Running the Numbers (Apr/May 2009)

I have entered another two month’s worth of numbers into my stats program and analyzed them. Of note is the fact that I was apparently hyperthyroid in Feb/March (and maybe before that) but went into remission in April. I’m gaining back my weight now, which apparently was caused at least partly by that.

I’ve created a new way of visualizing my data. The next image summarizes major symptoms since symptom onset (Fall ’06) through August ’08. Major events (such as the mold bloom) are marked at the bottom.

Symptoms 11/06 through 8/08
Symptoms 11/06 through 8/08

In this chart, each symptom is represented by a horizontal bar. The bar is either 0, 1, 2, or 3 units thick depending on the severity of the symptom. The horizontal axis shows the month and is annotated with significant events (such as our mold bloom). The vertical line in June indicates the onset of my oral treatment (IV treatment was given in 11-12 of 07). Symptoms shown in orange or red are most problematic; The bottom 9 symptoms have all stopped.

Symptoms 8/08 through 5/09
Symptoms 8/08 through 5/09

In this chart, the thickness of the bars indicates the number of days I had a symptom in a given month. For example, I was dizzy and had swollen lymph nodes almost every day in may. I had no significant memory loss episodes in the same time period. Medication changes are shown at the bottom.

5 thoughts on “Running the Numbers (Apr/May 2009)

  1. How did you make that graph? I am interested in making my own for symptom tracking too.

  2. Jennifer, I just sent another email to you regarding your trip to State College for a conference in June. I would like to meet with you. I’ve just forward this site address to my primary care doctor here in State College. I cannot describe the feeling that I have after reading your blog. At one point, I could not tell the Socail Security psychologist how old I was – I couldn’t remember. I have kept a journal and is mostly handwritten. Today, I am able to ‘chart’ my progress better than even a year ago. If I had not brought a journal with me to the doc each visit, I would never have been able to convey the symptoms and their waxing and waning. In ALL my tests, the one that came back early was for Babesiosis. My doc had to look it up, but he did. The Infectious Disease doc dismissed it and said that I didn’t have enough antibodies for it and “You don’t have it.” I could have written this journal myself. As in all lyme cases, some symptoms may have been at the max for you, but on the lower end of the spectrum for me and vice/versa. Thank you so much for beginning this and updating the blog. Ann

  3. For those wondering how I made these graphs, it’s not very straightforward.

    – They are made in Excel.

    – Create a table with a row for each time period (month in my case) and a column for each symptom. Each symptom has to be on the same scale. For example, in my second chart, I’m tracking the number of days I had a symptom, so the scale is 0-30 (I just assume I will forget to enter at least one day a month). To be more accurate I should actually normalize this based on the number of days I record data but I haven’t bothered with that here.

    – Next to each symptom create an additional column (I suggest naming them “blank”). This column should be automatically calculated to be the inverse of the data you entered. So for example, in my case it’s “30-” where value is the entry in the symptom column.

    – Select your data and create a stacked 100% area graph. In the new version of Excel you just click “charts” “area” and then the third image over. This will show up with the rows and columns reversed, which you can fix by opening the “formatting palette” and then opening “chart data” and clicking the right hand option next to “sort by”

    – The result should be like mine except that all the white stuff in mine will be colored (each “blank” columns). Then you have to right click on each of those in turn, click “format data series” and change the fill to white.

    – Next you go to the legend and delete each of the entries labeled “blank.” Once you’re done with that you may need to change the font size for the legend to get it all to fit and line up.

    If this all sounds complicated, IT IS. Sorry! I am hoping to write software for this once I figure out what’s the best visualization, but that’s probably a year or so off. Also, keep in mind that you’ll probably be recording your data in some other form (on paper or a different spreadsheet) and then have to translate that into days per month or whatever other metric you care about. Right now I record my data on paper and then enter it into SPSS which can aggregate it into days per month using a script I have written. SPSS is an expensive and very high end statistics package that is not very accessible if you don’t have some training with it.

  4. Jen, Thanks for the instructions. My Doc would have appreciated my ‘symptom chart’ in this format. Thanks, Ann

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s