Friday, July 11, 2008

Data Fiend

Tonight at dinner my husband referred to an incident from ~ 10 years ago that he described as "the most fascinating psychological window into your soul", meaning my soul. I hadn't really thought about it that way, but I can kind of see what he meant. So here it is:

When I was pregnant with my daughter, I had a very severe case of gestational diabetes. This was a big surprise because I had no apparent risk factors. I wasn't overweight and had no family history of diabetes. Doctors were also surprised by the severity of the diabetes. I had to take rather massive doses of two types of insulin by injection twice a day. I followed an extraordinarily strict (boring, depressing) diet exactly, but it took weeks of monitoring and ever-increasing doses to get things under control.

At various times of the day, I had to prick my finger and do a blood sugar test with a small monitoring device at home. Between these jabs and the twice daily injections for 10 weeks, I was a total pin cushion, and I was not always very cheerful about it.

I had a little booklet in which I recorded my blood sugar level, but I also started keeping track of the results in a spreadsheet and I graphed the results every day. I got interested in the shape and magnitude of some of the blood sugar highs and lows, but my initial sample spacing (in time) was too rough to get a satisfactory graph of these spikes, and there were other aspects of the data that I didn't understand when I did the minimum number of recommended tests.

So, despite my loathing for jabbing myself in the finger with a sharp object, I started collecting more data. I tracked the blood sugar spikes so that they were defined by more than one point and I could really see their shape and I was certain of their maximum values. I collected data day and night. I dreamed of a device that could provide a continuous readout of my blood sugar and make perfect graphs. Even with my primitive data collection techniques, however, I made beautiful graphs and I did things with the graphs in terms of how I analyzed them over different time periods and how I displayed the data. I was obsessed with these graphs.

Part of what fascinated me about all this was the fact that I had so much control over the data. In my research, acquiring data can be a very time-consuming and expensive process, and it is not always immediately clear what the results mean. With my blood sugar data collection project, I could get as much data as I needed and the only cost - other than a bit of pain and some scarred fingers - was the price of the test strips.

The first time I brought my graphs and spreadsheet to a doctor's appointment, the doctor was stunned. He called all the other doctors and nurses over to look at it. He asked my permission to make copies and fax them to other doctors. He asked me to start sending him my graphs between appointments. He stopped talking to me like I was a slow child and started discussing with me what the data might mean. He gave me suggestions for ways to get more useful data. All of this helped get me through a difficult time.

Although the anxious and painful aspects of having severe gestational diabetes are not something I want to remember, I very clearly remember the excitement of acquiring and graphing the data. I think that is what my husband meant about the "window into my soul", which I suppose must be a scientific soul.

My daughter was born completely healthy at 7 pounds 10 ounces, although there were some difficult moments at the end. I've heard that most women forget the pain of childbirth (and hence are willing to have more children, ensuring the survival of the human race), but it is difficult for me to forget because my daughter was born at exactly 5 pm and the doctors were listening to NPR news in the delivery room. My daughter arrived as the NPR top-of-the-hour theme song snippet played, so every time I hear this song (i.e., almost every day), I am reminded of that moment.

38 comments:

Anonymous said...

fantastic story! I LOVE your blog. btw I had gestational diabetes with my first baby, but it didnt occur to me to take extra data and graph it. I did notice that it is really true that refined suger and starch made my blood sugar spike, while natural fruit sugars, even cantalope, did not. This annoyed me, as previously I had argued with health foody types that sugar was sugar.

Anonymous said...

Oh my goodness, it is so funny to read this! I have a different health problem, and always ask for printouts of my lab results- and I have also arrived at a doctors appt with graphs in hand, I thought I was the only one! (although, i never tried to get extra data!)

After bringing the graphs and knowing the doctor for a few years, I had a particularly funny moment- there was a med student observing (well, pretending to be another doctor), and my real doctor was explaining the side effects of something, using all the appropriate terminology and telling me about some recent papers on it. The student looked totally shocked, until my doctor turned to him and goes "its ok, she's a scientist". I always hope that student remembers that you should never underestimate your patients!

Ace said...

Amazing story. Thanks for the blog.

estraven said...

Great story, and congratulations for your courage! I hate blood tests, and certainly wouldn't have made extra ones.

On the other hand, the day after childbirth I tried to explain to a very confused neonatologist how I wanted to use the facts
1) my blood group is A+;
2) my husband's is 0+;
3) both twins are 0-;
to estimate the conditional probability that the twins be identical. Apparently we each have an instinct for our own scientific speciality?

Mister Troll said...

I made similar plots of my wife's temperature when she had a vicious case of mastitis that refused to respond to antibiotics. Not being a scientist, she didn't see the point of the graphs. Too bad.

We are glad to hear your pregnancy ended well, and you and your daughter were healthy. But... didn't you get to pick what to listen to on the radio?

RoadGrl said...

My son was diagnosed with type I diabetes last November... I too am somewhat data-obsessed, and look for blood sugar trends with food types, etc.
He's hopefully getting a continuous glucose monitor soon, then I can really trend him out!
I've gotten my husband (former engineer) to critically evaluate meal makeup as a factor in glucose spikes.
thanks for sharing

Tammy B said...

I am a pregnant scientist at the moment, and I have had to monitor my blood sugar, and I also found it fascinating. My doctor was impressed with my spread sheet, but I hadn't thought of plotting out the curve of growth for the blood sugar reaction! What were your times bins?

I don't seem to have a problem with the diabetes, and I am to the point of questioning the methodology of my doctor because their measurements don't take into account the duration of a meal or that fact that if I have dessert, it is at the end of the meal and closest to the time when I am taking the measurement.

I'll try doing a plot of the typical blood sugar spike this weekend and see what I find!

Candid Engineer said...

Great post. I bet the doctors loved that one, because how could they possibly collect that data outside of a clinical trial. You gave them a real gift.

During the last year, I have been extremely tempted to go off of birth control so that I could practice natural family planning. Not for any religious reasons- no- but because I am inexplicably interested in gathering my basal temperatures to make graphs about my own body and see the wonder of fertility with my own two eyes. I am sure that if I told this to anyone (except maybe you), they would think I'm a total nut job.

Anyway, I'm with you on this one.

PhysioProf said...

Great story! Being a data junkie kicks ass!

Lisa said...

FSP, you are so cool! I heard at a talk that they have some meters that need less blood, so you can, supposedly painlessly, prick your arm instead. I've never measured my blood sugar by either method, though.

Candid engineer, natural family planning graphs were fun for me. I used to take my temperature every morning just after waking up at the same time (or I'd get up, get my temperature, and sleep in). I put it on graph paper right away but always wanted to transfer it into a spreadsheet. I am too lazy so far, though. I haven't done it recently since I am breastfeeding and there was no point until a couple months ago. (we were able to time the baby's due date within a week of when we wanted it, which was mostly luck but I think the temperature graphs helped a lot). Anyway, you could still take your temperature every morning on the pill, and see if there's anything interesting there. I bet there is. Also, beware the natural family planning books--you'll have to overlook/ignore a lot of the religious stuff if you get one similar to the one I have, which was quite bizarre in some parts.

Jenn said...

Candid engineer hit on one of my scientist loves... basal body temperature tracking for family planning. I'm such a data geek, but I love it! And my doctor actually takes the charts seriously too which is an added bonus. Great story. I love reading your blog. :) Thanks for writing!

drdrA said...

Great story! Strange timing as yesterday I posted about my own pregnancy/data junkiness...

Intuitive Investigator said...

Truly a window to your soul! Thanks for the wonderful blog! I love reading it!

Isis the Scientist said...

That is amazing! A few years ago, when I was trying to lose the weight I had gained after getting married, I calculated the error of my bathroom scale by taking repeat replicates.

Perhaps this is the approach to take with doctors. I never introduce myself as "Dr." when I take my son to the pediatrician, despite my appointment at a major university in the department of pediatrics. I am amazed at how many physicians talk down to their patients. Perhaps we should start teaching patients to graph their own body data....

Cloud said...

I didn't have gestational diabetes, but I do remember getting ridiculously fascinated by the biological basis for minor things in my pregnancy and having wild thoughts of doing the research and writing a book about the science behind things like food aversions, etc.

@Anonymous at 12:43: refined sugar is sucrose and fruit sugar is fructose. Sucrose is a disaccharide made up of fructose and glucose. Your body's systems are set up primarily to maintain stable blood glucose levels, and I think (but am not sure) that it is only blood glucose you are measuring when you measure blood sugar. Also, the body uses slightly different pathways to extract energy from different sugars, because they are different chemicals so different enzymes are needed to catalyze the various conversions. Some people have deficiencies in an enzyme that is only involved in use of one type of sugar, and sometimes these deficiencies can cause big problems if the person eats the wrong type of sugar. So in fact, one type of sugar is not equivalent to another, at least not for all people. However, I'm not aware of any data that indicates one type of sugar is "better" than another for the average person with no enzyme deficiencies.

Isis the Scientist said...

And a note to Candid Engineer, I have used natural family planning for 8 years with my husband. I have 1 child who was planned down to the day he was conceived. If you are a data junkie, I think this method is right up your alley.

sandy shoes said...

I'm currently doing something similar with my blood pressure readings. It's remarkable how differently the doctor treats me now. I'm awfully glad I don't have to prick myself to get data, though.

lost academic said...

Today your entry makes me enormously happy for so many small reasons.

Arlenna said...

This is awesome.

I had some spinal x-rays recently at the chiropractor, and I keep staring at them obsessed as well. I always want my doctor to give me actual data files for all my scans and numbers for all my tests, but I haven't found a doctor who talks to me like an adult yet.

The awesomest thing ever would be to have a full-body MRI and the software to troll through it. That would keep me occupied for years.

Anonymous said...

one of the professors in my grad school dept had an (in)famous matlab file called hiswifesname.m, plotting hormones and temperatures and things, with which he could predict her ovulation.

ScientistMother said...

Fantastic story. When monkey was born I really had to adjust to the fact that he wasn't an experiment. When he would not go down for a nap, it would not be unusual for me to state " I am not sure why he is not sleeping today, I've done everything as I did yesterday but the result is not repeating", I had graphed is eating and sleeping patterns. I have since learned that with monkey sample size is always a n=1

labbrat said...

Oooh, I am now very excited about plotting basal body temperatures for my future family planning years. Also, I have been thinking for some time that I would rather have a picture of my fetus' karyotype from the amniocentesis than a picture from an ultrasound. Will have to pick a doctor based on his/her willingness to fork over this data!

Paralith said...

You are my new hero.

Gaia said...

Great story. In your shoes, I would have done the same thing and would have been tempted to turn it into a fun (and possibly TMI) lab exercise for my time series analysis course...

Anonymous said...

We have a picture of the karyotype from amnio, for both our kids. And yes, I did get a kick out of it.

(but not everyone has karyotyping done, especially these days, when the ultrasound/blood test combo is a pretty good method of detecting Downs).

(and yes, FSP is my hero, to. In fact, I really really wish that we're all going to meet her some day. I want to put a picture of her up next to my picture of Marie Curie).

Science Cog said...

Nice story. You are a good writer. I read that most women who get gestational diabetes go on to acquire it in their forties or fifties. I had borderline gestational diabetes managed by frequent blodd-sugar monitoring and portion control. I was wondering if you managed to avoid diabetes to date and if so what is the best thing you are doing to keep it at bay. I'm rather worried about my current lack of motivation to eat healthy.

Female Science Professor said...

I have a fasting glucose test once/year, and so far so good.

Kathy said...

oh, dear FSP, you make me smile, and wish that I had been in a doctoral program at your school (instead of mine)...

you are such a neat person. :-)

Dr. MCR said...

Thanks for another great post; what an interesting shift with your doc! I had essentially the same thing happen RE: threatened preeclampsia and blood pressure; suddenly there's a realization that you actually *know* something about data/physiology/immunology and your doc starts to talk to you more like a peer *and* patient.

BTW, I am a new blogger, and I really enjoy your blog! I hope you'll check out my blog: http://phdandparenthood.blogspot.com/ - Thanks again for a great post!!

Anonymous said...

hey, that's great!

my husband (also scientist, with bad teeth, alas) analyzed his gum measurements from the dentist's office. You know how they take many, many measures of gum depth, and do this over time... apparently they don't actually analyze the data, they just eyeball it to see if it is getting better or worse. He photocopied their sheets and discovered a small, but significant, improvement in gum health (with repeated measures ANOVA, of course) after a series of extremely painful gum treatments. Viola!!!

Aspiring FSP said...

It's so great to get a glimpse into academia through your blog - as a humble graduate student looking in, it can seem like a secret society!

I have a question that I wonder if you or other professors (female or otherwise) might be able to address: I was born and raised in one particular area, and would like to study here, and eventually work here (I work in the field, so it makes sense to me to become very familiar with local systems); however, I have received advice a few times that suggests I will never be hirable here if I do MS,PhD, and post doc here. Thoughts? Thanks!

Doctor Pion said...

My grandfather, an engineer, graphed the winning percentage of pro baseball teams each monday for the entire season and used it to predict trends in the playoffs. For decades.

Lately, I have been recording the morning and evening spot prices of RBOB gasoline (from cnbc) on a calendar to help decide if I should fill up my tank early (price going up) or not worry (price stable or declining).

Arlenna said...

My husband and I have been obsessively tracking our miles per gallon, because starting this fall we'll have commutes of about 800 miles a month. We've found that we have to keep note of which gas station we buy our gas at, because the Marathon by my apartment's gas only gives me ~21 mpg whereas BP (with Invigorate!) gives me like 30 mpg.

and aspiring fsp: it's always good to get out and around and experience some different environments. Different mentorship and peer experiences will make you a better mentor and peer. But a lot of people get away with just getting various types of experiences in the same general locality/city, as long as you branch out some--you need to show yourself to be independent to get hired as such.

Ms.PhD said...

This story is exactly why I adore your blog. Thank you so much for sharing!

Anonymous said...

Great story! I made my boyfriend read it too stating that it sounded like something I would do. Perhaps not take extra data--though I suppose that probably was the "real" window to your soul part.
I recently started work in a scientific organization that requires a physical by their doctors when you start. I guess they're used to their patients demanding to know every little detail because when i went for my physical I was constantly asking "How does this work?" and "Ok, what are you doing now?" "What does that mean?" etc. etc. They all seemed very obliging to give me answers without talking down to me. I was pleased and I got to keep my chest x-ray. way cool! In my experience though, doctors tend to open up a bit when i start asking lots of questions. Even when I went to give blood the first time, the guy was talking to me like i was a kid until I started asking questions, then he started treating me like an adult (I was 18 at the time).

Anonymous said...

Great data discussion - on the other end of this spectrum, my son of 20+ yrs. tells me that the octane is relevant to Arlenna's mileage, with the fuel injection technology in late model cars. So that the 'cost' of the different octanes available is a trade-off with the mileage she will record.
More data to collect??
And also the delight in looking forward to the helpless babies and wild toddlers (believe, I knew of wild toddlers) actually having a thing or two to contribute to your own knowledge when they get an opportunity, I hope.

Chris S. said...

Wow! What an amazing birth story! I've been thinking of collecting data about my wife, who is pregnant. Weight is an easy one that comes to mind, but the size of her abdomen. Now that she's 28 weeks along, it's a little late to start, but I could probably get still get some good data. Have any other ideas what I could collect? I think I would post some results on our blog (not her weight, of course!). Thanks for the inspiration!

Anonymous said...

Lol! I had severe gestational diabetes that was hard to control with two types of insulin and what not. I did the exact same thing.

I also wrote macros in spreadsheet to highlight which readings are over the limit.

I was appalled that my insulin dose was based on just one number (reading, usually fasting one) given that blood glucose change is continuous.