I. Introduction
This paper investigates some of the social forces that shape death certification. It presents an ethnographic case study of the Office of the Chief Coroner (OCC) in Ontario, Canada, describing the role medical professional norms and values can play in the official identification and reporting of deaths that take place under medical care. Taking the experience of Ontario's physician coroners as a touchstone, it contributes to broader debates regarding the accuracy and reliability of death certificates as sources of epidemiological data and public safety reform. From an account of Ontario's medical coronial system – in which licensed physicians are charged with investigating death on the province's behalf – I argue that medical professional norms and values make it difficult for coroners to see, and easy for them to dismiss, lethal practice as delivered by the physician colleagues they investigate. From this argument I trace out some implications for death investigation and reporting systems throughout the world that rely on physicians to complete their forms and issue death certificates.
Unlike deductive scientific processes which revolve around the proof, or disproof, of an a priori hypothesis, ethnography takes an inductive approach, allowing a story to emerge from the data rather than using the data to test a particular story. Ethnographers acting as interrogators of documents and artefacts; interviewers; and observers of social activity draw on sociological theorisation to guide their analyses. They seek, through sustained, systematised observations of people enacting their culture (i.e. their moral values and social norms) to offer descriptions that illuminate both the specifics of a particular case-study, and broader issues in the way people interact. It is from these efforts to tell not just a local story, but to derive broader lessons that ethnography seeks to test and adjust theories of social action.
In this paper, the local story is one of in-care death investigations with the analysis of documentary, interview and observational data guided by sociological theories of professional culture. I tell the specific story of the investigative advantages and disadvantages that OCC coroners receive from their membership in and adherence to the moral values and social norms of the medical profession. The introductory section that follows draws out existing theoretical and empirical work focused on professional cultures broadly, and medical collegiality specifically, as a means to set up the argument that the values and norms of collegiality shape the content of death certificates in Ontario. At the local level this is a story of physician coroners adapting their official reports to suit the demands of collegiality. At the broader theoretical level the paper suggests modifications to existing theory to accommodate their use in the study, description and reform of death investigation.
While death certificates have long been seen as a source of public safety and public health data (Malthus, Reference Malthus1978 [1798]; Bayatrizi, Reference Bayatrizi2008; Chen, Reference Chen1996), significant concerns have been expressed regarding the accuracy of these key documents. When surveyed, nearly half the resident physicians routinely charged with certifying death in New York City hospitals acknowledged they had knowingly reported an inaccurate cause of death (Wexelman, Eden and Rose, Reference Wexelman, Eden and Rose2013). Epidemiologists have suggested that the inconsistent legibility and quality of death certificates necessitates caution in their use as measures of infection, cardiovascular disease, drug use, mental disability and maternal mortality, among other factors (Dash, Behera and Patro, Reference Dash, Behera and Patro2014; Eriksson et al., Reference Eriksson, Stenlund, Ahlm, Boman, Bygren, Johansson, Olofsson, Wall and Weinehall2013; Govindan, Shapiro, Langa and Iwashyna, Reference Govindan, Shapiro, Langa and Iwashyna2014; Mertz, Janssen and Williams, Reference Mertz, Janssen and Williams2014; Landes and Peek, Reference Landes and Peek2013; Saucedo, Bouvier-Colle, Chantry, Lamarche-Vadel, Rey and Deneux-Tharaux, Reference Saucedo, Bouvier-Colle, Chantry, Lamarche-Vadel, Rey and Deneux-Tharaux2014). Efforts to educate certifiers and standardise the way death certificates are completed have been seen as central to remediating these accuracy issues (Hanzlick and Goodin, Reference Hanzlick and Goodin1997; Goodin and Hanzlick, Reference Goodin and Hanzlick1997; Cambridge and Cina, Reference Cambridge and Cina2010). This approach has emphasised changes to individual characteristics – such as level of education, awareness or interest in the emerging standards for reporting death and its causes (Johnson, Hahn, Fink and German, Reference Johnson, Hahn, Fink and German2012) – but, as elsewhere, has not paid attention to the broader professional culture in which the certificates are being completed (Watling, Driessen, van der Vleuten, Vanstone and Lingard, Reference Watling, Driessen, Van Der Vleuten, Vanstone and Lingard2013).
Previous sociological work in the area of death reporting has highlighted the micro-politics of certification, showing how moral judgments are encoded in certifier and transcriptionist decisions (Prior, Reference Prior1985a, Reference Prior1985b, Reference Prior1989). Similarly, professional norms governing conversation have been identified as modulating certifiers’ capacities to identify discordant interpretations of death and its causes (Bloor, Reference Bloor1991). In his ethnography of forensic pathologists, Timmermans (Reference Timmermans2005) demonstrates how professional self-preservation plays a role in the way official death investigations are conducted. His analysis illustrates how existential professional concerns – concerns about the ability to maintain autonomous control over an area of expertise – shape the way forensic pathologists interpret and certify death. The present paper begins from these observations that the values and social norms of medical professionals are implicated in death determinations, and so the causes and factors that appear on death certificates. The focus of the analysis that follows is on the influence of a central medical professional value: collegiality.
Freidson's (Reference Freidson1988 [1970]) classic study illustrates how doctors have claimed, and maintained, a monopoly over the search for and treatment of disease. Medicine, as a profession, not only sets out and controls the terms and methods used in its area of expertise, but also enjoys autonomy, governing itself through collegial mechanisms. Doctors carry out their work – dealing with patients, educating, evaluating and disciplining one another – in a collegial environment free from outsider, non-expert intervention (Abbott, Reference Abbott1988; Evans, Cardiff and Sheps, Reference Evans, Cardiff and Sheps2006). It is, as Freidson and Rhea (Reference Freidson and Rhea1963, p. 123) note, the collegium
‘… which must provide information on actual performance before action may be taken. Indeed, consonant with the idea of a company of equals, it is assumed in the clinic that it is the collegium which performs everyday supervision.’
Bosk's (Reference Bosk2003 [1979]) ethnography of an elite teaching hospital's surgery department shows this collegiality in action; trainee physicians are indoctrinated into their profession's technical expertise as well as the self-regulating norms of its peer respect-and-review system. Bosk follows senior physicians as they assess and assign value to the deviance of their trainees; either writing mistakes off as acceptable technical error, or writing off the trainee as a bad doctor. Collegiality, as a core value of doctors, embeds commitments to autonomy, self-governance, peer-respect and peer-review. The present paper flows from asking the question: How might these facets of a physician-death investigator's professional values shape the decisions that are made and the certificates that are issued?
Where Innes (Reference Innes2003) describes how homicide detectives go about ‘making crime’ (Ericson, Reference Ericson1993) in the course of investigating murder, the emphasis in the analysis that follows is not on these relatively rare, criminally suspicious cases. Rather, I am interested in death investigations where criminal intent is less an issue than the professional competence and practice of those who have provided medical care. These in-care death investigations are both routine and fraught for the coroners at the heart of my case-study: routine, in that most Ontarians, like others in the developed world, will die while under some form of medical care; and fraught, in that they place physician coroners in the position of overseeing the practice of professional (and often personal) colleagues. I show that, in these circumstances, coroners may have expert intuitions about the (in)competence of a professional colleague, but that these intuitions very rarely come to the attention of either the public, or the professional licensing and governing body for the province's doctors, the Ontario College of Physicians and Surgeons (OCPS). Rather, coroners, enacting a core value of the medical profession, defer to colleagues and collegia, keeping their investigations as local as possible.
My analysis begins with an overview of the social and technical competencies that physician coroners bring to in-care death investigations. It describes how membership in the medical professional fraternity provides a basis of trust for interacting with health-care staff involved in a death, and also how expertise in deciphering the handwriting and technical meaning encoded in medical notes extends to a capacity to read incompetence and poor care between the lines of a chart. The analysis then moves on to describe how, once a coroner has identified any competence issues, a public report in the form of a death certificate or investigative summary is produced, but expurgated. The burden of my argument through this section of the paper is that coroners consistently follow the norms of collegiality, which has the effect of making (in)competence a private rather than public safety issue.
II. Case-study
The OCC is the public death investigation agency for the province of Ontario, and manages between 320 and 350 commissioned coroners at any given time. The office is charged, by law, with investigating suspicious deaths across a one million square kilometre jurisdiction with thirteen million residents.Footnote 2 Its motto – ‘We speak for the dead to protect the living’ – summarises a mission to ‘collect and analyze information about [a] death to prevent further deaths in similar circumstances’ (OCC, 2007). While the OCC's governing legislation does not specifically mention this preventative mission, Section 20(c) of the Coroner's Act (hereafter the Act) instructs coroners to consider
‘… the likelihood that the jury on an inquest might make useful recommendations directed to the avoidance of death in similar circumstances’
as they make their decision to convene, or refrain from convening, a public inquest. The office has built its mission statement, and defined itself as a public safety specialist around this instruction.
In an average year, 80,000 Ontarians die, and 18,000 of those deaths will meet the OCC's definitions of ‘suspicious’ and so be subject to a coronial investigation (OCC, 2007).Footnote 3 Of those 18,000, approximately 7,000 bodies are sent under a coroner's warrant to a forensic pathologist for medico-legal autopsy. This means that autopsies and coronial investigations are separate activities in Ontario, and performed by different personnel. Section 31(1) of the Act requires commissioned coroners to hold a medical licence with commissions issued by the Lieutenant Governor of the province acting through the Chief Coroner and the various Regional Supervising Coroners who recruit and develop physicians for the work.Footnote 4 Under the Ontario system, coroners are always qualified and licensed physicians, but not necessarily pathologists. Indeed, during my time at the office, none of the OCC's Senior Managers, Regional Supervising Coroners or Operational Coroners were either clinical or forensic pathologists. Rather, they were licensed physicians who came from backgrounds in hospital administration, emergency medicine, family practice and a range of other specialties. In everyday operations this means that family or ER doctors sign warrants that empower provincially endorsed forensic pathologists to conduct autopsies on their behalf. OCC managers and front-line staff describe this division of authority and labour between physician coroners and forensic pathologists as composing a ‘medical coroner system’.
A former Chief Coroner of Ontario, testifying at a 2007 public inquiry, described the range of approaches to coronial work that are deployed across common-law jurisdictions as he summed up the ‘medical coroner system’:
‘There are as many coroner systems as there are jurisdictions …
So if you ask somebody from the United States about a coroner system, what they generally are thinking about is a system that is a county-based system where the coroner is an elected official and holds an elected office and then runs the death investigation system by hiring pathologists …
A few states have developed systems where [the coroners] still are elected, but they have to be a physician …
Medical examiner systems [have also developed] in some of the larger [US] cities … as a means of dealing with the large volumes of homicides …Footnote 5
Within Canada, there are no two systems that are identical.Footnote 6 So there are coroner systems that are medical; [there are] coroner systems that aren't [medical]; [and] there are pure Medical Examiner Systems … I really think the Ontario System is really a blend of a bit of both.
[In] the Ontario Coroner's System, all of the investigating coroners and the people who go to the scene are licensed physicians.’ (Goudge, Reference Goudge2007)
In practice, this means that most of the province's coroners are full-time family practitioners who pick up extra ‘piece work’ for the OCC. For a fee, plus expenses, they conduct scene investigations and interviews in collaboration with local law enforcement, producing official determinations of death on behalf of the province, sometimes signing warrants for autopsy and providing context for the forensic pathologist who will conduct that work on their behalf.
III. Methods
My ethnographic study of the OCC began with an analysis of the office's legislative mandate and the limited materials it makes publicly available in print or on the Internet. After gaining approval from the University of Toronto's human subjects ethics review board, and negotiating access with the OCC, I moved from attending public inquests and forensic conferences as an observer to consulting the office's extensive internal operating manual as well as the paper and electronic forms that Ontario's coroners use to record their investigations and determinations. Along with searches of the primary and secondary literature, and the sociological literatures on the professions generally and collegiality particularly, this documentary analysis informed the interviews and observations at the heart of my ethnography.
Over seventeen months, from May 2007 to September 2008, I performed 104 interviews with all of the OCC's senior managers; Regional Supervising Coroners; forensic pathologists; forensic anthropologists; and a sample of the operational coroners from around the province. In addition, I interviewed homicide detectives and forensic identification officers, children's aid workers, forensic anthropologists and inquest lawyers. The interviews lasted between 90 and 120 minutes, and were taped and transcribed for analysis. My observations (88 hours) were focused on ride-along sessions in which I accompanied and sometimes assisted coroners as they were called to death scenes or sat, bored, in their cars. Those who agreed to these ride-along observations were a subset of those participating in the interviews. The observations were written up in field notebooks, and were also transcribed. Analysis was facilitated by NVIVO qualitative coding software following the constant comparative method (Glaser and Strauss, Reference Glaser and Strauss1968; Boeije, Reference Boeije2002). In this emergent, inductive method, themes are identified, revisited, expanded and collapsed prior to final interpretation. Excerpts from my field notes and interviews are given below to support this final interpretation. Field notes are rendered in italics, and interviews in indented regular type.
As part of anonymising my dataset, the subjects have been assigned random designator numbers ranging from 1 to 103. The study included 103 unique subjects, one of whom was interviewed twice, for a total of 104 interviews. Subjects have also been assigned a gender based on an assumed 50 per cent chance of being female or male. While this probability reflects a rough approximation of the ratio of females to males in the general population, it provides a skewed representation of the Ontario coronership. This is to say, the chance of being assigned a female gender as the data were randomised was much greater than the chance of encountering a female coroner in my sample. I justify this choice on the grounds of the increased anonymity it offers to the 16 per cent of my sample who were, in fact, female. While the data I present give an impression of gender equality, it is in fact the case that death investigation in Ontario is a predominantly male enterprise. Eighty-four per cent of my interviewed subjects were male. I suggest that this imbalance is in some part a reflection of who was willing to talk to a male researcher, and in some part due to the gender distribution across these occupations.
The terms ‘Senior Manager’, and ‘Operational Coroner’ used for attribution purposes in the analysis that follows are descriptive conveniences rather than reflections of actual OCC nomenclature. The senior management of the OCC is composed of the Chief Coroner and two Deputy Chiefs. All three of these positions are described in the text as ‘Senior Managers’. The OCC's front line investigators are described here as ‘Operational Coroners’. Their direct managers, and the middle management layer between the Chief and her deputies and the front lines, are referred to here as ‘Regional Supervising Coroners’.
As with all observational studies, my data come from a specific location at a particular moment in time. My analysis and theoretically informed efforts to transcend these specifics and so draw out generalisable conclusions regarding the interactions of medical professional values and death determination will have been shaped by my biases as a social scientist. Following best practices in ethnographic research, I have sought to dampen the effect of these race, class, gender and education biases with self-reflexive attention, and by seeking my subjects’ input and commentary on my emerging analysis.
IV. Three in-care death investigation competencies
The OCC's coroners arrive at in-care investigations with three related competencies that enable and frame the work: their status as health-care insiders and members of the medical collegium; their expertise in reading the handwriting and technical content of medical charts; and their capacity to read between the lines of medical charts.
4.1 Being collegial
An Operational Coroner described how his status as a physician and reputation as a collegial professional built trust and drew out information as he arrived at the site of an in-care investigation:
‘I find that health care professionals who were involved [in an in-care death] – not always the ones directly involved, but even some of them – do want to talk to you. They want to tell me. You know, the nurse will pull me aside and say, “Look, I have a concern here. This person had been through three walk-in clinics today. And now they're here in the emergency and they're dead. Like, did you know they'd been to three walk-in clinics?”
“No, I didn't. Do you know which ones?”
“I'll try to find out for you.”
I mean, people want to talk to us. Why? It's a good question. Maybe because we're physicians so we're colleagues so they see we're on their side. And you say, “Well, what about the College [i.e. the OCPS]?” No physician sees the College as on their side. Despite what the College might try and tell you.’ (Participant 6: Operational Coroner)
Another Operational Coroner who worked as both a clinician and coroner at a teaching hospital emphasised the importance of being perceived to be part of the local collegium and not part of an outsider organisation like the OCPS, or the OCC.
Q: ‘Are people from this hospital unhappy when you investigate them?’
A: ‘I think they're OK with me because they know me. Because they know I'm ruthlessly fair. I'm honest, but I'm ruthlessly fair. So they're OK with me personally. They don't like [the OCC] generally. But they're OK with me personally. Actually, I probably have a better chance, paradoxically even though I'm [part of the larger hospital structure], of being seen as an impartial agent than if I was from outside. Which is kind of weird.’ (Participant 26: Operational Coroner)
Trust, respect and key investigative information all flow from maintaining medical professional norms. Authority is not derived exclusively from the OCC mandate, or the threat of OCPS reporting, but to a significant extent from a coroner's health-care insider status, and a personal reputation for ruthless fairness and adherence to collegial norms.
4.2 Reading technique
Coroners bring technical expertise in understanding the content of medical notes, and reading the handwriting in which they are rendered. However, an Operational Coroner explained there are often limits to a generalist investigator's capacity to understand the details of a death that has occurred under a specialist's care.
‘It's not a pleasant feeling, but one thing we learn as doctors [is] to recognize our limitations fairly quickly. And we know in our mind when we're getting outside of our area of comfort. The first step is to be honest about it.
And it's not unusual for me [to] get a phone call from, you know, one of the top [specialist] surgeons in Canada, that [has] had a death on the operating table. And they were doing some sort of a procedure, and there was some sort of special instrument attached to here and there, and they were, you know …
[In that situation] the first thing to do is to say, “Listen. You're going have to dumb this down a couple of notches and explain to me what this apparatus is, and what it does. Was there a possibility of a malfunction, etcetera?”
And, you know, nothing puts a doctor more at ease than another doctor saying, “You're talking over my head. You're going to have to explain this in simpler terms.” You know, they don't feel like they have to defend themselves … They don't have to be concerned that they're in the middle of some sort of an investigation shortly after a death has occurred.’ (Participant 85: Operational Coroner)
While physician coroners bring a basic understanding of anatomy, biology and pathology to their in-care investigations, they are often out of their area of expertise when consulting charts or debriefing specialist colleagues. In these cases, as noted above, setting an appropriately collegial tone by removing the sense that a public investigation is underway, is seen as the best way to gather information. At the limits of a coroner's technical understanding, following collegial norms sets the conditions of trust that will allow an in-care investigation to proceed.
At a more mundane level, coroners bring their professional experience and technical expertise to deciphering medical notes. The following passage describes a coroner attempting to read a treatment chart.
‘Thumbing through [the deceased's] treatment chart [participant 98] squints, and then looks more closely at one of the pages. He takes his glasses out of his mouth and puts them on, squinting again.
“This is one of the reasons they get docs to do this,” he says “We can actually read some of this crap!”
Focussing intently and reading slowly he mouths the words written on one page. Looking up, he removes his glasses and uses them to point at the bottom of the page, saying: “Well, the [deceased] had some issues with cholesterol but I can't read that bit at all.”’ (Participant 98: Operational Coroner Ride Along)
At base, then, the OCC's operational coroners possess – more or less – the professional experience and expertise to be collegial in the face of specialist knowledge, and to grasp another physician's meaning by deciphering their handwriting.
4.3 Reading competence
Another coroner described how the ability to decipher handwriting extended into reading between the lines of medical charts.
‘It's kind of funny. In medicine, we have a thing people write down [on a treatment chart]: “Patient alert: NAD”, which means “No Apparent Distress”. The joke in medicine is that that really means “Not Actually Done”. We can [as physicians] all kind of see through that, to a certain extent. We can look at a note and say, “Oh, you can clearly tell from this note that this [patient] was well assessed”; or “You can clearly tell that this [care giver] is just writing a note [in the treatment chart] for the sake of it”.’ (Participant 6: Operational Coroner)
The coroner brings more than just collegiality and technical skills to deciphering poor handwriting, jargon and acronyms. His professional expertise allows him to read the competence of medical colleagues in the treatment notes they complete. Another Operational Coroner described how she used her professional experience as a touchstone when reading through the treatment charts of professional colleagues involved in an in-care death.
‘I just go with the facts. And stack them up against my clinical experience. I mean, if I go in and investigate a case, I have to say: “OK, in my experience, is this a reasonable scenario where somebody just made an error in judgement?” – and we all make errors in judgement – “or is this a case where, you know, the [treating physician] tried to manage this from his bed, rather than [getting] out of bed and [examining] the patient.”
… If I can read through something and find somebody who's actually cared about what was going on, just made an honest, you know, error in judgement, but made an attempt, then I'm OK. Alright?
If I find somebody who's been cavalier, or who has obviously been over their head, and chosen not to recognize it, or just didn't give a rat's ass, then I will pursue that issue. And you know, it may sound high and mighty on my part, but I have a good nose for caring. And I use it. I use the skills I have.’ (Participant 83: Operational Coroner)
When reading medical charts as part of an in-care death investigation, coroners parse not just handwriting and technical content but the moral intent and professional competence of colleagues.
To capitalise on their status as physicians, and so build trust with the subjects of their in-care investigations, coroners must enact the collegial values of the medical profession. These demonstrations of deference and respect for a peer's autonomy and expertise are the actions that validate a coroner's professional status. Without them, the coroner runs the risk of being seen as an interloper, albeit with a medical degree, from an outside agency like the OCPS or OCC. The competence with which they perform this collegial enactment shapes the flow of trust and information, and so the course of the investigation and the documents it produces. Coroners also bring a combination of technical and cultural expertise to their work. Their professional capacity to decipher the handwriting and technical matters contained in a medical chart – which varies according to the legibility of the script and level of specialisation involved in the case – informs the outcome of in-care investigations. In addition to these technical skills, their professional experience allows them to read between the lines of a deceased patient's medical notes. The hard technical and soft cultural competencies of medical professionals are thus central to the way coroners come to know about and certify death. The next section examines what coroners do with the information they gather.
4.4 (Not) reporting incompetence
When they decipher incompetence or poor care between the lines of a chart, coroners’ preferences for collegiality and respect for peer autonomy tend to limit how they act on what they have learned. Their first medical professional instinct is to avoid escalating or formalising competence and care issues.
Q: ‘Can you look at a file and say, “No. That's clinically incompetent.”?’
A: ‘I certainly can. And I do. In my head. But I can't put that on paper. Like [pauses] … I can't. I can say there are questions regarding care, or there are issues, you know, involving this [or that technique], but I'm not there to make any kind of a judgement.’ (Participant 23: Operational Coroner)
The coroner's reading skills provide her with an assessment that she withholds – as part of upholding collegial values – from her official report. Although she is investigating the death on behalf of the province, and arguably is there to make a judgment, she defers public judgment of a professional colleague. Rather than fulfilling a public duty, she fulfils a collegial one by keeping the discussion and determination of error within the bounds of the professional group. But what happens next? My observations suggest that in place of filing a formal publicly accessible report, coroners pursue one of a pair of options for dealing with incompetence or error uncovered during an in-care death investigation. Both of these courses of action are consistent with and framed by collegial values. They are: dealing directly with the local collegium; and reporting the case up the collegial hierarchy.
VI. Dealing directly with the collegium
An extended passage from my field notes illustrates, first, the instinct to avoid formalising professional mistakes in publicly visible reports, and second, the direct approach to dealing with a local collegium. The passage is condensed from several hours of field notes.
‘Investigating the death of a man who died shortly after arriving at hospital complaining of constipation, Subject 44 skims through a thick chart detailing the man's mental illnesses, pausing to look more closely at the treatment notes and test results of the last 24 hours of his life. Having taken out a blank Coroner's Investigation Worksheet (CIW) and filled in “arteriosclerosis” as the primary cause of death, she explains:
44: Hey, he's [over 50 years old], we can assume he's got some of that.
She then adds a secondary cause of “renal failure” and pauses, shaking her head.
44: I'm withholding hypovolemia here
Q: Hypovolemia?
44: Not enough liquids. I'm not going to put it down.
Q: Why? Doesn't it go with the renal failure?
44: You've got to think through who's going to be reading this later … You keep it off the form because it's just going to lead to questions later. Maybe from family, who knows? … Listen, [the emergency department doctors] didn't realize the guy was in renal failure so they didn't treat him for it, right? If I put down “hypovolemia” somebody's going to ask: “So why was he hypovolemic?” Right before they ask “Why didn't they do something about it?” This just keeps things neater. Less questions. Footnote 7.
With the CIW partially filled in, the coroner begins a conversation with an Admissions Nurse, inquiring about her vacation plans and asking for a supply of wrist identification tags. (She later explained that this hospital had the best tags in the region; specifically, the sort of tags that remained attached to a body all the way to the morgue.) Directed towards a cupboard in the copier room by the Admissions Nurse, the coroner accidentally meets the Chief of the Emergency Department while he is using the copier. As he recognizes the coroner the Chief visibly stiffens his posture saying that he knows why the coroner is in his department.
Placing the deceased patient's medical chart on the copier, the coroner opens the file folder to the page where the treatment notes indicate a blood test was ordered. The Chief leans forward and, indicating the signatures below the clinical orders and observations, explains “These are my best guys. My best guys.”
The coroner responds that she feels there is little to worry about from her point of view and the Chief's shoulders visibly drop. His speech relaxes and he explains he will be holding a meeting with all the staff to discuss the case. The coroner nods and the conversation turns to pleasantries about vacation plans and career trajectories. Eventually the Chief leaves the copier area and the coroner resumes her hunt for the cupboard containing the wrist tags. With the tags in hand she returns to complete her CIW and explains.
44: At the end of the day I truly believe in people's right to make a mistake. We've all been there – I did emerg for [more than 10 years] – and it's always easy for the armchair quarterbacks to come in and tell you what you should have done … When you've got someone who's admitting they messed up and is going to do something. Someone who's clearly concerned. We're just going to leave it alone. Not that [the patient] deserves any less investigation, but what more are you going to do? … I mean, hypothetically, at a coroner's inquest, what's the recommendation going to be? “Blood results should be looked at [by the doctors who order them] ?” I mean, “Hello!? What's the point?”
Q: So that's it for this file?
44: I might ask [the department chief] for a copy of the new protocol, [or] whatever comes out of the meeting to see that something's changed, but I'm not all that worried. I mean, did you see the guy? [He] knows this is a fuck up and he doesn't want it happening again.’ (Participant 44: Operational Coroner Ride Along)
The passage shows collegial values shaping the content of a death certificate. With only a superficial investigation of the body, arteriosclerosis and renal failure are entered as factors in the death. The medically induced hypovolemia that contributed to the renal failure is withheld. With the certificate expurgated to limit the public's view of the competence and error issues potentially involved in the death, the coroner has met the primary medical professional goal of keeping the discussion local and amongst qualified experts.
Next, she capitalises on the serendipity of finding herself in the copy room alone with the chief of the emergency department. Talking peer-to-peer, which is to say one experienced emergency doctor to another, the coroner uses read-between-the-lines skills to assess the chief's competence and moral intent in dealing with the error. Finding him to be serious about the incident she defers to his status as first-among-equals within the emergency department. Discipline and formal or informal consequences as well as more future-focused changes to policies and protocols that might prevent similar errors are deferred to the Chief, who has vouched for the two physicians directly involved. Both the content of the death certificate and the discussion surrounding the death, then, are shaped by, and expressions of, collegial values.
Similarly, a coroner described relying on the local collegium to deal with an in-care death in which a surgeon failed to review notes or respond to family concerns regarding a blood-thinning drug.
‘[The patient] was managed peripherally. No one ever revisited the internal medicine doctor's notes to say, “Hold [blood thinners] for a few days and then re-evaluate.” So that never got re-evaluated. [The patient] stayed in the hospital – off his blood thinner all that time – then got transferred to a nursing home for further convalescence, then keeled over because he didn't get put back on his blood thinner. [His] family had been screaming for weeks, “Why isn't he on his [blood thinner]?!” [They left] messages for the [surgeon] and so on, and nothing happened. And this guy then died.
… I ended up having to call and talk to the hospital administrator and say, “We [i.e. The OCC] are investigating this case, you know, you'd best [do an internal review]” So they ended up doing an internal review. And [the administrator] ended up sending a letter to this [surgeon] saying, “Listen, you've screwed up this case.” It was not an uncommon situation. So it got handled that way. But it's not my role to go in there and say, “You bad doctor. You're not allowed to do this”.’ (Participant 20: Operational Coroner)
While not dealing directly with the Chief of Surgery, the coroner avoids entering a public judgment of bad doctoring by working with the local collegium. The not uncommon situation of this surgeon peripherally managing his patients is handled by deferring to the disciplinary capacity of his immediate professional peers. Although she chooses to work at one level of remove, the coroner's follow-up to identifying competence issues and not publicly reporting them deals directly with, and defers to, the local collegium.
VII. Reporting up the collegial hierarchy
A second option for coroners who feel compelled to act on competence and error issues revealed during an in-care investigation is to report the case up the collegial hierarchy. This was the preferred option for those Operational Coroners working in smaller communities where their investigations were likely to include personal rather than purely professional colleagues.
An Operational Coroner in a community of 50,000 people described investigating a series of overdose deaths. Having developed the sense that a local colleague was over-prescribing narcotics, the coroner referred the case up the OCC and collegial hierarchy.
‘And so I called my Regional Supervisor, and I told him, “This is what I have. I don't want to go any further with this.” I said, “I want you to take it over and take it from here ’cause I have to work with these guys [i.e. the other doctors in the area].”
And he said, “Fine”.’ (Participant 99: Operational Coroner)
Similarly, an Operational Coroner described in-care death investigations in her community of 100,000 people.
‘Rather than me trying to, you know, call the doctor up and say, “[a family member] is questioning your care.” I mean, I don't see that as my role. It's also difficult in a small community, you know? These are my colleagues, so you may in your mind think, “Yeah, that could have been done a lot better”, but that can get difficult. And some of them are just judgement calls. Not, you know, gross negligence. I think I would have an obligation if there was obvious medical negligence. That would have to go further, and not just leave that up to the families. But I'd be discussing that with [my Regional Supervisor], and let him deal with the doctors because these are people that I deal with in other ways, right?’ (Participant 82: Operational Coroner)
When the Operational Coroner reads between the lines of a treatment chart that a colleague could have ‘done a lot better’ she finds herself in the difficult position of judging a doctor she deals with in her role as a member of the local medical collegium. These are physicians who she would prefer to engage with on a local, consultative level; and physicians from whom she may receive patients and to whom she may send them. As a result, the tendency is to refer these cases up the OCC chain of command to a medically qualified Regional Supervisor.
VIII. Avoiding the college by being collegial
Although in-care investigations revealing competence and care issues could be seen as opportunities to report to the OCPS, this was almost never the case.
‘I want to point out that I've never referred someone to the College [i.e. OCPS. In fact,] the number of cases that get referred to the College is very low. I mean, I wish I could remember, but I don't think it's even in double digits in a year. So it is very low. So that is not our primary responsibility.’Footnote 8 (Participant 6: Operational Coroner)
Similarly, a Regional Supervisor noted:
‘We don't tend to report clinical judgement issues [to the College] … I guess the phrase we use [is] – “We don't second guess reasonable clinical judgements.” But if people are outside sort of what's reasonable, then we get the [College] involved.’ (Participant 69: Regional Supervising Coroner)
To judge both the reasonableness of a colleague's actions, and the legitimacy of forwarding a case to the OCPS, coroners convene mini in-house collegia.
‘When I have cases where I've had gross incompetence … I contact [my regional supervisor and say], “Look. You know what? I think I see gross negligence here.” My regional coroner would then look at the case. If he agrees – which is fair, it's a second set of eyes to look at it, and a more experienced set of eyes – he'll refer that to the [OCPS], right?’ (Participant 6: Operational Coroner)
It is noteworthy that the coroner shifts formulations as he makes his report to his supervisor. While his between-the-lines reading of the case allows him a clear view of ‘gross incompetence’ in his presentation to his manager he only thinks he sees negligence. So he calls in an extra pair of expert eyes, convening a collegial discussion of what he has read as gross incompetence. These in-house OCC peer-to-peer consultations shape the content of death certificates, and at the same time limit the referral of competence issues to the OCPS. A Regional Supervisor described OCPS referrals:
‘It's not common. It's not common. But I will say, again, if you look at the published research on it – James Reason is the guru of error in large systems – ReasonFootnote 9 has looked at this, and actually 85 per cent of the time, [the error] has nothing to do with competence. It has everything to do with humanity. Well, it's only 15 per cent.
So of the cases we [as OCC managers] see, we should see about, you know, 15 out of 100 that are going to be in [the incompetence] rubric. So that's kind of what my yardstick that I look at to see whether or not I've got it right.’ (Participant 32: Regional Supervising Coroner)
The manager describes how the work of an organisational psychologist is used as a check on the OCC's mini-collegium negotiations of what constitutes reasonable behaviour. A ‘15 per cent’ guide provides a rule of thumb for patrolling the border between reasonable and unreasonable clinical actions. Collegiality, then, does not just see coroners consulting their superiors to confirm they have seen incompetence, but also those managers using managerial theory to justify a low rate of referrals from the OCC to OCPS.
IX. Working with local collegia
Like Operational Coroners, OCC managers prefer to keep things local. A Regional Supervisor described moving in-house OCC discussions about competence and care outwards, towards the organisations and collegia where in-care deaths occurred. Hospitals and care homes, he noted,
‘… are very receptive to us having what I call a “fireside chat” where we sit down behind closed doors. No lawyers. No press. No family. And say, “Here's what we think happened. Are we correct?”’ (Participant 22: Regional Supervising Coroner)
Another Regional Supervisor described dealing with competence and care issues in an emergency department where the hospital had recently been the focus of a public inquest. Although the manager applies more pressure to, and demands more action from, the local collegium than might be expected in a ‘fireside chat’, she too defers to local autonomy in the development, implementation and dissemination of reforms. The emergency department had twice discharged a mentally ill patient who then killed herself.
‘I sat down with the hospital, and I said, “There's a lot of pressure for an inquest.” They did not want another inquest [laughs], and I'm sensitive to that, and I say, “We've gotta make recommendations. We gotta address this concern because, clearly, this [patient] presented twice [at the emergency department]; got booted out twice; and kills herself.”
And we sat for three and a half hours, you know, myself and all the players and the [investigating] coroner and the chief of the emerg. And the chief of emerg. happened to be the doc who kicked her out. So you can imagine that the emotions were a little bit [strained], and there were some neck veins that were sticking up a bit.Footnote 10
After three and a half hours, we had probably eight or nine excellent recommendations. And, in the end, we had three pages of such good recommendations that I was thinking of going to the Chief Coroner and perhaps the Hospital Association, [and telling them], “Distribute these recommendations to all your psychiatric hospitals because they're so solid; they're so good.”
I thought the hospital should actually go public and say, “These are the results of our review.” Because, I mean, they hit a homerun. They made such excellent changes. And they put a ton of work in it. Way better than any [inquest] jury would come up with.’ (Participant 86: Regional Supervising Coroner)
With a public inquest just completed, and demands being made for a new one, the Regional Supervisor is unable to defer investigation, discipline and reform entirely to the doctors involved in the death. However, convening a closed-doors review that includes ‘all of the players’ but not the family of the deceased, she chairs a heated but collegial discussion that results in pages of excellent recommendations about the handling of mentally ill patients in the emergency department.
Although the Regional Supervisor thinks the hospital should go public with the ‘home-run’ document produced out of the review she has instigated and chaired, ultimately she defers to the hospital's desire to keep things local.Footnote 11 As a result, neither the Ontario Hospital Association, nor any individual psychiatric hospitals or emergency departments, are likely to hear what was discussed during this OCC in-care death investigation.
X. Conclusion
Medical professional skills and values make it relatively easy for OCC coroners to identify competence and poor care issues during in-care death investigations. Their hard technical and soft cultural skills allow them to read both the content and between the lines of medical notes. They combine their status as physicians with enactments of collegiality to build trust with the health-care personnel involved in the investigation. These enactments of collegiality come in the form of deference to colleagues’ expertise and clinical opinions, and efforts to ensure local, medical control of decision-making and disciplinary processes. These same enactments of, and commitments to, collegial values make it relatively difficult for OCC coroners to officially document any competence and care issues that they may identify.
The implications of these systematic cultural preferences for death certification are not clear-cut. On the one hand, medically qualified coroners are in an excellent position – technically and socially – to identify and elicit information about in-care deaths. It is difficult to imagine a retired police officer, lawyer, funeral director or other lay coroner with such a fine-tuned capacity to understand the content, handwriting and moral intent of care as it has been delivered. Similarly, lay coroners arrive at in-care deaths carrying the legal authority of their office, but not of the medical professional fraternity. Even if they attempt to enact collegial values by remaining non-confrontational and keeping the discussion local, as mistrusted outsiders their access to information seems likely to suffer. Without this trust and information, the death certificates and investigation summaries they produce will be incomplete.
On the other hand, the official, publicly accessible documents that medical coroners produce are routinely expurgated as part of enacting the collegial values that maintain trust and keep information flowing. This has two main effects: lowering the official count of medical errors and competence issues; and sequestering public safety decision-making into collegial forums. The first of these, in the Ontario case-study, results in low referral rates from the OCC to the OCPS. It also results in death certificates that, as they are shaped by collegial norms, underrepresent the incidence of medical error. This suggests medical error ought to join a growing list of epidemiologically unreliable death factors (Dash, Reference Dash, Behera and Patro2014; Eriksson et al., Reference Eriksson, Stenlund, Ahlm, Boman, Bygren, Johansson, Olofsson, Wall and Weinehall2013; Govindan et al., Reference Govindan, Shapiro, Langa and Iwashyna2014; Mertz et al., Reference Mertz, Janssen and Williams2014; Landes and Peek, Reference Landes and Peek2013; Saucedo et al., Reference Saucedo, Bouvier-Colle, Chantry, Lamarche-Vadel, Rey and Deneux-Tharaux2014). Future research to investigate underreporting will need to determine whether all death investigation systems that ultimately rely on physicians to complete their certificates and inform their conclusions are prone to this unreliability. It may be that medical professional values shape the content of reports and certificates regardless of the credentials and professional values of the authorised investigator (e.g. legal, lay) in a given jurisdiction. In any case, the data presented here suggest the critical importance of medical professionals’ values and norms – of their commitments to collegiality, specifically – in the death determinations they make. Making a place for, and improving understandings of how culture shapes the facts and documents of death, will be key to designing and implementing real reform. The present study deploys sociological theory to provide a first look at how collegiality benefits and detracts from the accuracy of death certification.
Although efforts to educate certifiers and standardise the way death certificates are completed are ongoing (Hanzlick and Goodin, Reference Hanzlick and Goodin1997; Goodin and Hanzlick, Reference Goodin and Hanzlick1997; Cambridge and Cina, Reference Cambridge and Cina2010), these tend not to address the cultural values – such as collegiality – that can drive reporting down, or underground. This suggests an opportunity in the development of medical school curricula and clinical level interventions to shape norms, and introduce boundaries to the application of collegial values to death certificate completion. A key intervention here might focus on altering the perception that keeping things local and professional is the best response when encountering issues of harm and competence. The success of such an intervention would seem to hinge on redefining medical error and competence as a public safety issues, rather than private collegial ones. To accomplish this may require drawing certifying physicians’ attention to the value of diversity in determining solutions to medical errors. For the moment the public – either in its own right or through proxy agencies like the OCC – is at best peripherally involved in determining which practitioners and systems of practice constitute public safety hazards. Limiting participation in the discussion of public safety tends to perpetuate the solutions and blind spots of those constantly involved in the process. Similarly, privatising these discussions to local collegia prevents broader system learning.
While framing in-care death investigations through the values and norms of collegiality successfully excludes much inexpert and inefficient lay participation, it also reduces the range of values and norms that shape how a death is viewed. As an Operational Coroner defers to, and seeks to assist, the department where lethal care has been delivered; or an OCC manager works in a legally protected private session with a hospital, an in-care death that begins as a mistake with no systemic implications remains a mistake with no systemic implications. Public participation in these cases is not so much about expertise as fresh eyes that might generate three pages of ‘home-run’ recommendations on how to handle suicidal patients, or new ways of ensuring emergency doctors read test results.
