Advances in computer technology have brought about an ever-expanding list of tasks which had previously required the manual ‘human touch,’ and now are programmed to automatically occur. The high profile examples of this phenomenon are self-driving cars, thermostats that adjust themselves based on the time of day, and the amazing interconnectivity of the “Internet of Things.” Computer-guided systems can also dramatically improve the effectiveness of life-or-death medical tasks in a hospital such as blood glucose management.

According to a recent seven-year study of patients who require IV insulin infusion, the use of algorithmically-based computerized systems can optimize the timing of blood glucose checks and dosing, thereby improving patient outcomes. This research, conducted by Vidant Medical Center, a 900-bed, tertiary care teaching hospital located in Greenville, North Carolina, evaluated the impact of EndoTool Glucose Management System and the results were remarkable.

The team of researchers was led by Dr. Robert J. Tanenberg, who in addition to being a professor of medicine specializing in endocrinology at East Carolina University and the Medical Director for Inpatient Diabetes at Vidant Medical Center, has specialized in endocrinology for more than 40 years.

This is his story, in his words:

Collecting and Analyzing Seven Years of Patient Data

When I arrived at Vidant Medical Center we did not have a computerized system for in-patient. I thought it was important for us to have a computerized system and we began using a computerized insulin protocol. We used this program for about five years and it was helpful, but there were some problems with it. For example, it wasn’t portable and it wasn’t wireless. So, we began looking for other options.

About seven years ago, we contracted with Monarch Medical Technologies for EndoTool. We have had tremendous success with this product and it has been a great tool for managing patient blood glucose levels.

Since we had data from seven years, we decided to collate this and evaluate both how quickly we were getting patients to target and preventing hypoglycemia. It’s important to note that we started with the surgical units and, over time, expanded to the medical units.

Basically, we have had really good success and our patients have done very well. What made all of this possible is the advent of electronic health records. This has allowed us to track our progress. Because of this, we published this poster examining hypoglycemia and time-to-target data at the American Diabetes Association’s 76th Scientific Sessions.

Insulin has been around for almost 100 years, and it saves lives. Before insulin, patients with type 1 diabetes would die and type 2 patients did not fare much better. It’s a very important medication, which we use all of the time.

Over the years, there has been a change in the way we manage diabetes in the hospital sparked by the Leuven and NICE-SUGAR studies. We all try to get patients into better glycemic control because the data is very clear – better control results in fewer mortalities.

There is also good data that suggests if we can control a non-diabetic patient’s stress hypoglycemia, we can have an 18-fold reduction in severe events and infection.

As a result of this data, we are using insulin aggressively in the inpatient setting. Insulin is a powerful drug and it must be used correctly. If it is not, the patient can become hypoglycemic. EndoTool is a way for IV insulin to be used safely and effectively.

Using Computerized Glucose Management Software

The glucose management tool we originally used at Vidant was not compatible with modern technology. Aside from the fact that the EndoTool is entirely wireless and therefore more convenient, it has an algorithm which was developed by Dr. Patrick Burgess, who is the founder and inventor of EndoTool.

He came up with the concept of looking at blood sugars and predicting where they are going. This is a fundamental change from any other technology. EndoTool takes into account many factors outside of just blood glucose levels – gender, weight, height, kidney function. All of these factors are entered and, depending on the stability of the blood sugar, the nurse is instructed to check more or less often.

The glucose management systems that use the linear protocols as the foundation, encourage the nurse to increase the insulin drip-rate and, in many cases, this can lead to hypoglycemia.

Results Were Compelling

There are several important findings from this seven-year study. We have data from more than 16,000 patients, with 492,078 blood glucose readings. The data shows that by using the EndoTool we achieved the target much faster than had been previously done – 98% of the patients achieved the target within two hours and the hypoglycemia rates were dramatically lower.

Hypoglycemia rates showed year-over-year decreases with significant reduction in frequency: from 1.04% of blood glucose readings in 2009 to .046% in 2015.

These positive results became even better when the newer version of the EndoTool was introduced last year. After we upgraded in the summer of 2015, we saw mild hypoglycemia rates drop to 0.297% of blood glucose readings.

Since we began using this technology seven years ago, we have expanded its use to include all parts of the hospital, including the ICUs and IUs, where the nursing ratio is 1 to 2 or 1 to 4. Now, for example, every patient that comes out of heart surgery is monitored by the EndoTool until their blood sugar is normal.

The Future of Inpatient Medicine

I think EndoTool is an excellent tool for insulin dosing. It has a non-linear approach and an outstanding algorithm, which gives it predictive capability. I believe this is the future of inpatient medicine.