Diabetic ketoacidosis (DKA) is the leading cause of morbidity and mortality in children with type I diabetes mellitus (DM). The frequency of DKA as the initial presentation for type I DM is approximately 30 percent in North America. Through our quality improvement project, we sought to create and implement interventions for reducing WVU Medicine Children’s DKA readmission rate in pediatric patients with type I DM. When starting this project, our one-year readmission rate for DKA was 23%, which is 15% above the national average. We hypothesized that development and utilization of standard interventions would reduce our DKA readmission rate. Patients who had DKA listed as their admission diagnosis were reviewed and those who had a prior admission for DKA within the last 12 months were enrolled in the study after consent was obtained. Intervention was a questionnaire completed by medical providers, primarily resident physicians, following a standardized interview of participants to discover and address causes leading to DKA admissions such as infections, dehydration, non-compliance, insulin pump failure, inadequate medical supplies, etc. Data was included for all patients with Type I DM admitted to our pediatric ICU with an admission diagnosis of DKA. We used the six-month time period from March 2017 to August 2017, prior to initiation of our intervention, for our baseline comparison. This period included 57 total patients, with 11 readmissions, yielding a six-month readmission rate of 19.3%. The 11 readmissions included seven unique patients, with some patients accounting for more than one readmission. For the six months from September 2017 to February 2018, during the post-intervention period, the number of patients admitted with DKA was 53, with seven patients readmitted with DKA during the same time period yielding a six-month readmission rate of 13.2%. We have achieved a 32% reduction in the readmission rate for DKA at our hospital since the start of the intervention compared to the six months prior to intervention. Using a linear trend extrapolation method, our improvement is expected to yield a 12-month DKA readmission rate that is well below the national average of 20%. This is a significant achievement, as this is helping our patients to avoid this potentially life-threatening complication of diabetes and hopefully helping them to achieve better control of their disease overall. This could also help with the huge financial burden that these readmissions place upon our patients and healthcare system, as virtually all of these patients presenting with DKA will be initially admitted to our pediatric ICU and will incur on average $13,946 in admission-associated billing.


This graph shows the projected readmission rates through 1 year of interventions as well as the 6 month data that has been collected.