On Risk and STDs

[mattbell]

I've recently been frustrated that people either tend to dramatically overestimate or underestimate risks of particular activities.  Neither one is good... people can end up paralyzed with fear or regretting their recklessness.  The frustration around unknown risk came up recently in  discussions of sexually transmitted diseases.  There are a lot of scary stories of bad things that can happen to you, and ... well... they can happen to you.  But if your response to these stories is to just sit around and remain abstinent until marriage like the Bush administration advises, well, you're missing the point and you're missing the fun. 

I decided to fix that.  After a few hours of research online, I now have as good a sense as I can about the risks of various STDs.  There are nice ways of quantifying risk.  For example, you can think of the risk of a particular activity or lifestyle in terms of the number of days it shortens your life on average.  For example, every time you skydive, you run about a 1 in 70,000 risk of dying.  If you're in your twenties, each skydive will cost you on average about six hours of your life.  I have a coworker who has done about 4000 skydives.  His lifetime of skydiving has carried a three year risk that he's fortunately beaten the odds on, but for him, it's worth it.

-------

Here's how this kind of thinking works:

There are some people who are very worried about catching HSV1 (oral herpes, aka cold sores) because it can lead to the fatal disease herpes encephalitis.  Crunching the numbers, catching HSV1 costs you about 4 days (or, if you prefer, 16 skydives :-).  Not too bad.  Which is good, considering 90% of people have HSV1 by age 50. 

Don't want to get Hepatitis B?  Get immunized.  It will add 9 days to your life.   

The big one, though, is HIV.  Because HIV is so deadly, there has been a lot of research on how dangerous the transmission is. 

Just one act of unprotected vaginal sex (receiving) from an HIV+ partner costs you 73 days (or 292 skydives).  If your regular partner catches HIV, they will very quickly suck the life out of you.   Use a condom, and it goes down to 1.5 days per act.  Unprotected fellatio with an HIV+ man costs the woman 5.8 days per act. 

Now there's a great reason to make sure your sex partner(s) are trustworthy and tested regularly. 

Then there's HPV.  What's frustrating about HPV is that the actual risks in transmission are not clear, and there's no way to test guys for it.  While the chance of HPV-induced cervical cancer killing you is low if you get regular pap tests, it can still lead to unpleasant surgery.  What's actually more dangerous is oral cancer caused by HPV.  HPV-induced oral cancer is currently shortening the average American's life by 22 days.  There's currently no oral pap test, and virtually no education on it.  This is one reason why I'm so excited about getting the HPV vaccine approved for guys -- from what I can tell of the pathology, there's no reason it wouldn't work to prevent oral cancers as well as it works to prevent cervical cancer.

This reasoning can be applied to other kinds of risks too. 

A lifetime of driving a car shortens your life by 91 days due to the chance of dying in a car accident.   However, all you have to do is wear your seatbelt, and you get 57 of those days back.  Sweet.  I'm looking to buy a new car.  I can get side impact airbags for an extra few hundred dollars.  It turns out that a lifetime of side impact airbags saves me 5.3 days.  Not bad. 

-------

Now, one important thing to remember is that all risks are based on averages.  If you're really choosy with your partners, the Hep B vaccine may only save you a few hours.  If you frequent gay bathhouses, it will likely save you hundreds of days. 

Also remember that the risks listed here are risks of death... there is also quality of life to consider as well. 

-------

I'm now feeling a lot better about STD risks.  It's a lot better than living in the dark.

Comments

[inferno0069.livejournal.com]

I have a coworker who has done about 4000 skydives. It's taken about three years off his life, but for him, it's worth it.

Not quite. The 4000 skydives he's completed safely have taken no time off his life. The next 4000 he may be planning, while their outcome is still undetermined, will take an expected three years off his life, unless his chances have been modified by his experience.

[nasu-dengaku.livejournal.com]

Yes, that's correct. I will phrase it slightly differently so that's clear.

[crasch.livejournal.com]

Fascinating. What sources would you recommend for the statistical data? Texts you would recommend for learning to make such calculations?

[nasu-dengaku.livejournal.com]

I got most of the STD data off the CDC website. The calculations aren't hard at all -- expected # of days lost due to an activity = 365 * number of years lost if the bad thing happens * chance of bad thing happening.

[ferrouswheel.livejournal.com]

Are these lost days based on your age? or on the average lifespan of person?

Thanks for doing the research :)

[nasu-dengaku.livejournal.com]

I assumed that the person is relatively young and has 40 years of life left. In some case where the data was an average over all Americans, I had to assume an average adult, which has 30 years of life left.

I'm ignoring any rapid medical advances, singularities and whatnot.

[lynthia.livejournal.com]

Your brain is clearly still a fascinating place to be. :)

[nasu-dengaku.livejournal.com]

Thanks. :-)

[yami-mcmoots.livejournal.com]

The HIV prevalence rate in the United States is only 0.6%, which means that unprotected fellatio on a randomly-chosen American man will take only 50 minutes off your life. With a randomly-chosen Swazi man it would be 2.3 days.

[nasu-dengaku.livejournal.com]

Good points. Yeah, don't go down on random men in Swaziland...

I considered posing the statistic as you point out with the inclusion of the HIV prevalence rate. However, I wasn't clear enough on a couple of things:

Knowing the per-act risk with a random American is useful in figuring out how risky casual sex is with random partners. What's not clear is how much more likely people who have a lot of casual sex with random people are to have HIV. There's also the factor that hopefully once people know they have HIV, they'll be good enough to stop having casual sex with random people. Thus, the selection of people that you're having casual sex with has a huge impact. -- eg whether they're trustworthy, they get tested regularly etc.

I also wanted to figure out how quickly HIV would spread in a network of people who are all having fairly regular sex of different kinds with one another. This has implications on who you need to ask for regular STD results from.

[nasu-dengaku.livejournal.com]

Oh... this may be a lost cause at this point, but I was never able to download the zip of those image files from your website. I got a "file not found" error.

[chellmer.livejournal.com]

I have a few friends with HPV, and while it's not a big deal for most, two of them have had to have more aggresive treatments to cut the pre-cancerous cells out of the cervix. This can build up scar tissue and cause problems with getting pregnant. Oh, and it fucking hurts.

But like you say, you can't be afraid of everything.

[nasu-dengaku.livejournal.com]

Yeah... and these risks are purely related to death, so they don't take into account quality of life changes. As you point out, these quality of life changes can suck.

[neoteny.livejournal.com]

So how would you evaluate non-life-threatening aspects for chronic conditions such as HSV-2, or being a carrier for HPV? Those things won't likely shorten your lifespan, but will definitely impact your quality of life.

[nasu-dengaku.livejournal.com]

I currently don't take those into account. Yes, getting one of those conditions can make you miserable. I suppose you'd want to figure out how much it's worth to you to not have a particular condition (in terms of money, quality of life etc) and then apply the same formula.

[radiantsun.livejournal.com]

This is great, would you mind posting it as a comment here: http://www.sophisticatedrelationships.com/blog/?p=61 as it has to do with risk assessment and I think your math of this is fabulous!

[nasu-dengaku.livejournal.com]

I can do that. I also have a page of advice and conclusions which I should post as well.

[radiantsun.livejournal.com]

Awesome!

[easwaran.livejournal.com]

Where did the transmission factor come from for unprotected oral sex on an HIV positive man? I was under the impression that there are only a few cases where that's the expected source.

Depending on the coding involved, you might make some sort of form using actual actuarial mortality tables, and the actual 1, 5, and 10 year survival rates for various causes.

And what's up with the oral cancer stuff? I hadn't realized oral cancer was even very common at all, but your numbers suggest that somewhere between 1 in 400 and 1 in 40 Americans get oral cancer due to HPV, which seems remarkably specific.

Also, that other factor of the percentage of the population with various diseases would be a nice coefficient to be able to add also (especially if the numbers are available by gender, orientation, age, and metro area or something).

[nasu-dengaku.livejournal.com]

Here's where I got the figure. Note that there's some debate, but I'm being cautious and taking the higher-risk opinion.

Data here:
http://hivinsite.ucsf.edu/InSite?page=pr-rr-05&doc=pr-rr-05-02
Discussion here:
http://hivinsite.ucsf.edu/InSite?page=pr-rr-05

It would be very nice to be able to break down risks by demographics as you suggest.

[nasu-dengaku.livejournal.com]

And for oral cancer due to HPV:

http://www.oralcancerfoundation.org/hpv/index.htm
http://www.oralcancerfoundation.org/facts/index.htm

http://www.webmd.com/sexual-conditions/hpv-genital-warts/news/20070509/hpv-linked-to-throat-cancer

[mary-wroth.livejournal.com]

You sort of address this at the end, but I guess I am still hung up on the averages issue versus what reality actually looks like for the individual.

The main issue (for me) continues to be that when attempting to quantify the risks associated with STD transmission, it varies SO MUCH from population to population. Sort of like the Swaziland comment above. The risk calculations seem like they need to be tweaked on a case by case basis in order to make any sense or to be reliable guideposts for adjusting behavior.

To me - it seems more effective to identify the behaviors that reduce risk generally. Rather than trying to figure out the average risk in X situation, especially when the risk assessment of the situation might be really off or really accurate.

[nasu-dengaku.livejournal.com]

Yeah, that is one of the dangers of basing an action based on averaged risk calculations. Unfortunately such hard data is often hard to come by. I did a lot of thinking about risk reduction strategies as well... perhaps I should include them in a separate post.

[serolynne.livejournal.com]

I know we discussed this in private e-mail.. but I just wanted to add to your thread just how important the Quality of Life impact STDs can have to some people, and that it's not all about how many days it may take off your life.

Most folks I know who are concerned about HSV really don't take into account the unlikely fatal impacts.. but rather the annoyance/pain of re-occurring sores, the stigma of an STD, having to disclose to future potential partners, possible insurance coverage denial, taking regular medication to reduce re-occurance, etc. HSV isn't likely to kill you... but it can be a damned pain in the arse (or vagina, or mouth) for life.

Same with HPV. Of the 6.2 million new infections each year, only 4000 women will die of HPV-caused cervical cancer. Pretty good odds if you ask me. But what if the much more likely cost of HPV to you as a man was having a piece (or all) of your penis removed - would you still consider the impacts of HPV to just be 'unpleasant'? That's about what women face with it, and quite often. As a woman with HPV, I have lost my uterus, my cervix, my insurability as an individual (which limits my professional options) and not to mention the 5 years of stress of dealing with it which impacted my mobility, my relationship and sexual options, etc. And if I happened to be a woman who wanted to bear children (which I'm not).. it would be far more than just an 'unpleasant' consequence for the fun of having had 7 partnerships in a lifetime that included sexual intercourse.



Oh.. and another fallacy with looking at the risks in this way is that if you have a fatal car crash - you don't just loose 57 days of your life as your numbers are trying to downplay the consequences of. You loose the REST OF YOUR LIFE. And unless you happen to be 57 days shy of your natural life span... And what about if that car crash was *nearly* fatal and left you in a severely incapacitated stated for the rest of your 40+ year remaining life?


But really.. life should come with a warning label. 'Caution.. living is dangerous to your health.' Heck, I do jump out of airplanes, have sex, snowboard, etc.

[nasu-dengaku.livejournal.com]

As always, you bring up some good points in a variety of areas. Quality of life impacts are a lot harder to quantify.

As for the car crashes, I wasn't meaning to say that not wearing your seat belt saves you 57 days in an otherwise fatal crash. I was saying that, all other things being equal, people who wear their seat belts for live 57 days longer (on average) than those who don't. It's true that death is all-or-nothing, so you'll never lose exactly 57 days. However, the # of days is useful (at least to me) as a way of getting some sense of absolute risk.

I also have not found a good way to account for and compare the risks due to injury as you point out -- it's a lot harder to get hard data on permanent vs temporary, how severe etc, and the penalty varies by person. (eg an injury would bother a couch potato far less than a marathon runner)

[spoonless.livejournal.com]

every time you skydive, you run about a 1 in 70,000 risk of dying.

There are a number of reasons why this statistic is misleading.

First, there are so few skydiving deaths per year that it's not a good sample size. In 2007, for instance, there were only 18 skydiving deaths in the US (out of about 2.1 million successful skydives) whereas some years it is as high as 40... it all depends on the year. But the more important things is that skydiving fatalities are almost never due to chance... nearly all of them result from people who are trying to do stupid tricks, like swooping across the ground while barely touching it, and running into a building. It's extremely rare that a beginner skydiver dies, because they are fitted with extra safe equipment that has lots of extra redundancy checks and watched over like a hawk while they go. Only when skydivers have been jumping for years and years do the start to get careless... "trash packing" their chutes between plane loads, swooping, having "low pull" competitions with each other, and BASE jumping. All of these are much more dangerous than just going on a regular skydive... and that's mostly where the 20-some odd death per year come from. So for someone to see "1 in 70,000" (or even "1 in a 100,000") and to think that's actually the risk they're taking if they make one jump is ridiculous. I would estimate it's more like less than 1 in a million.

One statistic often cited by skydivers is that you have more of a chance getting killed in a car wreck than while skydiving, but this statement is somewhat misleading and needs to be phrased very carefully to be accurate. (I've heard some skydivers incorrectly say that jumping once is more safe than getting in your car to drive home after jumping, which is not true.) The more accurate statement is that jumping about 17 times per year is as equally risky as driving 10,000 miles per year in a car. So if you drive at least 10,000 miles per year and skydive less than 17 times per year, you have a greater chance of being killed in a car wreck than on a skydive:

http://travel.howstuffworks.com/skydiving8.htm

However, this is again based on pure statistics, and ignoring the issue of control/safety. Wreckless skydivers have a much greater chance of dying, just as wreckless drivers do. However, in the driving case, you have a lot higher chance of being killed by some *other* wreckless driver, which means that you have less control over your own risk while driving.

[nasu-dengaku.livejournal.com]

Has anyone done a plot of chance of death during a skydive crossed with # of prior skydives? I'm curious if increased skydiver skill as the number of dives increases reduces the per-skydive risk, or if the desire of veterans to try new things and push the envelope increases the per-skydive risk.

It's a fight between competence and cockiness. Which one wins?

[spoonless.livejournal.com]

According to this page the current fatality rate in the US for tandem skydivers (those who jump strapped to a more experienced skydivers' back) is about 1 in 540,000:

http://theblueskyranch.com/sta/tb7.htm

Interestingly, he says it was a lot more unsafe in the early 90's... 1 in 80,000 or so. But with both of these, the sample size is even smaller than for skydiving in general... so it's hard to draw too many conclusions.

Based on the "1 in 540,000", it looks like I should take back my "less than 1 in a million" estimate for a first-time skydiver, since most first-time skydivers do tandem. Tandem involves higher velocities than usual (over 200mph as opposed to 120mph) so fatality may be more likely in certain situations. But overall, you would expect tandems to be about the safest, since the important decisions are made by someone experienced, yet they are unlikely to be wreckless when a first-time skydiver's life is in their hands.

I also found this comment from someone on a forum, although I haven't verified these statistics he's quoting:

(http://skysurfer.com.au/forums/index.php?showtopic=2575&st=0&p=29724&#entry29724)
"Stats from Parachute Industry Association (PIA) a few years back - 99% of skydiving fatalities are due solely to human error (poor maintenance, preparation, go low, no pull, midair collision etc). Less than 1% is true bad lack (sudden wind gust, um...can't think of any others). A main parachute malfunction is not fatal. Failing to pull your reserve can be, or not cutting away (releasing main) thus getting entangled."

Anyway, my guess would be that experienced skydivers have more fatalities than beginning skydivers, but after searching for a while I can't find definitive statistics on this. Interestingly, a few pages I found said that hangliding has a 1 in 2000 death rate... but that's only based on 1 or 2 deaths per year out of several thousand hangliders... and it's supposedly also pretty consistently due to user error, and not uncontrollable or unforeseeable factors.

[spoonless.livejournal.com]

ok, I need to learn to spell "reckless" :)

another thought on experience... I'm thinking there might be a window, just after getting off student status (obtaining a license), where skydivers are no longer supervised and starting to get cocky, but really don't have enough experience to make good decisions... maybe this would be the highest at risk group?

[serolynne.livejournal.com]

I don't know if there's a study on it or not, I'll have to ask my bro (a skydive instructor) on that.

But I can say from personal exposure via my brother... all of the skydive fatalities and injuries I know of are from more experienced skydivers. Not exactly a representative sample tho.

[spoonless.livejournal.com]

all of the skydive fatalities and injuries I know of are from more experienced skydivers.

I was thinking the same thing... seems like every time I hear about one, it's an experienced skydiver... however, then I realized that might just be because the total number of jumps per year are mostly made by experienced skydivers. So it could still be more dangerous for beginners, I'm not sure.

[serolynne.livejournal.com]

So true.

I had someone pose this question to me:

'So, you'll jump out of an airplane... but you won't ride a motorcycle?'


You bet! When I jump out of an airplane I typically have one obstacle to watch out for. It's very large, can't be missed, and it's inevitable that I'm going to impact it. But, I have a high degree of choice and control over how I land upon that obstacle - otherwise known as Earth.

If I'm on a motorcycle, there are obstacles all over the place each with their own agenda (or lack thereof). I'll drive my car to get places, thank you - and get my rush at 13,000 feet.

[spoonless.livejournal.com]

reichart, myself, and a bunch of others were standing around last weekend discussing how dangerous motorcycles were. They scare me too (although I've never done it).

I think you have to travel something like 20 times further in a car before an expected fatality than a motorcycle... and a lot of times, motorcyclists die and it's not even their fault... the car just didn't see them.