Week 4 at AFHSC

I realized today that it's been a few weeks since I've updated anything, so here goes. Things are going well at the Armed Forces Health Surveillance Center. I got data on hospitalizations for poisoning at the end of the second week, so I've been spending my time since then staring at numbers and analyzing data.

To do all this analyzing, I'm using a statistical software called STATA (I don't know what, if anything, it stands for). I used STATA last year in school (yay for flashbacks to Epi 3...), so I'm somewhat familiar with it, which is good, since we're getting our STATA review tomorrow. Yeah, about two weeks later than I needed, but whatever.

I spent most of my first day with the data playing with the software, trying to remember simple commands and trying to figure out why I couldn't get anything to work the way I wanted it to (I eventually figured it out), and then the next few days doing some more complicated analyses. Since then, I've been working on more complicated analyses and putting my presentation together.

My project is actually in two parts: the first part is to describe who has been hospitalized for poisonings (both drug overdose and toxic exposures, such as to petroleum or lead); the second part is to figure out which poisonings were intentional, and what the differences are between those who purposely poisoned themselves and those who didn't. The first part was easy enough; there are a few simple commands (that I had to do over and over again) to do all the descriptive epidemiology, but the second part required much more thought (and probably isn't all that accurate).

As far as describing who in the military has been hospitalized for poisonings, there have been almost 15,000 over ten years. There are some minor fluctuations from year to year, but for the most part, the rate has been fairly constant. While there were more men (because the military has more men), the rate was higher for women. The rates are also highest for the lowest ages (17-19), lowest ranks (E1-E4), those in the Army (because we're Army strong...?),  least education, never married, and white (although the differences in races were pretty small).

With the second part--figuring out which poisonings were intentional--I used a few different things. The first are that there are certain diagnostic codes that can be used to say that a hospitalization was due to suicide attempt. The second is that there was a paper a few years ago that used a combination of psychiatric diagnoses within six months of the hospitalization to say that it was likely intentional. They did this in a civilian population, though, which has different baseline rates of psychiatric diagnoses than military populations.

Long story short, according to my calculations, 63% of all hospitalizations for poisoning in the military over the last ten years have been intentional. Risk factors include being female, in the Army (again, we're Army strong), multiple deployments, being married (yes, married people are at a higher risk of trying to poison themselves), and working in healthcare. Protective factors--those that make people less likely to try to poison themselves--are age over 40, increasing rank, and being in the Air Force or Marine Corps.

So, it's been a lot of analyses and a lot of math over the last few weeks, but I actually have some results to present, so that's a good thing. Pretty much, all I have left for the next two and a half weeks is to do my presentation and write my paper. Fun and exciting, I know.