Lg Cwd
Need Help

Submit your question to our team of health care professionals.

Current Question

See what's on the mind of the community right now.

Meet the Team

Learn more about our world-renowned team.

DTeam Archives

Review the entire archive according to the date it was posted.

July 22, 2004

Hyperglycemia and DKA

Question from Piittston, Pennsylvania, USA:

My son has been on the pump for about one year now. He spent the night at his friend's the other night and while he was sleeping, his injection site pulled out. He woke up because he did not feel well. He noticed his site had come loose. We picked him up. He had large ketones and was vomiting. After he couldn't even keep down the water I was having him drink, I took him to the Emergency Room (ER). Although we got his sugar down at home from 478 to 115 mg/dl [26.5 to 6.4 mmol/L], he still had large ketones. When we got to the hospital, his sugar was back up to 178 mg/dl [9.8 mmol/L]. They hooked him up to an i.v. Even after four full bags of i.v. fluid, his ketones were still large. They checked his blood levels especially for potassium which they said was good, but he had acidity in his blood and he had 4+ glucose also. First, what does all that mean? Upon getting him there, his blood pressure was also high 148/98 and his heart rate was 107. The hospital contacted my family doctor who had them call an endocrinologist who was in our area. I wanted them to call his pediatric endocrinologist, but they didn't. He was released with large ketones even after all the fluid and they told us if he vomited again at home bring him back to be admitted. How come it took over nine hours in the ER and his ketones still weren't going down? Wasn't it dangerous to release him? He also had a severe pain in his head but they didn't seem to concerned? Was that from the DKA?


This sounds like classical pump DKA from catheter problems/falling out/occlusion. But, the treatment questions you ask raise the most important question which is why you did not have a plan of action for such events since they are basic training for pumpers. Everyone who pumps insulin should be aware that such problems occur and they need to know how to respond. I would strongly suggest that you review this material with your diabetes team so future problems may be prevented.

Also, why did you not call the diabetes team yourself, if the ER doctors did not do so on your behalf?

Without knowing details of these events, it sounds like there was a many hour deficit of insulin and thus the long time to recognize the metabolic emergency and make insulin and fluid corrections made for this episode of DKA. Vomiting is the one event that cannot be treated so easily at home but usually does not happen in DKA until many hours have passed with missed opportunities to deliver insulin via syringe at one to two hour intervals. All DKA treatment should focus on replacing lost fluids and salts, not just on insulin. When high sugars and ketones are present for such a long time, the body falls behind in both salt and water. Thus the dehydration. If acidosis then occurs, this is worsened.

Anytime there are such high values on a pump, assume that the catheter is blocked, kinked or dislodged. Therefore, insulin must be given by syringe and not via the pump at that moment. It is also possible that there is a pump mechanical or electrical malfunction as well, but this is less common than catheter problems. Giving insulin via syringe guarantees that the insulin is in the body. With fast acting analogs, this may need to be repeated at one to two hour intervals until the blood glucose levels start to move downward. High sugars over 250mg/dl [13/9 mmol/L] and moderate to large ketones suggest the need for a large bolus injection in the neighborhood of 20% of the total day’s insulin requirements when not ill. Day to day algorithms and sliding scales are insufficient under such DKA circumstances. With high sugars and only small ketones, then the booster sick day dose is an extra 10% of the usual day’s insulin requirements (basal and bolus added together). This is no different from when using injection therapy or pump therapy. If not responding, headachy, confused, dehydrated or persistent vomiting occurs, then we want a call within a few hours.

Ketones almost always take much longer to clear than glucose levels. This occurs because, during the treatment of DKA when it is successful, the total amount of ketones is responding, but the amount that is easily measurable at home or in most hospital laboratories is only part of this total ketone load. The measurable part rises because the biggest ketone portion, called beta hydroxybutyric acid, isn’t usually being measured directly. Often several hours or even a day or so must pass for all the ketones to clear the system even after insulin is restored and fluid and electrolyte management correct. It’s all very complicated and potentially very dangerous. Unfortunately, many people die during episodes of DKA because of delayed treatment or over-reliance on insulin and under-appreciation of the water and salt imbalances involved. If brain swelling occurs, this is particularly dangerous (called cerebral edema).

Your last question about when someone should be discharged from hospital or emergency room is also impossible to answer since it depends upon the cause of the problem in the first place, how that patient responded to treatment and how well they can then manage at home after the crisis is over.

Bottom line: learn from what happened so that it can be prevented and set up a better system for contacting your diabetes team under such emergency situations.