Dry January 2021

Is 2021 a good time to not drink booze for a month?

Reasons it is not: Things are tough and if a pint makes them less tough for you in a dark scary January rock on for yourself. This post is not trying to rag on anyone.

Reasons it is:
1. Health short term. Several studies have shown giving up booze for a month improves sleep and liver health. Reduces weight, cholesterol, Blood pressure and blood glucose

"Richard de Visser, University of Sussex, surveyed 857 participants of Dry January 2019 before and after their month of abstinence, finding 71 per cent slept better, 67 per cent had more energy, 58 per cent lost weight, and 54 per cent had better skin".
"Our study saw a weight loss of around 2kg, a decrease in blood pressure of around 5%, and improvement in diabetes risk of almost 30%." 

2. Speculative Coronavirus: Someone who has slept better, has lower blood pressure and weighs less is probably slightly less likely to have really bad covid effects. Those links at least suggest improving each should bring benefits. I can find no specific moderate alcohol and covid research but I would lay a bet that improving these biomarkers improves slightly overall outcomes. Even a very small effect is also fairly easy to achieve, all it is is a month of not doing something.

3. Pubs are closed. Meeting your friends for pints are the best drinks and you are not going to have those this January anyway.
Helping out pubs by going there through the usual January lull is good for them. But they are closed now so you can't do that this year.
Having to explain to your mates that you are on not drinking can be a bit intimidating. Though in my experience people are fine with it in practise. But you do not have that worry this year as you won't be out with them. These usual pub related reasons to not do dry January are not present this year.

4. Long term effects: Dry January does seem to reset drinking and reduce long term quantities. 'The research by de Visser found that in August the following year after taking part in Dry January, the 800 abstainers had cut back. The number of days they drank alcohol fell from an average 4.3 to 3.3 a week, the units consumed had slid from 8.6 to 7.1 per session, and frequency of drunkenness had fallen from 3.4 days per month to 2.1.' 

Replacement Options

You could really be thirsty. Drinking lots of water and other non alcoholic drinks to replace any booze you were drinking might make things easier.
Alcohol free beers have improved a lot in recent years. I think they used to have to evaporate off the alcohol. But now new yeasts can make beer tasting flavours without producing alcohol so this removal stage is skipped.
Alcohol free beers I have tried include

Lidl Non IPA is my favorite 9/10
BrewDog Nanny State quite nice though I could not imagine drinking loads of them 8/10

Heineken 0.0% Non-Alcoholic Beer decidely meh 3/10

Bitburger Drive grand 6/10

Budvar 4/10

Perlenbacher 0.0% 7/10

Beck's Non-Alcoholic I didn't like this. It tastes chemically 2/10

Carlsberg non alcoholic Better than Heineken 6/10

Guinness "Pure Brew" 6/10

Dungarvan Main Sail I really liked though it might be hard to get 9/10

In general the east European boring lagers do a non alcoholic beer and any I have tried have been grand.

I do not have great taste in beer. So get a bottle of each one you see and try yourself to find one you like. I do not see the point of alcohol free spirits. Red grape juice is worth getting if you drink red wine. Some people find they eat more sweets during dry January.

If you can get some sort of exercise challenge for the month I feel it gives you something useful to focus a bit on. Starting Strength weightlifting is good though it might be hard with the gyms closed. Couch to 5km is easier given the current circumstances. Some people swear by Yoga or hot baths to help them chill out.

That is my four reasons it might be a good idea to not drink this January. And a few replacement options that can help. If you do try Dry January let me know how you get on?

2020 books review

 I read 33 books in 2020. 

My favorite non fiction was We Are the Weather: Saving the Planet Begins at Breakfast by Jonathan Safran Foer it is about Climate Change and what the individual can do about it. It is slightly poetic and surreal which I like in non fiction books. Runner up was The Ancestor's Tale: A Pilgrimage to the Dawn of Evolution by Richard Dawkins. He is a really good writer and knows evolution really well.

My favorite fiction was Crime and Punishment by Dostoyevsky. I love his awful protagonists. On that note Exciting Times by Naoise Dolan is probably the book I would recommend that you might actually read. Pretty unpleasant, as in believable, characters, great dialog. And a quick fun read.

I love detective books. Classic stuff without autopsies and graphic murders. And I never get who did it. The best I read this year was The Inugami Curse by Seishi Yokomizo. If you know of any good classic crime fiction translated into English please let me know in the comments. 

Vaccinate the old fast

The CDC give Infection Fatality Ratios for people of different ages as

0-19 years: 0.00003
20-49 years: 0.0002
50-69 years: 0.005
70+ years: 0.054

This mens 5.4% of all over 70s who are infected with Covid they believe will die.

The number of people in these age groups in Ireland is

425k over 70s

55-69 year olds =720k 

18-54 is 1,350,000

0-18 about 1,200,000 people

The ages do not quite line up but they are close enough.

If all over 70s in Ireland got infected with covid 23,000 would die. Assuming the infections were spaced out enough that the medical system stayed working.

All 55-69 about 3,600 would die

All adults 1350000 * 0.0002= 270 deaths 

Children=36

This ignores long term issues that Covid could cause. But hospitalisation numbers track Deaths pretty clearly. And long term issues seem to track hospitalisations.

The numbers are pretty stark though and indicate to me we should rapidly vaccinate older (and vulnerable) people as quickly as practical.

What jobs have most risk of Covid?

This data is Norway but I will assume the US and other countries would be similar. The full paper is Occupational risk of COVID-19 

"The difference between the first and second wave may be due to changes in the test criteria. In the first wave, there was a shortage of test equipment, and patients, risk groups and healthcare personnel were prioritised for testing. In the second wave, close contacts and people with mild symptoms could also be tested. " so I'll use the figures from the second wave for the risk.

How many people have these jobs? The US bureau of Labour Statistics has good numbers on occupations here 

I took all these jobs and got the numbers of US workers with them and their chance of getting ill. It works out that about 10% of workers account for 30% of cases (in workers). That is still about 16 million people. But it does give a possible next group for vaccines after medical workers. I put the combined risk and numbers employed data into a spreadsheet here 

 

Detecting the Sex of a Chicken Egg

 'About seven billion male chicks - not wanted for meat or eggs - are killed around the world each year, usually in shredding machines or by gas.' This is really wasteful and cruel.

There are genetic tests that can be done on eggs to sex them. But they are not widespread yet due to the expense and complicated process they require.

In 2018 it was discovered that looking at the blood vessels in human eyes can be used to detect gender. This was not something widely expected by experts in looking at retinal scans. 'gender (area under the receiver operating characteristic curve (AUC) = 0.97)'.

I think it would be really interesting to make a dataset of chickens eggs, their sex when they hatch, size measurements, weight and photos of the eggs under various types of light. Possibly Infrared is the most useful kind of light.  With this it might be possible to make a classifier that can tell you if an egg is likely to be male.

Raspberry pi cameras have advanced a lot recently so making the rig to take these measurements would be much easier than it used to be. Tying these to the eggs eventual gender would at least involve some smart labelling.

Do you know anyone with the necessary chickens and interest to try make up such a dataset?

Short Supply of Flu Vaccines

 

On February 21, 2020 i blogged here about how 'Ireland should make the Flu Vaccine Free for All'. I thought this because the risk of Covid was high enough to make it really worthwhile this year.  It seems that there is now not enough flu vaccine to meet demand in Ireland this year. 

I wrote to various politicians. Including current Health Minister Stephen Donnelly.

	22 Feb 2020, 17:58
to Stephen.Donnelly

Dear Mr Donnelly 

Congratulations on your reelection as a TD. 

I would like you to consider it be Fianna Fáil policy for the government to offer free flu vaccines to everyone in Ireland this autumn. 

Firstly, Research from Canada says offering such free flu vaccination is cost effective[1]. Secondly, Ireland has particular issues with hospital overcrowding during flu season which makes reducing these cases very valuable to us. Thirdly, possible covid19 spread means any spare capacity we can get in our health system is particularly valuable this year[2].

Orders for doses of flu vaccines have to be made relatively soon. I understand the process of forming a government at the moment makes making such an increased ordering decision for use this Autumn more difficult.

Thank you for your attention.

   Best regards,

[1] Systematic review of the cost-effectiveness of influenza immunization programs by Eon Edward Kwokho Ting https://tspace.library.utoronto.ca/bitstream/1807/70706/1/Ting_Eon_201511_MSc_thesis.pdf

[2] http://liveatthewitchtrials.blogspot.com/2020/02/ireland-should-make-flu-vaccine-free.html

And his office replied

Dear David,

 

We thank you for your correspondence, and considerate thinking.

I have raised your concerns and proposals with Stephen Donnelly.

 

Kind regards,

 

 Office of Stephen Donnelly T.D.

There could be many reasons that there is not enough flu vaccine this year. But I think this exchange shows that 'No one thought or suggested before March that we would need more flu vaccines' is not a reason.

Judging Trump's Presidency

Early 2017 I gave some metrics here that could be used to judge Trump's presidency a success on his own terms. Dues to the pandemic I do not think it is fair to judge from 2020 on. But how did he do in the previous three years?
The metrics were

1. Better healthcare. Cover Everybody, cost less and have lower deducables

Didn't happen
'In 2018, 8.5 percent of people, or 27.5 million, did not have health insurance at any point during the year. The uninsured rate and number of uninsured increased from 2017 (7.9 percent or 25.6 million).The percentage of people with health insurance coverage for all or part of 2018 was 91.5 percent, lower than the rate in 2017 (92.1 percent).' 


2. More GDP Growth. 'If Trump can deliver an average of more than 3% over his 4 years in office I think an impartial observer would agree the economy has done well.'

Didn't happen



3. A balanced Budget.
Didn't happen
"The budget deficit increased from $779 billion in FY2018 to $984 billion FY2019, up $205 billion or 26%. The budget deficit increased from $666 billion in FY2017 to $779 billion in FY2018, an increase of $113 billion or 17.0%.^[10] The 2019 deficit was an estimated 4.7% GDP, up from 3.9% GDP in 2018 and 3.5% GDP in 2017. The historical average deficit is 2.9% GDP" 
4. Infrastructure improvements are a big part of Trumps promise. These are measured here
Didn't happen
2018 US was 14th in the world in world bank LPI rank. Infrastructure score 4.05
2016 US was 10th in the world. Infrastructure score 4.15

On none of the four metrics declared in January 2017 has Trump's presidency succeeded in, in the time before the pandemic.

Ireland really should make the Flu Vaccine Free for All

I argued here on February 21st that Ireland should make the flu vaccine free for all this year. This I argued because it would reduce overcrowding, increase productivity and mitigate some impacts of covid-19.

Today the minister for health has announced that more people will get the vaccine for free this year.

Defined at risk groups of adults is described here using 76 words. That is just too complicated. Many who qualify won't get vaccinated without clear messaging.
A simple message of "The flu vaccine is free for all this year. You can reduce the risk to you and spread in the community by getting vaccinated" would increase uptake even in the most at risk groups. The cost of expanding free vaccination to all, based on the experience of Ontario, would be about 60 million euro.

Flu compared to Covid-19 is a small issue. But it is one we have a fair amount of control over. Reducing the number of people having to go to hospital is really useful at the moment and free flu vaccines for all would do that.

Heatmap

Code at https://gist.github.com/cavedave/8ff22e94c882e9f6be933f99f2ae0b50

Give everyone LCT strips

The idea previously of LCT bracelets won't work. Arm temperatures are just not accurate enough to show fever. In some cases when you get a fever your lower limbs actually gets colder.

But that does not mean these cheap thermometers are of no benefit. There is an interesting model here from Paul Romer about how even an inaccurate test can reduce spread. Here is his simulation for a test with a 20% and 40% false negative rate results in large reductions in the spike int he number of infections.

and with 40% False Negative Rate

One such test might be the thermometer test for fever it is cheap but not very accurate. About 30% of people might not have fevers in spite of having covid-19.  It seems to be unknown how infections these asymptomatic people are.

Another paper on Wuhan says "Per person, these undocumented infections were half (52 percent) as contagious as documented infections yet were the source of two-thirds of documented infections" but undocumented does not mean asymptomatic.

There is also this interesting article on the issues with temperature checking in Asia. ‘Thermometer Guns’ on Coronavirus Front Lines Are ‘Notoriously Not Accurate’ it seems outdoor ones are pretty useless. And they need to be aimed very carefully.

All these bits of evidence combined seem to suggest that encouraging and facilitating everyone to check their temperature everyday at home might be worthwhile. LCT strips in particular are still available and are really cheap. 

Here is a supplier that can provide a million reusable LCT a month at a price of 3.6 cent each. There are 1.7 million dwellings in Ireland. Buying one for every dwelling in Ireland would cost 62,500 euro. 138.45 million housing units in the United States = 6 million dollars

Supplying these to households as a cheap if not very accurate test could well meet a cost benefit analysis.

Automatic Temperature Scanning 2

Here is a design to find fever in streams of people entering a workplace, school or public building

This can be done by hand but I want to make it cheaper

Ultrasonic detector tells someone is in front of it. (MLX90614)
Gets them to the right distance with a display (not shown)
Mirrored picture in an antiviral metal helps the person line themselves up in front of the IR sensor.
Flashes red if they are above 37.5C

Temperature Measuring Bracelet

If it is a good idea to check peoples temperatures a lot there are probably three distances you want to check them at

1. Personal level

2. Places lots of people go through like employee bathrooms

3. Public places like airports.

Each of these has different requirements.

For Personal thermometers the idea is you have a temperature sensor showing you do not currently have a fever. And if you do that is rapidly detected to reduce the number of people you infect. 

Personal thermometers they need to be really cheap. Not need much power. And be socially fashionable to show off.

Liquid crystal thermometer fits most of these requirements. These were suggested by Tara Whelan

Can they be made durable enough and fashionable enough to get people to wear them?

Earlier posts: Why temperature checking would help . 

Edit: the physiology of the arm means these won't work as bracelets.

Automatic Temperature Scanning

I blogged last night about how automatic temperature reading of people several times a day might reduce the number of infectious people out in public

Temperature Scanners Everywhere  Is that post if you want to see the reasoning. 

How about we put a remote thermometer and displays on bathroom mirrors?

What would happen if you 

1. Attached Infrared thermometers (slightly reassembled) to the mirror

2. Put a switch on it that turned it on when someone stood in front of the mirror

3. Output the reading to a display on the mirror

There are several instructable here on how to make the handheld type of forehead thermometer at a distance https://www.instructables.com/id/Arduino-Laser-Infrared-Thermometer/ this is much the same mechanism but in a different container.

All these parts combined and mass produced would probably cost around 10 Euro. Private and Employee bathrooms might be best for privacy reasons. There is no network connection or face recognition involved here just a very simple temperature reading.

Temperature Scanners Everywhere

'Why did Singapore’s containment strategy work, in such a high-density location?  Their program is mostly symptom/temperature checking-based'

It does seem like having symptoms really ups how infectious you are
"while we are learning that there are people who can shed COVID-19 virus 24-48 hours prior to symptom onset, at present, this does not appear to be a major driver of transmission. " WHO

'Some spread might be possible before people show symptoms; there have been reports of this occurring with this new coronavirus, but this is not thought to be the main way the virus spreads. ' CDC

Singapore is doing twice a day checks for some people
'All employers should require their employees to check their temperature at least twice daily, check for respiratory symptoms, and for those unwell to visit a doctor immediately.' During SARS they did something similar 'Singapore became the so-called country of thermometers: temperature monitoring was made mandatory in schools and temperature screening was instituted at entry to public buildings.'

In China 'In every business I enter (whether that’s a restaurant, supermarket, or retailer), someone takes my temperature and asks me to write down my phone number. Taikooli, the large outdoor mall by where I live, has blocked off most entrances to corral people into temperature check zones'

Airport screening does not work as it only catches you if you are sick at just the right moment so the checks have to be regular.

 What is the most efficient way to get peoples temperatures checked twice a day? Non-Contact Infrared Thermometer are cheap ($10). Paying people to hold them is expensive. Grabbing a thermometer others have grabbed sounds unhygienic. And trusting people to measure themselves expects a lot from them. But a thermometer held on a stand that checked people, that's cheap. Some logging would probably be necessary but lets work out the basic concept first.

Would twice daily temperature checks reduce R0? The idea is to reduce the period of time someone with symptoms is not isolated.
If so is there a way to set up a machine to check your temperature so they do not have to be manual?

Corona Virus and the Wheat on Chessboard problem

Exponential growth is really counter intuitive. But think of a chessboard where on the first square you have one grain of wheat. And every square you double the amount of wheat

Entire World Population number of wheat is square 34.

UK and Italy population are each square 27

US is Square 29

Ireland is square 24

Cases of Coronavirus currently UK, US and Ireland are about 18 squares behind the entire population number. Italy about 13 squares behind. Or 5 ahead of us. We at the moment move onto a new square every 4 days according to here. So Italy now is Ireland, UK or US in three weeks time at that really fast growth rate. Infection rates are not exactly exponential, eventually they start running out of new people to infect. But thinking about doubling is a useful shorthand to see how things can get scary fast.

I work with colleagues in China. The region they are in Liaoning has 43 million people. This would be Square 26.5. Liaoning, Mainland China
Confirmed: 125 Deaths: 1 Recovered: 110 Active: 14 These 125 cases puts them on square 8. They are at the same ratio as we are 18 doublings behind everyone being infected.

They wear mask all the time outside and in the office. Each person is only allowed to go to office one day in 3. All schools are closed.

They are at the same point as Ireland, UK and US and are on lockdown compared to here. Increasing this time to move to the next square like China and Singapore have done is really important

Drone Food Delivery

Manna sat they are going to deliver food March 2020 in UCD. 'Drone delivery company Manna plans to pilot its takeaway food delivery service on the campus of University College Dublin from the end of March.' Which will be great if it happens.

I can't find any videos of someone making an order and getting it delivered. Or anything close to a test run of that. Including from students seeing a delivery test and recording that.

All there is is slick PR video that is not a real delivery

I've asked some people involved if they had done test runs in UCD and I was told they did test runs in 'Last 3 years in wales and here. Tens of thousands of flights. UCD Just another step in the process.' 

Physical startups are hard. Airbnb and Uber took existing hardware, a mobile app and lots of negotiations with a regulator to create businesses worth billions. Manna seem to be negotiating with the regulator 'Irish Aviation Authority questions new Dublin food drone delivery service'. They could use the delivery company app or a version of that that let you indicate you were in UCD and what drop point you wanted to use. Getting that to only to all UCD students would be tricky but not impossible. If it not all students but only a few then it is a demo not a pilot. The drone they are using is well known so that is existing hardware. Both AirBnb and Uber did lots of physical testing of their product in the early phases and doing this with drones is really visible.

I could be wrong to be suspicious that this is going to be a demo not a full pilot for lots of people over an extended period. If I am wrong I will admit that next month. It will be cool if they can get this delivery pilot working then.

Ireland should make the Flu Vaccine Free for All

Ireland should make the Flu vaccine free for all people this year.

Firstly it is cost effective in general.
SYSTEMATIC REVIEW OF THE COST-EFFECTIVENESS OFINFLUENZA IMMUNIZATION PROGRAMS:A CANADIAN PERSPECTIVE
Runs the numbers for Canada and it is cost effective to freely vaccinate everyone
'Generally, from the societal and healthcare system perspective, vaccination was cost-effective. For pregnant and post-partum women, vaccinating all versus only high risk was cost-effective. For children (6 months to18 years), vaccinating all versus only high risk was cost effective, especially for infants, toddlers, and adolescents. For healthy working age adults (19 to 64 years), results were mixed, and sensitive to vaccine efficacy, uptake, and productivity loss. For adults with co-morbidities and healthcare workers, vaccination was cost-effective.'

Think of it this way a vaccination in Ireland costs 20 euro. About 10% of Irish people get the flu every year. Would you pay 200 euro not to get the flu? In lost earnings alone it is probably worth that.

Secondly Ireland has an overcrowding crises in its hospitals.
HSE apologises to those affected by overcrowded emergency departments as some hospitals ban visitors
It said the flu season arrived up to four weeks earlier than last year and has lead to a  “significant surge in patient demand”.

Flu and bed shortage responsible for record trolley numbers, says Taoiseach

Free flu shots will not prevent this overcrowding but it will reduce it. The evidence from Canada is based on a cost benefit analysis that does not include the costs of treating people as badly as we do in overcrowded hospitals. Which makes vaccinations more cost effective here. Vaccination would not be a long term solution to all the overcrowding problems but it would reduce overcrowding somewhat this year.

Lastly there is a non trivial chance that the COVID19 virus will lead to a surge of sick people in Ireland As such it is vital we have as much spare capacity in the medical system as possible. Free flu vaccination is one quick way to add that capacity by reducing demand.

Many in Ireland are already entitled to free vaccination. Roughly 1 million of the 4.8 million of us. 650,000 of us are over 65 for example and people with certain health complications or jobs are also entitled to free flu vaccination. At 20 euro each the remaining 3.8 million of us would cost an extra 76 million euro. Which in health spend is low.

Because flu vaccines are cost effective, especially once loss of earnings are taken into account. Because our medical services are overstretched. And especially because of the risk of COVID19. Ireland should order extra flu vaccinations this year and offer them to everyone for free.