At the Caressant Care home in Woodstock, where she worked as a registered nurse from June 2007 to March 2014, Wettlaufer confessed to killing seven elderly residents. Inspectors learned that another 180 residents died during her shifts, according to ministry documents obtained by the Star through a freedom-of-information request.

Wettlaufer’s method of killing was a lethal injection of insulin, which can result in many hours of physical distress before the victim dies. So, inspectors also searched Caressant files for residents who died within 24 hours of the killer’s shifts, and found 55 of them.

At the Meadow Park nursing home in London, where Wettlaufer worked for about four months after being fired by Caressant, the same exercise was conducted. They discovered that 14 residents died during or shortly after her shifts, including the resident she confessed to killing, Arpad Horvath.

Wettlaufer did her killing on her eight-hour evening and night shifts. At night she was the only registered nurse on site. Was she responsible for any of the other 248 deaths that occurred on or near her shifts, deaths she didn’t confess to?

Wettlaufer has denied killing more people. But the striking number of deaths on her shifts reveals the full investigative challenge police faced when, in September 2016, she confessed without prompting to killing eight nursing home residents, and assaulting or trying to kill six others.

The numbers also reflect the importance of data collection for raising red flags in nursing homes. The commission examining what allowed Wettlaufer to continue killing undetected is likely to make more data collection part of its recommendations. It closely examined, for example, a new ministry study that for the first time flagged nursing homes where the ratios of unexpected deaths are highest.

Read more:

The nurse on the night shift

Warnings early in Wettlaufer’s career

Nurses college under fire over killer’s case

But the commission showed little interest in specifically exploring the tracking of deaths during nursing shifts, despite testimony from expert witness Beatrice Crofts Yorker, a California professor of nursing and criminal justice, who conducted a landmark study on health-care serial killers worldwide. She noted that tracking deaths that occur in the presence of specific health-care workers can result in life-saving red flags.

When this exercise is applied to Wettlaufer, the results compiled by the Star seem striking.

Between July 2011 and June 2012, for example, 50 residents died at Woodstock’s Caressant Care home, including residents who died within 30 days of being taken to hospital. The death/shift data obtained by the Star indicates that 37 residents died on Wettlaufer’s shifts during that 12-month period, including three she confessed to killing.

Ontario Provincial Police Staff Sgt. Carolle Dionne told the Star that officers who investigated Wettlaufer “identified all those that passed away during Wettlaufer’s shift and/or in the timeframe soon after her shift.”

They reviewed medical files, interviewed nursing home staff and consulted coroners, Dionne added, noting she had discussed the Star’s questions about deaths on Wettlaufer’s shifts with Det. Insp. Rob Hagerman, who headed the investigation. “No bodies were exhumed,” she said.

In the end, no charges were laid beyond the ones Wettlaufer confessed to. That doesn’t surprise Jim Van Allen, former manager of the OPP Criminal Profile Unit.

“The deaths are historical so you’re relying on paper records. And who completed those? She did,” Van Allen said, referring to Wettlaufer.

He said the OPP likely held round-table discussions with medical experts to assess possible foul play for residents who died during her shifts. Exhuming bodies, he adds, would not have been seen as useful, given the emotional strain on relatives and the difficulty in detecting insulin overdoses in autopsies, particularly in bodies buried for years.

Van Allen believes a cold but reasonable calculation eventually set in with police and Crown prosecutors: “You got her in custody for eight (murder) counts. She’ll never get out, she’ll probably die in prison, so why would you try and convict her for 10?” Wettlaufer pleaded guilty to all charges in 2017.

Deaths under Wettlaufer’s watch were noted in a colour-coded calendar produced by ministry inspectors with the heading, “Caressant Care — Woodstock — EW shift schedule & resident deaths.”

The calendar contains the dates of Wettlaufer’s more than 1,000 shifts at Caressant, which has 163 beds at its Woodstock home. Yellow is used to highlight dates for “deaths occurring during EW shift”; the colour grey notes “deaths occurring within 24 hours of shift”; and red indicates the murders Wettlaufer confessed to. It does the same for her 39 shifts at Meadow Park.

The death/shift calendar’s existence was mentioned in only a sentence or two during commission testimony by ministry inspector Rhonda Kukoly, who did not reveal its contents. A ministry document attached to her affidavit notes the 14 people who died during Wettlaufer’s Meadow Park shifts. But for Caressant it states that “data collection (is) in progress.”

The calendar is not a stand-alone exhibit among the 171 on the commission’s website. Asked if the commission had it, Mark Zigler, its co-lead counsel, directed the Star to a 21-gigabyte sharefile, a subfolder with a volume number, a second subfolder called “images,” and finally a 13-digit exhibit number to find the calendar. He warned it would take “a long time” to download the sharefile. The fastest the Star was able to download it was 30 minutes.

In an interview, Zigler expressed concern that the death/shift information could be used to speculate on whether Wettlaufer killed more people.

He noted that many deaths at nursing homes are expected. He pointed to the new health ministry study, unveiled at the commission, indicating that about 20 per cent of nursing home residents die each year in Ontario: 21,074 during 12 months in 2015-2016.

Police found no evidence that Wettlaufer killed more people; it couldn’t even find forensic evidence linking her to the killings she confessed to, Zigler noted, given the difficulty of detecting insulin overdoses when death can occur days after the lethal injection. Only two of the victims Wettlaufer confessed to killing were exhumed, and cause of death could not be determined.

“This is all 20/20 hindsight, years later,” Zigler said.

“The last thing we would want to do is engage in that wild goose chase because there are families involved. So is it possible that she killed other people? It may be, but one couldn’t definitively prove it based on what we have.”

He also pointed to testimony from Kukoly, the ministry inspector, who described the death/shift calendar as useless, because some residents Wettlaufer killed died on the shift of other nurses.

“If the experts in the field are saying that’s not a particularly relevant statistic,” Zigler said, “then you have to be careful with what you do with it, because you can scare a lot of people about something that may not be factually relevant.”

Zigler said it’s likely the commission’s report — expected in July 2019 — will note the need for more nursing home data. But the specific data points that would be most useful for raising red flags will be left to experts, he added.

Yorker, who was specifically invited by the commission as an expert witness, told the inquiry in September that “serial murder in health-care settings is a phenomenon.”

Citing her landmark studies, she said there have been 131 prosecutions of health-care professionals for serial murders worldwide since 1970. Ninety have been convicted for killing a few hundred people, although the suspected number they killed is at least 2,600.

“These numbers are disturbing and demand that systemic changes in tracking adverse patient incidents associated with (the) presence of a specific healthcare provider be implemented,” Yorker, a professor at California State University, Los Angeles, noted in one of her studies, which formed part of her affidavit to the commission.

The tracking done by the Star — indicating 37 out of 50 Caressant deaths occurred during Wettlaufer’s shifts in a 12-month period — came from two commission documents: Wettlaufer’s death/shift calendar, and the ministry’s new, province-wide analysis of nursing home deaths, which noted the total number at Caressant for three 12-month periods, from 2011 to 2014.

The documents did not estimate the number of deaths that can be expected on a nurse’s evening or night shifts. During the years Wettlaufer was at Caressant, the home had three registered nurses working the day shifts, two working the evening shifts and one the night shift, according to testimony from its former director of nursing, Helen Crombez.

“The fact that Elizabeth Wettlaufer worked on the evening/night shift is consistent with many of the known circumstances in other health-care serial killers,” Yorker stated in her affidavit. “There are fewer health-care providers present, fewer family members are present, and a single RN is typically in charge on the evening and night shifts.

“We probably are not detecting the full extent of health-care killings in both long-term care and in home care,” she told the inquiry, noting the difficulty in gathering hard evidence.

The case of former nurse Niels Hogel illustrates the shocking point. In late October, he admitted to killing 100 patients under his care in two hospitals in northern Germany. He had been in prison since 2015, convicted of killing two patients and trying to kill two others. The 41-year-old admitted to the extra murders in a new trial.

Asked by a commission lawyer what she would consider suspicious signs, Yorker included health-care professionals who fabricate their credentials, who purposely make patients in their care sicker — a syndrome known as Munchausen by proxy — and “being associated with your presence to a discernible increase in adverse patient incidents including cardiac arrest and death. Those are red flags to me.”

Asked what information nursing homes should be sharing with the ministry and other employers, Yorker replied: “Are adverse patient outcomes associated with the presence of a particular health-care provider or particular nurse?”

She added that judges have rejected statistical correlations in caregiver presence during deaths, no matter how strong, as evidence of murder. But the tracking can flag health-care providers who are outliers and trigger quality-of-care scrutiny of their work. Wettlaufer struggled with drug and alcohol abuse, was suspended from her nursing jobs multiple times and was fired twice.

Yorker noted that internal investigations by health-care facilities are often triggered by spikes in deaths, or complaints about a nurse from co-workers, or patients and their familes.

The Star outlined Wettlaufer’s death/shift evidence in a detailed email to Yorker. She said she would prefer to comment after the commission’s final report.

The commission spent little time exploring the need for nursing homes to track the presence of specific nurses when patients have adverse outcomes. It focused instead on the new ministry study analyzing deaths in Ontario’s 633 nursing homes, from 2010 to 2017.

The ministry applied a statistical method called the Standardized Mortality Ratio, used for the past decade in Canadian hospitals. It adjusts for factors such as age and level of sickness, making possible comparisons among hospitals or nursing homes.

The model’s algorithm allowed Health Ministry analysts to calculate the number of residents expected to die at each nursing home in certain years, and compare that to the number who actually died.

Any death above the expected number could be related to factors, such as quality of care, specific to the nursing home, according to testimony by Dr. Michael Hillmer, director of research in the ministry’s analysis and evaluation branch. An especially high number of unexpected deaths could be seen by ministry inspectors as a red flag for closer scrutiny.

In the three 12-month periods analyzed, when Caressant had about 230 “active clients” in each period, it ranked 64th, 61st and 64th among Ontario’s 633 nursing homes. That means it was in the top 10 per cent of nursing homes with more actual deaths than expected ones, Hillmer said.

During those periods, it respectively had 55, 50 and 53 deaths. “Expected deaths” were 39, 36 and 39.

During testimony, Hillmer agreed with Caressant Care’s lawyer, David Golden, that his analysis can’t be used to draw conclusions about the quality of care in nursing homes.

Golden listed several factors that the Hillmer study did not account for when it calculated expected deaths, including the ethnic makeup of residents, differences between urban and rural homes, or whether a resident has a personal physician that provides care. Golden also noted that by including residents who died 30 days after they left the nursing home, the study did not account for events that might have caused death after the resident entered hospital.

The public report does not name other nursing homes. Hillmer noted, however, that 300 of them had deaths higher than the number his model predicted. He testified in September that the ministry hasn’t decided whether this death-ratio analysis will become a yearly, provincewide exercise.

He said his type of analysis would not have flagged Wettlaufer’s murders because the one, two or three per year she confessed to at Caressant were not enough to noticeably spike its ratio of unexpected deaths.

He was asked by commission counsel Rebecca Jones if he had tracked caregivers present during time of death, or whether deaths occurred on the night or day shifts.

“We don’t have that data, and I think even if we did it would produce an enormously complex model,” Hillmer replied, adding that he “would be skeptical that that much extra data would be a useful exercise.”

He was not asked if it would be useful for nursing homes themselves to collect the more specific death/shift data, or to comment on Wettlaufer’s death/shift record.

Yorker, the California professor, agreed that red flags would not have been raised solely based on the deaths that Wettlaufer ultimately confessed to.

“One of the things that makes her different from the ones that we studied,” she said, “is that there were no clusters or significant increases in patient deaths, and that begs the question: had she not turned herself in, could this have continued?”

Sandro Contenta is a reporter and feature writer based in Toronto. Follow him on Twitter: @scontenta