of a car.

We established through alibi evidence that Susan Nelles was unable to have made some of those telephone calls, for example, she was on an airplane to Vancouver when one of them was made. She was out of Toronto when some of the lipstick markings were placed, a bridesmaid at a wedding in Ottawa, for instance, when one of the marks had to have been drawn.

Another bizarre incident occurred at the hospital after the arrest of Miss Nelles when two nurses found Propanolol, another heart drug, in the food they had brought to the hospital from their homes. During a 2:00 morning lunch break, one nurse found the drug in her soup, which was in a Tupperware container, while the other nurse found the drug sprinkled in her salad in a similar plastic container.

We were able to establish through cross-examination of most of the witnesses called for the Crown that Susan Nelles had not been in or near the hospital since the day of her arrest. We were therefore able to argue that some human being with an aberrant personality was loose both in the hospital and outside. A disturbed person was still operating, undeterred by Nelle’s arrest. We argued that this evidence was consistent with the fact that the real killer had not been apprehended. Judge Vanek concluded on this point:

Any inference of Nelles’ guilt from the fact that there was a decrease in the number or deaths on Wards 4A and 4B after her arrest is offset by equally valid inferences of innocence for the reasons previously stated [i.e., the locking up of all Digoxin and the realization by the killer that he/she might attract attention by further killings] and the long list of unusual incidents and mysterious happenings at the hospital.

After we learned that the exhumation of baby Lombardo had confirmed that her body was loaded with Digoxin, although not prescribed, and we confirmed that Susan Nelles was on holiday for the five days that Lombardo was in hospital, we could see that the prosecution was demoralized. Their case was disintegrating. We met with the prosecution and police at the El Toro Steak House in Toronto for dinner. We gave them copies of our psychiatric and psychological reports that indicated Susan Nelles was sane and well adjusted. We suggested that the prosecution be stopped. We told the prosecution that we intended to file those reports or call the doctors who wrote them at the preliminary hearing to show that Nelles was not an aberrant personality. We felt they would be potent evidence at the hearing.

The Crown responded by requesting an examination of our client by a psychiatrist. We said we would allow it, but we wanted the examination to be by doctors who were not usually consultants to the Crown at that time. We asked for doctors who were staff at teaching hospitals in Toronto and professors or associate professors of psychiatry at the University of Toronto. We gave them a list of seven or eight names of suggested doctors who had those qualifications. However, the Crown would not accept any of the doctors on the list, so we were never able to agree on a doctor to conduct the examination. In any event, the Crown was not prepared to stop the prosecution, even if a psychiatric report obtained by them confirmed that Susan was sane and well adjusted.

In the end, we decided not to file our psychiatric and psychological reports. We were concerned that an adverse inference might be drawn against us because we were unwilling to permit a doctor selected by the Crown to see her. We had in mind the case of R. v. Sweeney, a murder case in which the defence was insanity.  The defence had psychiatric reports, and the Crown had asked for a psychiatric examination of the accused.  The defence agreed on the condition that the Crown psychiatrists confer with the defence psychiatrist before examining the accused.  This offer was declined by the Crown, so no examination was done.  At the trial, however, the defence proceeded to call its psychiatric evidence in support of the defence of insanity.  The Crown then led evidence of the refusal of the defence to permit their doctors to examine the accused.  The accused was convicted and subsequently appealed to the Ontario Court of Appeal.  That court held that the evidence of the refusal of the defence to permit the psychiatric examination was relevant and admissible:  it implied that the defence of insanity was either contrived or weak.  Mr. Justice Zuber said in the decision:

It is, after all, the accused who has raised the defence and made his sanity an issue. Can it be said that he has the exclusive right to call psychiatric evidence and also to deny the prosecution even the ability to explain why the Crown has called no evidence to meet this issue? In any view, the rational conclusion in this case is that the impugned evidence was properly admitted as a fact to be weighed in assessing the merit of the defence.

In light of this decision, we decided to keep our psychiatric reports in our file. We knew our case was very good, and we did not want to provide the Crown with an iota of an inference that the judge might consider evidence within the rule in U.S.A. v. Sheppard.

At about the same time, the prosecution requested a lie detector examination of Susan Nelles. We were still hoping they would withdraw the charges. We wanted to convince them to do so, but we had doubts about accuracy of the lie detector. Eventually we agreed to permit her to undergo a lie detector examination on two conditions.  First, if the results were negative, those results could be filed on consent before Judge Vanek.; They would provide the final nail in the coffin that the Crown’s case had become. Second, if the results were positive (i.e., the machine was in error), no reference could be made to that fact at the preliminary hearing.  The prosecution refused to accept our conditions, and no lie detector test was administered.

In a long judgment on May 21, 1982, Judge Vanek found that the circumstantial evidence tendered against Susan Nelles was not sufficient to meet the test laid down in U.S.A. v. Sheppard and that she should be discharged on all counts.  In his reasons, the judge went further and, in effect, exonerated her from all complicity in the murders by stating, “There’s evidence that points in a different direction.” Subsequently, Mr. Justice Grange, after a lengthy intensive public inquiry, confirmed that Susan Nelles was indeed innocent and suggested that the Province of Ontario pay her compensation.

Overall, I found the case a fascinating one because of the unusual and sometimes bizarre facts we had to deal with, the sensitive tactical decisions we had to make, and the daily courtroom drama in which I was involved.

Since 1953 Cooper, Sandler, Shime & Schwartzentruber LLP has been specializing in all areas of Criminal Law in the Toronto area. Our Toronto criminal lawyers at Cooper, Sandler, Shime & Schwartzentruber LLP serve the Greater Toronto Area and Ontario, Canada.

Cooper, Sandler, Shime & Schwartzentruber LLP
439 University Ave. Suite 1900
Toronto, Ontario M5G 1Y8