305 - Risk

I am quite happy that the phrase perceived risk explains any sense of adventure; think of trips in the hills and of moments where any member of group can decide something is exciting, even as others in the group see no reason for excitement. What any climber reckons as an exciting climb is well past the point where I'd rate the same thing as simply suicidal, but that is more a matter of competence.¹ There are few places on Dartmoor (walking) that count as exciting, but I have been places where, for my companion of the day, the risk rating passed beyond all previous experience and therefore (having survived) was exciting in the extreme. This excitement is one way in which we deal with risk, but we might recognise that these are generally things that we are doing immediately and the (perceived) danger is that we might fail to control our bodies well enough to achieve the objective. Often this is a matter of little more than balance or unfamiliar ground or unusual sights.

There is a different sort of risk to associate with risk for others. Often this might be described as a transfer of responsibility. Suppose you have a garden with a river along one edge, not fenced. How do you feel about there being primary-school age children in the garden? Is what bothers you not that a child might get wet but that you will be blamed for any event that transpires?

What action do you take for any perceived risk of this class, the conceivable event? There are three choices: accept the risk, avoid the risk or reduce the risk.  The two extremes are obvious, the intermediate one less so. Reduction of risk (my dyslexic fingers tried to type re-education) rather implies a reduction to an acceptable level. That in turn suggests that there are quite likely to be a number of actions available, some or all of which must occur before a risk is deemed to have dropped to within acceptable levels. I was always aware that any risk to be considered during activity in the hills was (in my head, almost always) less than that of being on the roads going to and from the hills. But that is mostly because I have no control of the other idiots on the road, while I probably have a good deal of control what a party does in the hills. So the unknown elements of risk are dramatically reduced. ²

When completing a risk assessment, the general process requires one to identify a risk, then to itemise the things that can go wrong — if you like, the severity of what could occur. For each such occurrence, you rate that severity on a scale (say 1-5, 5 bad). Then you assess the likelihood of this bad thing happening, usually on a scale of three, Low, Medium and High. Multiply likelihood by severity and then, for everything that is extreme, list actions to reduce the risk (while ensuring that all these actions will occur). Thus the identified risk events are reduced to become acceptable risks. Those actions will include avoidance: I've known people I'd rather NOT take on a trip—because I can't keep everyone safe with that individual along—so the mitigating action, on the assumption that I fail to persuade the problem child to not attend, might be to add an adult. Avoidance might be 'we won't do that' and 'we won't go there', but one solution might include a competence test to be passed first. And so on.
The weakest part of this process is the assessment of likelihood. The declaration of severity needs some tests suitable for the context; I might suggest that a High risk event could be translated as "If this happens we won't get to do this activity again", while Moderate maybe translates as "The event is suspended while we deal with this" and Low is 'one person is inconvenienced, perhaps two'. On occasion (in the hills, I've managed this) a medium risk event can turn into a positive learning experience for all, so that everyone learns that there was a risk and how to deal with it. Changing a hill route (and explaining to all why this is happening) would be a very simple example of this.


With biological risks we are far less certain and excitement is irrelevant. For a start, nothing is immediate. That means that even those who are clued up may well realise they've been putting themselves at risk only after they've made some actions (think of any first response moments), and discover whether they have erred only days later, far too late to say to oneself anything equivalent to the hill moment "Hang on, let's pause and think about reversing our route". I spent time sailing with Army and Navy colleagues and what happens there is that you discover you're being exciting when the boat heels in a way that is extreme and sudden, which not only makes you realise you've got things wrong, but that there is not much you can do to repair the situation immediately (pretty close to my idea of disaster). This doesn't happen with bio-hazards.

I'm also thinking of farming-type threats such as foot-and-mouth in the UK or the biohazards as perceived in NZ, where we simply accept that there are restrictions or required behaviours (mostly, don't go there); we comply or not until either the danger goes away or we are refused access entirely. This is a version of isolation: The risk is the people so remove the people from the situation.


For things like Covid, and we might include future attitudes to influenza or even the (more) common cold, we receive instruction from the experts in public health, which some of us understand and many do not. Medically we are safe if we cannot come into direct or indirect contact with an infected person. In such circumstance it is irrelevant whether a person knows that they are infectious. Statistically we are appreciably safe if we can stay two metres from any individual, on the understanding that airborne droplets are the vector. The other vector of which we need to be aware is collecting virus from contaminated surfaces, hence loads of washing and cleaning. Indeed, from there, suitable caution about anything carried into one's personal space, such as deliveries, food particularly. All the other procedures that we are told about (PPE for example) apply when the distancing is not possible.

But if we do not understand the meaning of the rules, then we cannot appreciate the risks, relative or absolute. Worse, most of us have no idea how to compare the risks we are told about. Still on Covid, I guess we're pretty sure we don't want to have it, mostly because we don't know how badly we'll be affected until quite a bit later. So the risk of being ill, possibly dying, is enough to prefer to avoid the possibility of catching it. Of course there are those who refuse to consider that they might get ill and I have long assumed that these are often the same people who don't much care if they contaminate others.

For example, changing the proximity rule from two metres to one: for a start, what have we been measuring as two metres? Distance from mouth to mouth? Closest approach of nearest two bits of body? The possible differences in answering that distinction for two metres is what makes my perception of one metre a problem; if one metre is between centres, then we are within touching distance, which doesn't quite occur at two metre centres.  Further, when I looked studies comparing such a change I found assumptions that risk at two metres was half of the risk at one metre; that makes the risk at distance zero extremely high. This is a bad assumption to make, and just from the modelling point of view an inverse square rule would be far more sensible, temporarily ignoring the persistent problem of what distance we are measuring. The inverse square rule would say that 1.4 metres was twice the risk of what it is at 2 metres, which would be translated as 1.5 metres for public consumption. But we are far from certain what we mean by 'twice the risk'. Twice <very small> is still <very small>, but twice <significant> is scarily large. There is a clear need for some absolute measure, then. 

As explored already (Essay 304), repeating something generally pretty safe is less safe. For example, medical procedure risks are established from historic records, such that there is a distribution (assumed Normal) and hence the distribution has a mean and standard deviation. A typical medical process will be declared safe when errors (failure events) fall outside 3σ, meaning 99.73% of the time this is safe. Which also means that at around 250 times this same process gives you a 50% chance of having failed by now. So there are tables one can look at that take a daily event of risk [Nσ, where N is usually an integer or plus a half, and σ is the standard deviation from the mean] and these tell you how long to failure; a 2σ daily event will fail on average every second week, a 4σ event about twice a lifetime.  But if the event is, for example, visiting a Covid patient and you're seeing 20 a day, then the 4σ failure is every second year. Which tells us a lot about the necessary levels of safety in hospital, because 4σ safety is well outside the usual experience. Quite how one sets up a routine as 'safe' to such exacting standards, I have no idea, as 4σ safety requires the sort of thinking that most of us rate as well past 'impossible' and well into the 'unreasonably paranoid' category.


So, if one has been, let us say, isolating while the coronavirus issue plays out, then the risks—the places in which failure-to-be-safe can occur—all lie in those instances where something crosses your house boundary. Which puts your emphasis on safety into being as certain as you can that what enters is safe; that you have dealt securely with what is not known to be safe e.g. the packaging. For added security the places where possibly contaminated surfaces could have permitted transfer have been cleaned, and that the surfaces that could have had transfer have been cleaned. This seems straightforward in principle. Having an idea of the expected survival time of virus on classes of surface does not replace cleaning. Yet I expect that for many it does exactly that; this is 2σ behaviour rather than 4σ, though I say that from a measure of paranoia more than data. If you consider yourself to be in a more at-risk category the actions for dealing with the risk are identical; what has changed is your measure of the severity of the result of the risk event occurs. Therefore your action, in terms of risk mitigation, is to be much more careful (than other folk at lesser risk) about the performance of those mitigating actions. You will be attempting to move the 2σ security towards 4σ.

If going out, then the risk is surely the behaviour of everyone else, because you are very well aware how safe you have been and intend to be. So the two metre rule applies and while outside this should be manageable.  I notice two extremes; very large spacing and no allowance at all. Those that translate 'two metres' to be 'as much space as I can have', are, from my own observation, to a large extent also the people who have taken 'isolation' to include 'no interaction'; these are people who will not respond to a cheery greeting and I fear for their mental health. At the other end of the spectrum are those for whom caring about distancing is simply not within their awareness, so one's best practice if caught within proximity is to escape as quickly as possible. The official advice has said we're at <undefined measurable risk> if within two metres for 15 minutes, not the 0.15 seconds it takes to whizz past on a bicycle, nor the 1.5 to 2 seconds you might be within range if walking through someone's bubble. Which last implies a difference in speeds of 8kph, for which I describe three situations; a cyclist overtaking a runner, a runner overtaking a slow walker (especially those with a dog or a small child) or two walkers passing in opposite directions with no change of speed. ³  

We have been given no advice what the risk is when briefly within the bubble of another person. There is a lot of evidence about relative risk in essay 304 but not very much about absolute risk. Yes, one is far more at risk if older, but is that relative term really comparing you with yourself at a younger age? For instance, if I accept that I might be at 4 times the risk of death or damage if I catch CV-19 at 65 than the same me at 25 (numbers picked out of the blue), that still doesn't tell me whether the 25-year-old is at risk and if I were the sort of person to have little social interaction at 25, why would I care about old folk catching it? In a sense, why is this my (25-year old) problem? I can see how it might be their's, not mine. And that explains the observed behaviour of so many young people; they simply have no reason to care about others. So my behaviour at 65 is simple; I do what I can to avoid me catching it; simple self interest.  But my expected behaviour at 25 is far harder to grasp; "I'm expected to do <all these undesirable things> to protect everybody else? Why would I do that? Give me some good reasons that are mine, not yours!"  Fundamentally, I think this attitude problem is not about risk, it is about caring about others.

Let's be clear, that the risk of catching Covid-19 is quite different from the death risk (survival probability sounds so much nicer) if you have caught it. Further, what absolute risks you can find start with the assumption that you are already ill enough to need to be in hospital. So for example, if you're so ill you've been admitted to hospital then it is relatively easy to find an approximate likelihood of survival. You can then scare yourself thoroughly by adding in (actually, multiplying) the several factors that make your own case worse.

Again, if you have ethnic difference from the bulk white British population, the risk (of dying if you've caught it) is apparently higher and the relative risk is quite well understood. But that still doesn't tell us what the absolute risk is, which is why the advice for action is exactly the same—only more so; you need to be especially vigilant for the same level of absolute risk. More care of the same few actions. 

I'd love to find figures for increased risk for levels of obesity. I pick on this because it is one of the few recognised disadvantages (increased risk factors) any of us can do something about. There is a clear correlation between the BMI of an individual and how severe a case of Covid-19 hits. To repeat, that's increased risk of death or long-term damage, not an increase in the likelihood of catching it.⁵  This is a major confusion, one of the many that has become clearer as time passes. I think part of the problem is that until we learn to listen carefully, we simply do not connect the ancillary, qualifying detail with the numbers that are given. We hear, for example 'obese people get it worse' and interpret that immediately as 'more at risk of catching it' when it would more nearly true to say that they are 'more at risk if they catch it' a small but significant difference in very missable words, almost between if and of.

So in terms of what-can-you-do, the risk for Covid-19 separates clearly into 'what you can do to avoid catching it' and 'what you could do to improve your situation if you were to catch it'. For the first this is a list we all know, but to which we could pay attention. For the second, we cannot change our ethnicity or our disadvantages but we could do quite a bit to reduce BMI towards the recommended zone. 30% of the population of Britain is obese or worse and, while there are calls to call this disease and not a life choice, it behooves us all to be as fit and healthy as we can without adding to your other risks, which might reduce to simply eating less. I was reading a site this week that recommended a shortfall of 500 calories a day between what you eat (input) and what you consume by movement (output). This sounds like a lot but it would produce a modest and steady loss of around a kilogram a week. More exercise will help change shape but loss of mass will only occur when output of energy exceeds input. Few of us can do that without recognisably eating less.


DJS 20200615

small edits 20200628



1 I discovered long ago that my capacity to enjoy a climb is very low indeed. The point at which I can tell you I'm going to come off in the next minute is depressingly early. Simply put: this is not exciting, beacuse the coming off moment is inevitable. Strangely enough I feel entirely different about climbing on snow and ice. I have learned to keep a short length of rope on my rucsac—it's always there for when I might need it and it would only be removed to be used. It is the right sort of length to help me clamber down a 3m drop but not so long I'd contemplate an abseil. I can abseil, but I choose not to. I can't/won't climb and I avoid the need. Quite why I find this so completely uninteresting I am not sure; I think it falls into the same class as cricket because it is exercise that generally doesn't include an aerobic element. Thus I like squash but not tennis, hockey but not golf, and rowing must have a sliding seat.

2 For those who think the hills are dangerous places, let's enumerate some of those things you reckon as dangerous;- 

1. Being caught out by the weather; you're already out in it, you've had plenty of opportunity to look beforehand and while you're outside, what exactly is preventing you from looking at the sky for an immediate forecast? Get rid of the earplugs for a start.

2. Getting into difficulties: Such as what, being caught outside by weather? Covered. Picking a wrong route? But you're there; why have you not turned around? Where are your 'escape routes? If you didn't plan this then you are right, you shouldn't be out. It amazes me how hard it is to (learn to) admit an error and turn around; yet, once learned it becomes really easy to do. No shame, just sense.

3. Accident: how extreme an accident do you have to have before you can't cope? How would you get yourself into such a situation? How often have you had accidents in your other experience? If that last answer is really few, how will this happen on the hills? Answer: it happens with people who say they can when they can't, people who say they're okay when they aren't, people with so little experience that they don't know what they don't know. You learn to plan significant flexibility when hillwalking with people whose abilities you don't already know. 

4. Equipment failure: again, how extreme a failure requires you to need external help? Sole coming off a boot? That's why you carry a roll of gaffer tape (for me, parcel tape is most of my first-aid kit). And you tell yourself off repeatedly for permitting any failure to have occurred. 

5. Inexperience: this is the hard one; carrying too much, attempting too much; failing to plan (and add a long list of things that can be failed). The worst here is failing to recognise when you have a problem and failing to react to the problem when it is eventually recognised. That way lies disaster.


3 A little light maths of around GCSE standard.

Imagine a circular bubble in plan around a body, radius 2 units. Plot a chord whose closest approach to the centre is one unit. No, really, draw it. That defines a sector of central angle 120º and of chord length 2√3 units. Call that 3.5 metres in 1.5 seconds, implying about 2.3 m/s or 8kph. 

Repeat, but allow the body in the centre to have size, say of half a metre radius, then the chord that takes a metre off the bubble gives a 3-4-5 triangle shape and the chord length is 4 metres. Which either increases the speed differential for a 1.5 second intersection to nearer 10 kph, or increases the time for an 8kph speed difference to 1.8 seconds. You would do better to worry about breathing the expelled air of the person you're so very briefly meeting.

4 However, young (under 60) patients [..] with a Body Mass Index (BMI) of between 30 and 34 were almost twice as likely to be admitted to ICU compared to patients with a BMI of less than 30. This likelihood increased to 3.6 times in those patients with a BMI of 35 or greater. ("Normal" BMI is 24.9 or below.)  Alkazeera

67% of men and 62% of women in England are overweight or obese, BMI>25. Essay 304

5  I read here that the PM's BMI is 36. Crikey, no wonder he had it bad and Matt Hancock didn't. Acceptable is 20 to just under 25. BMI is your metric weight in kg divided by your metric height squared, kg/m².  20-25 Okay; 25-30 overweight; 30-35 obese; 35+ morbidly obese.

Examples:  tallish guys 75/1.8² = 23,  smallish girls 55/1.6² = 21.5.  

Bodies you might recognise: 

BoJo 110/1.75²=36, the Donald 109/1.9²=30, Arnie 113/1.88²=32, J-C van Damme 80/1.77² = 25.5 (Stallone, same figures), Usain Bolt 94/1.95²=24.7,  Michael Johnson 77.1.77²=23,  Steve Cram light at BMI of 20, Mo Farah lighter 58/1.75²=18.9. Seb Coe is the same height but was even lighter, BMI=17.6. 

Hint for better running; reduce body fat = lose weight = faster running because you're carrying less. I'm no fattie and I'm 20kg heavier than Seb Coe for the same height and with a far smaller lung capacity. And most of you think of me as a good runner!

Email: David@Scoins.net      © David Scoins 2018