In a hastily organized media conference on Saturday, October 15, 2011, the City of Ottawa’s chief medical officer of health, Dr. Isra Levy, announced that a local, privately owned “non-hospital” medical clinic failed to follow proper infection control measures, resulting in the potential exposure of 6,800 patients to Hepatitis and HIV.
According to Dr. Levy, there was no evidence that a single patient had been infected as a result of treatment, and following consultation with infectious disease specialists he confirmed that the estimated rate of possible infection was “very low”:
- 1 in 1 million for Hepatitis B
- 1 in 50 million for Hepatitis C
- 1 in 3 billion for HIV
On his Twitter feed Dan Gardner, author of the critically acclaimed book, Risk: The Science and Politics of Fear, described the risks cited in this case as “indescribably tiny…dwarfed by the risk of driving to the corner store.”
Despite the exceedingly low possibility of infection, the announcement by Ottawa’s health authority predictably generated outrage and intense public and media scrutiny.
Ottawa Public Health (OPH) first became aware of this clinic’s problems in July 2011, when the Ontario Ministry of Health and Long Term Care advised that an inspection by the College of Physicians and Surgeons of Ontario discovered infection prevention and sanitation protocols had not always been followed. It was then that OPH began its own investigation to assess the risk to public health and identify all patients who might be affected.
This involved a lengthy process of tracing several thousand patient records over a 10-year period. This volume of patient records, combined with restrictions on patient confidentiality set by Ontario privacy laws, made the task of informing those affected extremely difficult.
The final list of patients who may have been exposed to infection was not confirmed until Thursday, October 13th. On Friday, October 14th, OPH put its risk communication plan into effect. The first step involved finalizing the preparation of registered letters that would be immediately sent to all 6,800 patients. This included coordinating with the physician at the centre of the health scare, a professional obligation involving medical errors. Second, it involved notifying local physicians to ensure they would be able to address public demand for information and requests for blood testing. And it involved training as many as 50 public health nurses who would be redeployed from other units (e.g., sex education, home visits with new parents, etc.) to staff a call response hotline.
This plan was developed over the course of the health department’s three-month investigation. Given the possibility of an information leak, only a select number of key individuals were involved in the investigation and planning process.
A threatened media leak
Ottawa Public Health originally intended to hold its media conference on Tuesday, October 18th, at which time all information about the findings would have been disclosed. By this point, all affected patients would have been informed directly about what had occurred, physicians would have been prepared to respond to demands for information and testing, and the call response unit would have been up and running.
On the morning of Saturday, October 15th, Dr. Levy’s office was informed that a national news organization had become aware of the investigation and was preparing to break the story on the basis of inaccurate information.
This placed the public health authority in a difficult situation: the risk that a news report containing misinformation was real—certainly not unprecedented—and had the potential of creating vastly more harm than good.
OPH was faced with three options:
1. Do nothing and respond to the report and the fallout that would ensue after the fact.
2. Provide full disclosure of the situation, including identifying the name and address of the clinic and physician and the types of procedures which had placed patients at risk.
3. Provide partial disclosure that would strike a balance between patient needs, the public interest and the capacity of the system to absorb increased demand for information, testing or treatment.
The risks that kill people and the risks that upset people are completely different.—Sandman, 2007
The health department scrambled to organize a media conference for later that afternoon. At this time, Dr. Levy announced what had occurred, confirmed that there were no known cases of anyone becoming ill and reported the very low numerical probability of infection. He acknowledged that some people might feel anxious or nervous about the announcement, and offered an explanation about what actions his office had put into place and would be following in the coming days, including a promise for new information early in the week.
To this extent, he acted in a manner consistent with the basic tenets of risk communication. He did not over-reassure, acknowledged that people would feel anxious about the announcement and described the discovery and response processes.
However, when pressed by journalists for a fuller disclosure of information, Dr. Levy refused to identify the name or location of the medical facility, the physician who operated it, or details about the patient population affected (i.e., children, adults, seniors, etc.).
This was a risky move for two major reasons.
First, it guaranteed that the health department would clash with the media over competing values: whereas the health department values only pertinent information in the interest of protecting public health, journalists value full disclosure, immediacy and thrive on controversy and outrage. Second, the decision to provide only very general information risked intensifying ambiguity and uncertainty, where the objective of risk communication is to lessen it. People aspire for control over their lives, even if they cannot change what might happen.
Ottawa Public Health called a second media conference on Monday, October 17th, where Dr. Levy disclosed all of the known information about:
- where the breach had occurred (a private health clinic operated by Dr. Christiane Farazli on Carling Avenue in the city’s west end)
- what caused the lapse in infection control (improper sanitation of equipment associated with the performance of endoscopies)
- what patients should do next (contact their physician or the public health department’s call response centre to discuss whether they should be tested)
The news media’s framing of risk has more to do with its reproduction of moral outrage than with “scientific” notions of calculable risk. —Brown, Chapman & Lupton, 1996
Ottawa Public Health and Dr. Levy in particular, came under fire for the decision to provide only partial disclosure in its first media conference.
In a post to his Greater Ottawa blog on October 17th, Ottawa Citizen reporter David Reevely initially described Dr. Levy’s shift from partial to full disclosure as a “volte-face” move, a “classic emergency communications error,” and mused about whether the public health unit might be “sitting on something more shocking.” (He later revised his position, explaining the full context of Dr. Levy’s shift in tactics, characterizing it as a “judgment call…that makes a whole lot of sense when viewed from inside.”)
In a story published on October 18th, the Ottawa Sun did not report the low levels of infection risk but did note the “potentially fatal” nature of Hepatitis and HIV and cited demands from evidently uninformed patients for full disclosure: “You can’t keep the public in the dark…We have the right to know— it’s not fair…. Especially HIV, when there’s no treatment.”
On CTV National News, public relations consultant Barry McLoughlin characterized Dr. Levy’s decision to not release all of the information at once as “a mistake” that intensified public anxiety.
And in an October 18th editorial, the Ottawa Citizen blamed Dr. Levy for causing “undue public concern by mismanaging the release of the information.”
Risk communication: normative and situational perspectives
These criticisms and the demand they represent regarding full disclosure are consistent with normative recommendations for risk communication.
The World Health Organization defines risk communication as “an interactive process of exchange of information and opinion” among authorities, citizens, news media and other stakeholders.
In the past authorities typically acted on the basis of what they believed was the best course of action. Oftentimes this meant shielding the organization itself from blame. Risk communication hinges on therecognition that citizens deserve to be treated honestly, respectfully and with a view to enhancing their autonomy. The objective is to reduce uncertainty so that people will be capable of making informed decisions that affect their lives. Organizations achieve this objective, in part, by communicating as openly as possible.
Notwithstanding the normative appeal of full disclosure, the ability to report all information needs to be considered against a variety of situational factors, including the seriousness of the threat (i.e., the scientifically measured level of hazard or harm), the organizational resources required to manage the response that full disclosure will produce, and the conflict between patient rights to privacy and the public and media’s right to know.
The focus on whether the release of partial information was sufficient needs to be determined in light of the probability of harm and in relation to the ability of the health system to absorb the effects of full disclosure.Given the low hazard for harm and the state of system readiness, and the fact that this event was not caused by the public health department itself, it’s not unreasonable that OPH proceeded cautiously in its first communication with the media and public.
The problem, however, is that this limited the flow of information to journalists, whose occupational values—more information is always better—and “nose for outrage” positions them in opposition.
According to the U.S. Centers for Disease Control and Prevention, “scientists want data to be released when it’s ‘seasoned’—the media want fresh data now.” Consistent with previous cases of low hazard/high outrage events, the Ottawa health department and media differed not only in their treatment of information, but also their definitions of how to define what’s in the public interest. The health department’s partial disclosure not only strained its relationship with the media; it also kept the wider public under-informed and in a state of uncertainty.
Risk communication conclusion
The question of when to release risk information is a serious one, not to be taken lightly. It is vitally important to communicate openly and to communicate early. As the CDC advises, public health authorities need to “be first, be right, be credible.” And according to the World Health Organization, “the benefits of early warning outweigh the risks,” even when faced with uncertainty and the possibility of error.
Although prescriptive recommendations such as these are important in guiding decision-making about disclosure, such decisions cannot be made by virtue of normative standards alone.
Rather, as argued here, they must be made in relation to situational factors. They need to be made in a context that acknowledges:
- It guaranteed that the health department would clash with the media over competing values.
- The resources that will be required to manage the system impacts such announcements tend to produce.
- The legislative environment that balances patient privacy rights against the rights of the public to know
In this case, it’s possible that a full disclosure of all available information in its first media conference would have created undue pressure on local physicians, public health clinics and hospital emergency rooms. Keeping in mind that risk is about both uncertainty and possibility, the scenario of an overwhelmed healthcare system surely played out in the health department’s decision making.
It’s important to note that this risk event was not caused by the health department itself, but by a private clinic regulated by the province of Ontario. Ottawa Public Health responded to an investigation by the College of Physicians and Surgeons of Ontario, and to an alert by the Ontario government. It proceeded with its own investigation and a strategy of public disclosure only when it became evident that the other agencies involved would not do so. The decision to provide only partial disclosure was made on the basis of the health department’s interpretation of the scientific evidence relating to infection risk. That this decision wasallegedly forced by a news organization threatening to break the story with erroneous information, is significant in terms of assessing the response.
Ottawa Public Health acted appropriately in balancing the needs of patients in relation with system capacity, but only to the extent that this event involved infinitesimally low levels of health risk. Had the probability of infection been higher, or had there been evidence of patients who had actually been infected, its response (and this assessment) would likely have been different.
Response problematic in one area: social media
The OPH response is problematic in one other way.
In the most recent edition of his book Ongoing Crisis Communication, W. Timothy Coombs describes the “increasingly important” role of social media for issues management and as a channel for responding to public questions and sharing information. It’s unclear to what extent social media sites are used by Ottawa Public Health to scan or monitor media and public discourse; but for the dissemination of public information sites have been used only sparingly.
For example, (at the time of writing this post) the health department’s under-used Tumblr account does not contain a single update about the infection scare, although it’s been used for other health information purposes during this time. And while its Facebook page and Twitter account have posted synced updates to a low number of fans (363) and followers (5,000+), the fact that both were dormant in the 36-48 hours following the initial media conference suggests social media outreach represents a low priority within the health department’s communication plan.
Given that the period immediately following a public announcement is a critical time when reporters and members of the public are discussing an event and forming their initial impressions, social media platforms present an important space not only for assessing the tone of the public conversation, but for also correcting misinformation if and when it occurs.
Risk events such as the Ottawa endoscopies infection scare can be disorienting because of the intense feelings of uncertainty, anxiety and fear they produce. But to the extent that these situations are potentially destabilizing, they also afford unique opportunities to think critically about how we discuss and practice risk communication.
** This post originally appeared as a guest column on PR Conversations. I thank Judy Gombita and Heather Yaxley for the invitation and their community of readers for the excellent comments and feedback.