Inpatient diabetes management: examining morning practice in an acute care setting.

Cohen Linda S, Sedhom Laila, Salifu Moro, Friedman Eli A

The Department of Nursing Services, State University of New York Downstate Medical Center, Brooklyn (Ms Cohen),

College of Nursing Graduate Programs, State University of New York Downstate Medical Center, Brooklyn (Dr Sedhom)

Diabetes Educ. 2007 May-Jun;33(3):483-92. doi: 10.1177/0145721707301351.

PURPOSE

Morning diabetes management in an inpatient acute care facility was examined.

METHODS

A descriptive, nonexperimental research design was used to study the effect of the following variables on patients' prelunch blood glucose: duration of time between (1) blood glucose monitoring and insulin administration, (2) insulin administration and breakfast, and (3) blood glucose monitoring and breakfast. A nonprobability convenience sample was used to examine 40 adults with diabetes who were hospitalized in an urban, academic medical center.

RESULTS

The chi2 test and measurements of central tendency were used for statistical analysis. The mean interval of time between (1) blood glucose monitoring and insulin administration was 93 minutes+/-52.82, (2) blood glucose monitoring and breakfast was 121+/-47 minutes, and (3) insulin administration and breakfast was 73+/-37.06 minutes. Insulin was administered in 28% of patients<45 minutes before breakfast, whereas in 39% and 33%, it was administered between 46 and 90 minutes and >90 minutes prior to breakfast, respectively. There was a statistically significant difference (P=.033) between mean prelunch glucose levels for subjects who received insulin>45 minutes before breakfast and mean glucose levels for those who received insulin<45 minutes before breakfast. Eighty percent of patients whose breakfast was >45 minutes following insulin had prelunch glucose values of >180 mg/dL (10 mmol/L) versus 20%<180 mg/dL (10 mmol/L). Of those whose breakfast was <45 minutes before insulin, 43% had blood glucose levels>180 mg/dL, and 57% had blood glucose<180 mg/dL (P=.026). A logistical regression model revealed that patients had a 5.3 times higher risk of having a prelunch blood glucose level>180 mg/dL when their breakfast time was >45 minutes after receiving insulin as compared to those whose insulin was given<45 minutes before breakfast (relative risk, 5.3; 95% confidence interval, 1.2-25; P=.031).

CONCLUSION

Current practice suggests patients are at a higher risk of prelunch hyperglycemia, and multiple opportunities exist to improve care for the inpatient with diabetes.