Targeted glycemic management and glycemic control in hospitalized patients with severe hyperglycemia

Program: Abstracts - Orals, Featured Poster Presentations, and Posters
Session: SUN 839-872-Diabetes & Obesity Management
Clinical
Sunday, June 16, 2013: 1:45 PM-3:45 PM
Expo Halls ABC (Moscone Center)

Poster Board SUN-867
Cynthia F Yazbeck*1, Amy Calebrese Donihi2, Kristin Delisi1, Emily Martin3, Nisha Parambil4 and Mary T Korytkowski1
1University of Pittsburgh Medical Center, Pittsburgh, PA, 2University of Pittsburgh School of Pharmacy, Pittsburgh, PA, 3VA Pittsburgh Healthcare System, PA, 4University of Maryland at Baltimore Washington Medical Center
Hyperglycemia occurs frequently in the inpatient setting and is associated with increased morbidity and mortality.  Despite this, there is a persistence of clinical inertia for directing interventions to reduce the frequency and severity of hyperglycemia events in the hospital.  To address this, we implemented a quality improvement project that was directed at notifying the primary service for patients who had ≥ 2 blood glucoses (BG) > 300 mg/dL at least 4 hours apart during a 36 hour period.   The objective of this project was to determine whether direct notification of the service resulted in consultation with the Inpatient Diabetes Service (IDS) or improvement in glycemic control. 

A daily report identifying all non-critically ill patients with ≥ 2 BG > 300mg/dL between December 2011 and May 2012 was reviewed. The primary team was alerted to the occurrence of severe hyperglycemia with the recommendation that an IDS was available to assist with glycemic management.   Patients for whom consultation was obtained (Group 1, n = 20) were compared with those for whom consultation was declined (Group 2, n = 16).  Glycemic data from patients for whom the primary service was not called were used as controls (Group 3, n = 22).  Patient demographics included age (Group 1 vs. 2 vs. 3: 54 ± 17 vs. 59 ± 15 vs. 64 ± 17 years, p = 0.17); weight (71 ± 14 vs. 97 ± 37 vs. 90 ± 23 kg, p = 0.009); A1C (7.9 ± 1.9 vs. 7.6 ± 1.3 vs. 8.9 ± 3.1%, p = 0.37);  hematocrit (28.5 ± 4.5 vs. 31.6 ± 3.5 vs. 33.1 ± 9.1%, p = 0.12) and gender (%female: 45 vs. 56 vs. 23%).

Mean BG for the 2 days prior to contact (303 ± 50 vs. 319 ± 38 vs. 266 ± 58 mg/dL, p = 0.008) and on the day of contact (296 ± 66 vs. 298 ± 57 vs. 238 ± 57, p = 0.006) was significantly lower in group 3.  There were no group differences in mean BG on days 2-4 following the contact (199 ± 47 vs. 201 ± 51 vs. 218 ± 45, p = 0.35); however, the percent reduction in BG was significantly greater in Groups 1 and 2 (30 ± 22 vs. 27 ± 16 vs. 3 ± 30, p = 0.002).  The percentage of BG in goal range (70-180 mg/dL) increased from 13 ± 18 to 46 ± 21% in Group 1 (p < 0.001), 11 ± 15 to 41 ± 22% in Group 2 (p < 0.001), and 14 ± 21 vs. 30 ± 20% in Group 3 (p = 0.013).  The percentage of BG > 300 decreased from 55 ± 19 to 14 ± 15% in Group 1 (p < 0.001), from 63 ± 22 to 12 ± 16% in Group 2 (p < 0.001), and from 37 ± 29 to 15 ± 13% in Group 3 (p = 0.002).  

In summary, we observed that a phone call alerting the primary team to the presence of severe hyperglycemia triggers a therapeutic response that results in reductions in mean BG and the % of BG > 300 mg/dL and an increase in the % of BG within goal range. These changes were similar to those observed in consultation with the IDS.  This suggests that providing alerts related to hyperglycemia can be useful in overcoming clinical inertia.  Despite this intervention, mean BG levels remained above goal in all groups, highlighting the challenges of achieving recommended inpatient glycemic goals.

Disclosure: MTK: Principal Investigator, Sanofi, Consultant, Regeneron, Speaker, American Association of Clinical Endocrinologists. Nothing to Disclose: CFY, ACD, KD, EM, NP

*Please take note of The Endocrine Society's News Embargo Policy at http://www.endo-society.org/endo2013/media.cfm