Week 3: Advancing with Data

Hi everyone,

My data collection is progressing much faster! I learned a few tricks which save me some time. So there are a few different formats the histories are stored in: ones which are photo scanned versions of handwritten documents, and ones which are typed right into the computer. Although there are also various versions of these, the ones which are photo scanned usually have little to no detail about alcohol and tobacco usage. They just have boxes which the physician just checks; however, it doesn't provide information such as how long the patient has been using, and whether or not the patient quit. Sometimes, these boxes are even left blank. And when they are filled out, they sometimes vary from the other histories. On the other hand, the histories which are typed have more information. For example, it would say if the patient quit smoking 20 years ago, instead of just writing "yes." Additionally, it also says if the patient has been using any other street drugs, such as marijuana, cocaine, heroine, etc. I actually created a separate column in my data chart to record if the patient has used any of these other drugs. So, I decided to avoid scrolling through the photo scanned copies unless there is no other option.

While I looked through these histories I noticed things that a lot of the patients had in common. Here is a list of those common illnesses:

Sleeping problems, such as sleep apnea

Breathing problems, such as asthma and dyspnea

Acid reflux/ Gerd

Decreased sensation (Makes sense since neuropathy affects nerve endings.)

Difficulty with concentration

Kidney disease/failure (I guess this is even more evidence that kidney stones and diabetes are linked.)

Congestive heart failure/ cardiovascular disease

Cerebrovascular disease/stroke

Peripheral vascular disease

Hypertension

Blindness/ Vision problems

Skin infections/ Skin breakdown

Dental disease (gingivitis, periodontitis)

Gangrene

Osteomyelitis/ Osteoarthritis

Depressive symptoms

I just thought it might be interesting to connect diabetes to other illnesses!

So the ‘surprising news’ is that I will be shadowing an endocrinologist, Dr. Jyothinagaram, who specializes in diabetes and commonly treats patients with diabetic neuropathy. Starting Monday, I'll be able to see patients with him and see first hand the consequences of diabetic neuropathy. It will also be really interesting to see how the history/physical, the main source of my data, is done.

Furthermore, I will be able to observe a surgery! Although it probably won't be as big as an open-heart surgery, it will still be a significant one so I can actually see inside the human body. There's a good chance it will be an open-abdominal surgery, where I can witness many organs. I don't know when this will be, but I'm very excited!

Week 2: Captivating Events

Hello all,

So the 'Master Spreadsheet' with all the data is on its way to completion. I just have to go through about 200 individual patient records (population size has to be large), scroll through the "history/physical" section for each one, check the lab results for each one, and input the necessary parts for each one into the chart. Not time consuming at all.

In the meantime, there were some interesting things which happened this week. On Wednesday, a couple of urologists gave a presentation on kidney stones: what they are, how they happen, how they are treated, and how to prevent them. They are pretty much stones made up of chemical crystals that separate out from urine. These stones begin to form in the calyx (a cup-shaped part of the kidney). They either stay in the kidney, or move into the urinary tract, where they can block the flow of urine and cause pain.

One of the main reasons kidney stones form is dehydration: if you don't drink enough water, you won't have enough urine to dilute chemicals, forming crystals which develop into stones. In addition, eating foods which contain a lot of protein or salt can lead to kidney stones. There are also kidney infections which can slow down urine flow, or change the acidity in the urine, thereby causing stones. Although the symptoms of kidney stones depend on the stone's size and location, many stones cause sudden/severe pain and bloody urine. Others cause nausea or burning urination. Some say that kidney stones hurt more than gettting shot. There are a few ways to remove the stones, such as using sound waves to break up the stone into smaller pieces so they can be passed through the urinary tract (Shock Wave Lithotripsy), and making an incision on the back so smaller stones can be surgically removed (Percutaneous Nephrolithotomy).

Here's a diagram of where the stones may be.



This talk, although unrelated to my main project, was interesting because the urologists pointed out that obesity was a big factor in many patients who had kidney stones. So I did a little research and found that diabetes and kidney stones are linked. Type 2 diabetes may likely cause urine to be very acidic, which in turn can lead to the development of stones.

On Thursday, my mentor, Sumit, and I attended a meeting about the ongoing diversion in Banner locations across the valley. A diversion happens when all the beds in a hospital are occupied, and incoming ER patients are redirected to a different hospital. This raises the concern that critically ill patients may not be able to get the timely medical care they need. To minimize such occurences, nurses from various departments came together to see which patients can be discharged in order to make more room for new patients. We had to leave the meeting early, before any conclusions, but I'm sure they made some space to fit in more patients.

Oh, and I also had lunch with a bunch of doctors in the Physician's Lounge (Sumit let me in), which was pretty cool.

I think I'll have a few surprises for you next week. Be sure to check in!

Week 1: The Beginning

Hi everyone! Today concluded my first week of my internship at Banner.

The first thing I learned from my mentor, Sumit, was how the entire healthcare system is structured. Because the main theme of my project is healthcare delivery, this helped me understand the process better. There are two main categories: the providers (hospitals, clinics, etc.) and the payers (insurance companies). Patients are seen by providers such as physicians and nurses as either an inpatient or outpatient. For my research, I will be looking into the records of these patients who came to Banner.

Before I get into details of the project, let me summarize how diabetes works. After you eat, the pancreas secretes the hormone called insulin which regulates the blood sugar level; it signals other cells to absorb the sugar, which can then be used for energy for your body. Insulin can also help store this sugar in the liver, and release it when your blood sugar level is too low. Now, when you eat constantly, and your blood sugar level is always high, the receptors on the pancreas start to get desensitized, and fail to properly detect sugar levels. When that happens, insulin isn't used adequately and the sugar in the blood does not get absorbed anymore, leading to high blood sugar levels, and eventually, diabetes. When this happens, it's known as diabetes type 2, which occurs mostly in adults.

There is also something called type 1 diabetes which results because of a congenital disease where the pancreas produces little or no insulin. In my research, I will be focusing on type 2 diabetes.

When doing prior research, I found that there were a few studies done on diabetic neuropathy; however, they were longitudinal studies done for only a couple of years. It's important to note that diabetic neuropathy progresses over many years, so the patients' health could have very well changed after the studies were finished. So, we decided it would be more insightful to do a retrospective study, analyzing patient records beginning from at least 5 years ago. In order to find a list of worthy patient records, we needed an end-point: a below-knee amputation. Patients who had an amputation show that their condition worsened to a point where they had to lose the area which was most affected by diabetic neuropathy. By selecting these patients, we can analyze all the variables which may have led to their amputation.

From the list of patients who received an amputation, we will separate them into two categories: those who were regular smokers or alcoholics, and those who were not. We predict that the average age of those who did not smoke or drink will be much higher than the average age of those who did. We also hypothesize that the HbA1c levels (which tell us the average blood sugar levels of a patient over a period of 3 months) will be higher in patients who were smokers or alcoholics.

The main program I will be using to analyze patient data is called Cerner, where all the records are stored. Firstly, we will get a list of all patients with diabetic neuropathy. Then, we will filter that to get a list of those who received a below-knee amputation. From there, we will separate the patients into smokers/alcoholics and those who were clean, and analyze the variables.

That was the overview of this week. Come back next week to stay updated!