Reporting, horizon scanning and risk assessment systems

DILYS MORGAN: Reporting, horizon scanning, and risk assessment are important activities to be able to detect outbreaks early and to inform an appropriate response to these.

Horizon scanning is a component of what we call epidemic intelligence activities.

This is composed of three parts: Scanning, which is the identification and verification of incidents or outbreaks with a potential threat to public health via the systematic review of formal and informal reports; Assessment, which is determining the potential public health threat; and thirdly, Communication, notifications to those who need to know, and the speed of notification here is obviously how urgent is that report you’re receiving. Before starting these activities, it’s very important to define what threats are you looking for? Are these all threats, all infections, or a subset of infections? For example, you may just be interested in vector-borne diseases.

The risks to whom: this is your at-risk population, and it very much depends whether you work at a local, national, or international level. And where in the world are you looking? Many news feed may not be relevant for your at-risk population. Horizon scanning is used to gather information about current outbreaks, incidents, especially those of public health significance. This is achieved by reviewing different sources such as surveillance sites and news reports; aggregate surveillance websites such as ProMED, GPHIN, EOS, FluTrackers; or from reviewing the scientific literature. You will also see links to these sources at the “Also See” section at the end of this step.

Scanning is not just occasionally looking at ProMed, but involves a systematic review of set sources looked at at a pre-defined time, logging the significant incidents, and following them up.

There are key criteria to assess the public health significance of these reports.

These include: events with serious health impact; unexpectedly high rates of illness; potential for spread beyond national borders; potential for interference with international trade or travel; strength of national capacity to be able to contain the outbreak; or a suspected deliberate release. Or more often, it’s a gut feeling in the person doing the scanning. For example, during the Ebola outbreak in West Africa which we were monitoring very closely, once we knew there were cases in Sierra Leone this was obviously of significance, knowing the country, knowing the infrastructure, then we thought this was vitally important to start informing people about the severity and potential impact of this outbreak. Once you’ve got a report, you have to assess the credibility of that report.

There is lots of fake news out there. So if you do receive a report from a source you don’t feel is particularly credible, then we have to follow it up by looking at the better website like the government ministry website of the countries those reports are coming from. Since most emerging infections are zoonosis, that is, they originally come from an animal source, it’s very important to monitor outbreaks of animal diseases or diseases happening at the animal-human interface. Once these are detected, a “One Health” approach is best because you have to involve the veterinarians, the human health practitioners, and the environmental specialists to be able to properly assess the impact of such events.

Everyone has access to a wide range of news nowadays, and international events are perceived as a risk to all, irrespective of whether they have been exposed to that threat or not.

Once an event has been verified of public health significance, a Rapid Risk Assessment is undertaken, this is usually within 24 to 48 hours, to assess that threat to public health.

A Rapid Risk Assessment is necessary to: inform those who need to know, for preparedness and forward planning, to provide appropriate guidance and advice often to a wide range of stakeholders, to inform media, and to advise policy and policy makers. Initial assessment of emerging infectious disease events can be challenging. They must be undertaken in a very short period of time, often when information isn’t available and the events are evolving rapidly. If your Minister of Health asks you for an assessment of Zika in the next country, he doesn’t want it in two weeks time when you’ve done your modelling and the vector transmission studies are available. He will probably want it the next day.

However, despite these limitations, it’s important that you compile all available evidence from all sources before you start the Risk Assessment.

A number of tools and approaches are available for a Rapid Risk Assessment. And these are shown in this slide. They use a variety of approaches, whether they’re a narrative risk assessment, or they use decision trees or algorithms. However, if you are undertaking this work on a regular basis, it’s important that you develop your own system that you are competent in using, so you can produce risk assessments in a robust and systematic way.

Risk is the product of “probability x impact”, and although risk managers and communication teams like you to communicate things in terms of “risk”, it is in fact much more informative to report them separately. This is because a “low-probability, high-impact” event such as a case of rabies or variant CJD, is very different from a “high-probability, low-impact event” such as an outbreak of the common cold. The risk to a population from a communicable disease is dependent on the likelihood of transmission in the specified population, that’s the probability, and the impact or scale of harm should the infection occur. But communication teams, risk managers, and policy makers would prefer you to use an overall term for “risk”.

This may be issued in the format of a risk statement, or after reviewing the evidence, it may be decided that you move to a formal risk assessment process. You may need to review your Risk Assessment frequently as more information on the outbreak becomes available.

A Rapid Risk Assessment will determine: the degree of escalation of communications; whether a response is indicated; the urgency and magnitude of the response; the design and selection of critical control measures; and will inform the wider implications of further management of the incident. Serious attempts should be made to estimate how good the quality is of the evidence you’ve put in underpinning your Risk Assessment. The quality of the evidence underpinning the risk assessment should be expressed as the confidence in the output from that Risk Assessment. This is equivalent to the news reader who tells you we’ve got a 95% chance of rain tomorrow, and in fact, there is only a 5% chance he is right.

Once the Risk Assessment is completed in whatever format, it is important to be able to communicate the output to those

who need to know: risk managers; policymakers; and others.

It is essential to: use simple clear language; use consistent risk descriptors; don’t assume others understand terms or words you assume are obvious; clearly state any uncertainties or gaps in knowledge; don’t make the documents too long or people will never read them; and you may need to justify why you’re not following the guidance of others; and be prepared for the wider repercussions from the media. For example, the following slide demonstrates what happened in a popular newspaper in the UK after we published our evidence-based, robust guidance on sexual transmission of Zika.

However, good Risk Assessments do not always lead to appropriate interventions, good advice, or appropriate newspaper headlines, and often, much of what we do in the Risk Assessment is managing expectations of others.