SARS is serious
During the SARS outbreak, I have become deeply concerned by the failure of medical professionals to understand basic statistics.
By 14 April, 1190 people had contracted SARS and 47 of these had died. Such numbers have been widely cited as meaning there is a 3.9 per cent chance of death. The calculation is this: the chance of dying equals the number of deaths divided by the total number of cases, including the number still sick, and the result is multiplied by 100. But this must be an underestimate, as some of those who are still sick might die.
The correct way to estimate the chance of dying is to divide the number of deaths by the number of deaths plus the number recovered, and multiply by 100. By this calculation, the chance of dying becomes 9 per cent worldwide, and 17 per cent if China’s suspect figures are excluded. The rate for Canada has been as high as 33 per cent. These figures show SARS is an extremely serious disease that warrants aggressive quarantine.
Debora MacKenzie writes:
鈥� The standard measure of the deadliness of a disease is the case fatality rate, defined as the number of deaths divided by the total number of cases. The case fatality rate, which for SARS stood at 5.9 per cent on 24 April, will indeed be lower than the chance of dying so long as a significant proportion of the total number of cases are still sick and might die.
Epidemiologists do realise that the case fatality rate is likely to be an underestimate at this stage in the epidemic, even if they do not mention it. But while using the case fatality rate is not ideal, accurately calculating the chance of dying is not as simple as it appears. For instance, if people either die in a week or take several weeks to recover, using Lassesen’s method would underestimate the proportion of sick people who will yet recover and thus overestimate the chance of dying.
All figures relating to SARS need to be treated with a healthy dose of scepticism at this early stage. If authorities have not reported all deaths, the case fatality rate could be higher, while if infected people with only mild symptoms have been overlooked, the rate could be much lower.
DU lingers on
I’m glad to see that New 杏吧原创 is drawing attention to the immediate and longer-term dangers of widespread contamination with depleted uranium munitions in Iraq, and previously Kuwait and Kosovo (19 April, p 4). However, there are a few points I’d like to direct your readers’ attention to.
Saying that tank shells are “tipped with depleted uranium” is a pretty severe understatement 鈥� a 120-millimetre tank shell contains about 5 kilograms of uranium-238. The website of the Federation of American 杏吧原创s () has more complete information including diagrams of these weapon systems and munitions.
The scientific picture is also not nearly as uncertain as the Royal Society report tries to make out. I suggest you check the Uranium Medical Research Centre’s website (). The European Committee on Radiation Risk also released a report in February on the effects of low-level and internal radiation which re-examines the epidemiological basis for assessing radiation risks to humans. It illustrates that the “standard” of the last 50 years unofficially established by the International Commission on Radiological Protection is deeply flawed, underestimating the risk for low-level and internal exposure by a factor of between a hundred and a thousand.
It is also worth pointing out that in Iraq, these munitions have been most heavily used in urban areas and along main roads. The US has announced that it will not even be attempting to decontaminate Iraq. Also, the US Army’s Jefferson Proving Ground in Indiana, where uranium-238 munitions have been tested for some decades, has been declared a “National Sacrifice Area” 鈥� because of the technical difficulty and cost of decontamination.
Kidney cleanout
You cite a three-day study showing that any beverage is as effective for keeping the body hydrated as eight glasses of plain water (5 April, p 26).
Water is passed through the kidneys, whose jobs include maintaining proper balances of water and many different acids and salts in the body, and removing toxins and waste. Consumption of plain water allows the kidneys to perform this function while passing water with lower concentrations of the materials being eliminated. Most other beverages will result in higher elimination concentrations, because they contain additional substances that must be eliminated, such as phosphoric acid in soda pop, or sulphites in table wine.
While the cited two groups of athletes drinking plain water or drinking only beverages might have exhibited the same hydration level after three days, those on plain water were giving their kidneys a rest.
Three days is not enough time to see kidney failure problems emerging. Things might be very different if the study had lasted for decades instead of only three days.
I have personal experience with a family member who certainly drank her eight glasses of soda pop every day. She died of kidney failure.
Test them all
The arguments about genetically modified crops seem to be based on the proposition that it is possible for the technique to be proved either safe or unsafe (5 April, p 44).
This is in the interests of the proponents of GM organisms, because once you accept the idea that altering the genetic code of a plant won’t affect how safe it is, there will be little opposition to the creation of all sorts of new plants.
It is less in the interests of those who disagree with the use of the technology, and so I was surprised that Colin Tudge focused on this theme. Surely each GMO should be treated and tested entirely on its own merits, and the debate we are having about GMOs in general should take place for each new variety?
We need filters
Your correspondent Tim Lister, in failing to see the necessity for better spam filters, has himself missed the point (12 April, p 28).
Whitelisting (the practice of only accepting mail from known sources) does not work because many email users, especially businesses, expect 鈥� indeed, wish to 鈥� receive mail from people they do not know. They may be potential customers, afterall.
Blacklisting does not work because many spammers use a “one-time” sender’s name for each message.
The world does need better spam filters, and I am pleased to note that such methods are now becoming widely available.
Clocking on
I thought you might be interested to know that in 1956, IBM 鈥� which was then heavily into timekeeping systems 鈥� developed a pendulum clock that was radio-controlled by a short-wave signal from the National Institute of Standards and Technology radio station WWV. The concept for this clock was about the same as that described in your article about antique clocks going “nuclear” (29 March, p 18).
It was produced for only about a year. IBM then sold all of their timekeeping business to a company by the name of Simplex. The son of a former employee of Simplex donated one of these clocks to us and we have since put it into operation and displayed it in our lobby.
Superunlikely
The example of room-temperature superconductivity that you report is unconvincing (12 April, p 17). It is clear from the original paper by Johan Prins that it would be quite impossible to detect superconductivity with his measurements.
In his experiment he found that the current through his crystal did not change detectably when the thickness of the vacuum gap between the diamond and positively charged probe was reduced to zero. Since he was applying 1000 volts and drawing only half a milliamp, the resistance of his circuit was 2 megohms. And since the noise in his current was at least 1 per cent, any resistance in the gap less than 20 kilohms would be undetectable.
If the gap were made of copper the resistance would be about 0.04 micro-ohms, so the only conclusion that can be drawn is that the resistivity of the electron gas is no greater than around 1011 times that of copper. This cannot be regarded as evidence of superconductivity.
Blooming crystals
The article on flower-shaped crystals of indium nitride suggested these were the first flower-shaped crystals to be created (1 March, p 15).
In fact, this is a good example of man imitating nature. Natural, flower-shaped crystal groups much larger than the bonsai crystals you illustrated have been known to mineralogists for centuries. For example, the delicate and unbelievably symmetrical groups of haematite crystals found in the Swiss Alps are known as iron roses.
Joanna Marchant writes:
鈥� Green is correct to say that a number of crystals reminiscent of petals and flowers occur in nature. To clarify, the bonsai crystal we described is reported to be the first crystal group that consists of 6 petal-like crystals arranged around a circular centre.
TB testing
We were gratified to read your report of the “new” tuberculosis test that could soon replace the familiar 100-year-old “tuberculin skin test” (12 April, p 21). Your readers might like to know that a very similar test already exists, has been approved by the US Food and Drug Administration and was launched worldwide years ago.
Like the ELISPOT test developed by Ajit Lalvani’s team at the University of Oxford and described in your article, our company’s “QuantiFERON-TB” test detects interferon secreted exclusively by the white blood cells of individuals infected with Mycobacterium tuberculosis, the bacterium that causes TB.
Not only is our test already “out there” rather than in the trial stages, it’s simpler because we can test whole blood without having to separate out the white blood cells. One analyst can perform hundreds of QuantiFERON tests in a single day.
Lalvani’s findings do illustrate the value of a test based on proteins absolutely specific for TB infection, however, and we’ll soon be launching a version of our test embracing these improvements.
Man with balloon
Discovering the reason behind the counter-intuitive effect of damage to the brain as it floats in its saline solution was brilliant work (5 April, p 21). However, there is a very simple experiment anyone can do to give the same fascinating result.
Buy a helium balloon with a ribbon, tie it down in a car (so that it isn’t touching a seat or the roof) and (with the windows closed) drive around 鈥� or preferably get someone else to drive around and make sure their view is not obscured and that they are concentrating on the road, not your balloon. Better still, do it on a bus.
When you accelerate, it moves forwards in the car, when you go around a left hand corner, it moves left and so on (the opposite of what happens to, say, something dangling from the roof). I have been fascinated by this for years, but never imagined there was a practical issue associated with it. And most people don’t understand why I drive with a balloon in the car.
Going to the dog
Following up your correspondence on pubs with names like The Office (29 March, p 30), has anybody told you about the little town of Chatteris in Cambridgeshire where there is a very handy pub called Walk the Dog?
“Yes, dear, I’m just going to鈥�”