Improving Inpatient Diabetes Management in a Community Hospital: A Resident-Run Quality Improvement Initiative

Objective:
Diabetes mellitus is one of the most encountered co-morbidities in outpatient and inpatient settings. Treatment varies from oral hypoglycemic agents (OHAs) to injectable agents to insulin based on average glucose levels. Guidelines recommend discontinuation of OHAs during inpatient hospitalization and using long and short-acting insulin as treatment, to avoid unpredictable glucose levels. In our community hospital, it was noted that if home OHAs are discontinued on admission, physicians tend to use only the insulin sliding scale (ISS) or short-acting pre-prandial and bedtime insulin as treatment. Recent studies have shown that stand-alone ISS is inferior to a basal/bolus insulin regimen in achieving adequate blood glucose control in hospitalized patients. Our aim is to improve blood glucose control in our community hospital by implementing an insulin policy as a standard treatment for diabetes management.

Methods:
We used the Institute Healthcare Improvement (IHI) model to guide our quality improvement (QI) project. Root cause analysis through a fishbone revealed contributing factors. A multidisciplinary team, which included IT, pharmacy, and resident physicians, was created to implement the proposed solutions. Several PDSA cycles were used to test changes.

Results:
Despite guidelines to assist in the management of inpatient diabetes, our hospital lacked a standard approach. PDSA 1 involved education about guidelines, discontinuation of OHAs, and utilizing insulin for management. Post-implementation, >50% of physicians still ordered OHAs. PDSA 2 cycle, a pharmacy checkpoint was placed to discontinue OHAs, and physicians were prompted to use insulin. This led to the elimination of OHAs and the initiation of insulin. However, stand-alone ISS was being used, causing hyperglycemia. PDSA 3 re-education was provided. Policy changes were implemented to incorporate a basal insulin regimen along with ISS for inpatient management. We proposed 10 units of long-acting insulin plus ISS. The new policy was implemented with the aid of the pharmacy. A review of post-intervention data from 250 patients through a 6-month period showed that OHAs were continued in only 1.2% of the patients. Stand-alone ISS therapy was started in 24% of the patients, whereas 62% of patients received a combination of basal insulin along ISS. We noted hypoglycemic events (glucose < 60 mg/dl) in 7.6% of the patients, and hyperglycemia (glucose >160 mg/dl) in 60% of the patients.

Discussion/Conclusion:
Limiting the use of OHAs in the hospital resulted in fewer hypoglycemic episodes. Prior to our QI, we noted an inconsistent approach to diabetes management leading to hypo and severe hyperglycemic events. As a part of this QI initiative, a new hospital-wide policy, which included basal and short-acting insulin, was incorporated with the help of a multidisciplinary team. We noticed that incorporating a standardized approach led to an improved trend in glucose control and fewer complications secondary to severe hyperglycemia or hypoglycemia.

The next step is to implement an EMR conversion system from OHAs to basal/bolus insulin combination and maintain sustainability.