Hacker News Reader

Comment by lmilcin | Hacker News Reader

I hope you understand that HIV takes years to decades to develop symptoms.

A lot of people could get infected before anybody realises there is an epidemic.

Is this strain properly detected by current HIV tests? If that's the case then we already have statistics about the HIV epidemic. No need for speculations.

A new strain, due to nature of exponential growth, can spend a long time before it noticeably moves the total number of cases and causes people to notice change.

Think about it this way:

Year 1) 10.000 normal cases + 1 new = total 10.001

Year 2) 10.000 normal cases + 10 new = total 10.010

Year 3) 10.000 normal cases + 100 new = total 10.100

Year 4) 10.000 normal cases + 1.000 new = total 11.000

Year 5) 10.000 normal cases + 10.000 new = total 20.000

Year 6) 10.000 normal cases + 100.000 new = total 110.000

You see how for a long time, if you were looking at a total number of cases, nothing was happening and then it suddenly became absolute, unmitigated disaster?

This is one of the unfortunate characteristics of exponential growth. If there is a new exponential growth that is masked by noise -- it could be very invisible for a long time and then suddenly become a huge problem.

---

Edit: here another example with more realistic reproduction rate (2x annually and delay of 5 years before developing symptoms)

Year 1) 10.000 existing steady state + 100 new = 10.000 with symptoms

Year 2) 10.000 existing steady state + 200 new = 10.000 with symptoms

Year 3) 10.000 existing steady state + 400 new = 10.000 with symptoms

Year 4) 10.000 existing steady state + 800 new = 10.000 with symptoms

Year 5) 10.000 existing steady state + 1.600 new = 10.000 with symptoms

Year 6) 10.000 existing steady state + 3.200 new = 10.100 with symptoms

Year 7) 10.000 existing steady state + 6.400 new = 10.200 with symptoms

Year 8) 10.000 existing steady state + 13.000 new = 10.400 with symptoms

Year 9) 10.000 existing steady state + 26.000 new = 10.800 with symptoms

Year 10) 10.000 existing steady state + 51.000 new = 11.600 with symptoms

Year 11) 10.000 existing steady state + 102.000 new = 13.200 with symptoms

Year 12) 10.000 existing steady state + 204.000 new = 16.400 with symptoms

You see, adding a delay to symptoms of exponential growth makes it even more heinous.

Year 11 when a somebody that watches statistics of symptomatic patients might notice a rise is already when you have over 10x more new cases than the steady state you think that is happening.

We could complicate this example (some cases get symptoms earlier than 5 years, etc.) but the basic reality is still the same -- when there is a noise and the signal is delayed, the moment you notice a rise in total number of cases and you suspect it being caused by an exponential growth, you should be worried that the reality is much, much, much worse.

By that same token you can’t get 10x growth per year and hide for decades. Even doubling every year would be 1M million cases in 20 years.

Hope you realise that mine was completely made up example just to show the basic characteristic of exponential growth in presence of static component.

I chose 10x growth factor to make sure the reader gets a point you can have absolutely disastrous epidemic brewing and still be invisible for many years.

With smaller reproduction rate and other confounding factors detecting a new epidemic like that could be even more difficult if you only look at total number of cases.

Sure, the point still stands.

If it doubles say every 18 months numbers get really obvious before the number of cases dwarfs exiting new cases let alone existing cases. Something like 3-6 years warning before the number of new cases double. T minus 6 years it’s a 6.5% bump, T - 3 years it’s a 25% bump, T-18 months it’s a 50% bump.

For a disease like HIV where existing cases dwarf new cases you get a lot of time to prepare before the number of existing cases significantly increase.

It’s particularly interesting because the trend with HIV so far has been towards less damaging variants, not more. Fatality rates for HIV are almost at point where there’s no longer any reduction in life expectancy, due to mix of better medicine, but also due to HIV evolving into less deadly strains.

To see a reversal in this trend is extremely note-worthy, and somewhat worrying.

Maybe one or two observations should not make us jump to a conclusion of reversal in trend.

Trends don't imply constant variability on each and every single observation...

When making statement like this in a political environment like ours you should consider the facts that underpin this. With the great majority of people practicing safe sex, there's evolutionary pressure on the virus not to kill its host before a passionate slipup, which may well be 10 or 20 years. Also, there's very effective medication available. None of these facts is true for covid. There's no pressure at all for the virus to become less virulent, it's an airborne disease, and what medications are available are not particularly effective.

Majority of people who practice safe sex do not affect HIV transmission. HIV transmits mostly through networks of people who do not practice safe sex.

Unlike a normal virus where people exchange it at random, here people that are more likely to get the virus also seek out other people that are more likely to get the virus. Normal r-value calculations do not apply because you no longer have homogenous population.

You do not need a lot of these people to keep the virus alive. A person that practices unsafe sex once is more likely to do it more than once, with other people who will do the same. In effect forming a network within general population.

While people who practice safe sex might become collateral damage (like spouses of people who practice unsafe sex or patients at a clinic that used untested blood) they are very unlikely to transmit it further (safe sex practices are very effective at preventing spread of HIV).

Also no need to bring politics to it. Viruses do not care about politics.

> HIV transmits mostly through networks of people who do not practice safe sex

This is correct. Over two thirds of HIV cases around 2009 were transferred genitally; over 80% if one includes the intestinal tract (rectal) [1].

> Viruses do not care about politics

But people do, and their behaviour influences how viruses like HIV spread.

[1] http://perspectivesinmedicine.cshlp.org/content/2/11/a006965...

Do you have a source about HIV becoming less damaging? I haven't heard about that before.

> It’s particularly interesting because the trend with HIV so far has been towards less damaging variants, not more.

False. HIV is as deadly as ever.

> Fatality rates for HIV are almost at point where there’s no longer any reduction in life expectancy, due to mix of better medicine

True. It is due to improved treatments, which are a mix of medicines (where the mix is designed to inhibit the development of resistance to the full cocktail)

I believe seroconversion symptoms happen weeks after infection. And testing for HIV is effective at that point.

Yes. Seroconversion symptoms are very common and there is a queer community term for it: "HIV Flu."

Most people do not immediately go for comprehensive testing the moment they feel worse for a little bit. And most doctors that see flu-like symptoms will assume you have a flu.

The reality is that almost every person diagnosed with HIV is diagnosed because:

1) They developed full blown AIDS

2) They got contacted by somebody who just discovered they have HIV/AIDS to get checked

3) They went through some routine unrelated test (for example they wanted to give blood)

Routine full panel std checks when you have a new partner or if you have an accident like say when visiting a sex worker are a thing, enough that you can order one at a bulk lab diagnostics service for relatively little (it's still expensive but not prohibitively so, and some nonprofits offer them for free)

People are offered life saving vaccines for free and still don't take it.

I don't want to throw completely baseless statistics, but I would guess over 95% of the world either a) does not have access to that kind of thing, or b) has other priorities, financially, or c) does not care about this because they are either not informed or they know this exists but is not important enough for them to actually do it.

And if you are from US remember, western countries where these things are easily available and people can generally afford it are only small percentage of the world's population.

Op was about circulation in Europe, where I am guessing they have these things.

And if people are not doing it enough in western nations where we do have access, then my posting this serves to increase awareness in this small hn community.

Reply on news.ycombinator.com