Computer Algorithm Improved ICU Glucose Control

HOUSTON – Nurses using a computerized algorithm maintained better glycemic control, compared with expert nursing care without the algorithm, in a prospective, randomized study of 300 patients in one institution’s ICU.

Baseline characteristics did not differ between the 151 patients randomized to nursing care and the 149 patients cared for by nurses using the LOGIC-Insulin computerized algorithm developed by Dr. Dieter Mesotten and his associates at Catholic University of Leuven, Belgium. Mean blood glucose levels during ICU care also were similar between groups: 106 mg/dL in both groups.

Sherry Boschert/IMNG Medical Media

On other measures of glycemic control, however, the LOGIC group scored significantly better than did the nurses group, he reported in a poster presentation at the annual meeting of the Endocrine Society. The LOGIC group had a significantly lower mean score on the Glycemic Penalty Index, a marker of efficacy of blood glucose control that was the primary outcome measure of efficacy in the study: 9.8 vs. 12.4 in the nurses group.

Patients in the LOGIC group were less likely to develop critically low glucose levels, they spent more time in the study’s target range of 80-110mg/dL, and they had narrower swings between minimum and maximum blood-sugar measurements, indicating less blood glucose variability.

Mean scores on the Hyperglycemic Index were significantly lower in the LOGIC group (2.5 mg/dL), compared with the nurses group (4.2 mg/dL). Patients in the LOGIC group reached the target range of 80-110 mg/dL faster (in 2 hours on average instead of 3 hours in the nurses group) and were within the target range 69% of the time while in the ICU, compared with 60% of the time for patients in the nurses group. The mean daily difference between minimum and maximum blood glucose levels was 31 mg/dL in the LOGIC group and 37 mg/dL in the nurses group, Dr. Mesotten reported.

The patients had been admitted to the ICU after cardiac surgery (49% in the nurses group and 51% in the LOGIC group), following transplantation (17% and 13%, respectively), because of medical problems (15% and 17%), respectively, or for other reasons.

In general, maintaining tight blood control in critically ill patients is labor intensive and difficult. The LOGIC algorithm may have made this task more successful but also appeared to add a bit to the nursing team’s workload by increasing the frequency of glucose measurements. Nurses in the LOGIC group measured blood glucose levels every 2.2 hours on average, compared with every 2.5 hours in the nursing-only group, a statistically significant difference.

Patients averaged 64 years in age, 60% were male, and 21% had diabetes when admitted to the ICU.

Dr. Mesotten reported having no financial disclosures. Belgian public agencies funded the study.

HOUSTON – Nurses using a computerized algorithm maintained better glycemic control, compared with expert nursing care without the algorithm, in a prospective, randomized study of 300 patients in one institution’s ICU.

Baseline characteristics did not differ between the 151 patients randomized to nursing care and the 149 patients cared for by nurses using the LOGIC-Insulin computerized algorithm developed by Dr. Dieter Mesotten and his associates at Catholic University of Leuven, Belgium. Mean blood glucose levels during ICU care also were similar between groups: 106 mg/dL in both groups.

Sherry Boschert/IMNG Medical Media

On other measures of glycemic control, however, the LOGIC group scored significantly better than did the nurses group, he reported in a poster presentation at the annual meeting of the Endocrine Society. The LOGIC group had a significantly lower mean score on the Glycemic Penalty Index, a marker of efficacy of blood glucose control that was the primary outcome measure of efficacy in the study: 9.8 vs. 12.4 in the nurses group.

Patients in the LOGIC group were less likely to develop critically low glucose levels, they spent more time in the study’s target range of 80-110mg/dL, and they had narrower swings between minimum and maximum blood-sugar measurements, indicating less blood glucose variability.

Mean scores on the Hyperglycemic Index were significantly lower in the LOGIC group (2.5 mg/dL), compared with the nurses group (4.2 mg/dL). Patients in the LOGIC group reached the target range of 80-110 mg/dL faster (in 2 hours on average instead of 3 hours in the nurses group) and were within the target range 69% of the time while in the ICU, compared with 60% of the time for patients in the nurses group. The mean daily difference between minimum and maximum blood glucose levels was 31 mg/dL in the LOGIC group and 37 mg/dL in the nurses group, Dr. Mesotten reported.

The patients had been admitted to the ICU after cardiac surgery (49% in the nurses group and 51% in the LOGIC group), following transplantation (17% and 13%, respectively), because of medical problems (15% and 17%), respectively, or for other reasons.

In general, maintaining tight blood control in critically ill patients is labor intensive and difficult. The LOGIC algorithm may have made this task more successful but also appeared to add a bit to the nursing team’s workload by increasing the frequency of glucose measurements. Nurses in the LOGIC group measured blood glucose levels every 2.2 hours on average, compared with every 2.5 hours in the nursing-only group, a statistically significant difference.

Patients averaged 64 years in age, 60% were male, and 21% had diabetes when admitted to the ICU.

Dr. Mesotten reported having no financial disclosures. Belgian public agencies funded the study.

HOUSTON – Nurses using a computerized algorithm maintained better glycemic control, compared with expert nursing care without the algorithm, in a prospective, randomized study of 300 patients in one institution’s ICU.

Baseline characteristics did not differ between the 151 patients randomized to nursing care and the 149 patients cared for by nurses using the LOGIC-Insulin computerized algorithm developed by Dr. Dieter Mesotten and his associates at Catholic University of Leuven, Belgium. Mean blood glucose levels during ICU care also were similar between groups: 106 mg/dL in both groups.

Sherry Boschert/IMNG Medical Media

On other measures of glycemic control, however, the LOGIC group scored significantly better than did the nurses group, he reported in a poster presentation at the annual meeting of the Endocrine Society. The LOGIC group had a significantly lower mean score on the Glycemic Penalty Index, a marker of efficacy of blood glucose control that was the primary outcome measure of efficacy in the study: 9.8 vs. 12.4 in the nurses group.

Patients in the LOGIC group were less likely to develop critically low glucose levels, they spent more time in the study’s target range of 80-110mg/dL, and they had narrower swings between minimum and maximum blood-sugar measurements, indicating less blood glucose variability.

Mean scores on the Hyperglycemic Index were significantly lower in the LOGIC group (2.5 mg/dL), compared with the nurses group (4.2 mg/dL). Patients in the LOGIC group reached the target range of 80-110 mg/dL faster (in 2 hours on average instead of 3 hours in the nurses group) and were within the target range 69% of the time while in the ICU, compared with 60% of the time for patients in the nurses group. The mean daily difference between minimum and maximum blood glucose levels was 31 mg/dL in the LOGIC group and 37 mg/dL in the nurses group, Dr. Mesotten reported.

The patients had been admitted to the ICU after cardiac surgery (49% in the nurses group and 51% in the LOGIC group), following transplantation (17% and 13%, respectively), because of medical problems (15% and 17%), respectively, or for other reasons.

In general, maintaining tight blood control in critically ill patients is labor intensive and difficult. The LOGIC algorithm may have made this task more successful but also appeared to add a bit to the nursing team’s workload by increasing the frequency of glucose measurements. Nurses in the LOGIC group measured blood glucose levels every 2.2 hours on average, compared with every 2.5 hours in the nursing-only group, a statistically significant difference.

Patients averaged 64 years in age, 60% were male, and 21% had diabetes when admitted to the ICU.

Dr. Mesotten reported having no financial disclosures. Belgian public agencies funded the study.