Lower glucose target linked to improved mortality in critically ill

 

In critically ill patients, treating blood glucose with a low target of 80-110 mg/dL was associated with a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL, according to results of a retrospective cohort analysis.

With the computerized intravenous insulin protocol used in the study, the strict target could be achieved with a low rate of hypoglycemia, the authors wrote. The analysis was published in the journal CHEST^®.

monkeybusinessimages/Thinkstock

These findings do not suggest that clinicians should practice counter to current guidelines, which recommend against intensive insulin therapy, noted Andrew M. Hersh, MD, of the division of pulmonary and critical care at San Antonio Military Medical Center, and his coauthors.

However, it does raise the possibility that earlier investigations finding an association between intensive insulin therapy and excess mortality “may have been accurate only in the setting of technologies which led to high rates of severe hypoglycemia,” they wrote.

The retrospective cohort analysis by Dr. Hersh and his colleagues included 1,809 adult patients treated at three different ICUs in two hospitals between January 2010 and December 2015. Treatment was delivered with a computerized ICU insulin infusion protocol that allows clinicians to choose between two blood glucose targets: 80-110 mg/dL or 90-140 mg/dL. The lower target was chosen for 951 patients, and the moderate target for 858 patients.

The most common primary admission diagnoses in the cohort included chest pain or acute coronary syndrome in 43.3%, cardiothoracic surgery in 31.9%, heart failure (including cardiogenic shock) in 6.8%, and vascular surgery in 6.0%.

While patients in the low blood glucose target group had a higher rate of moderate hypoglycemia, both groups had a low rate of severe hypoglycemia, at 1.16% in the low target group and 0.35% in the moderate target group (P = .051).

 

Unadjusted 30-day mortality was significantly lower in the 80-110–mg/dL group compared with the 90-140–mg/dL group (4.3% vs. 9.2%, respectively; P less than .001), according to the investigators.

Furthermore, logistic regression analysis showed that patients treated with a target of 80-110 mg/dL had a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL (odds ratio 0.65; 95% confidence interval, 0.43-0.98; P = .04).

These results advance the debate over appropriate blood glucose targets in critically ill patients, as they suggest that the effects of targeting blood glucose and the effects of severe hypoglycemia “can be separated,” the investigators wrote.

Current guidelines on intensive insulin therapy are based in part on findings of the NICE-SUGAR trial, which found that among adults treated in the ICU, intensive glucose control increased mortality. However, a post hoc analysis suggested the mortality increase in NICE-SUGAR was “largely driven by a significant incidence of moderate hypoglycemia, and to a greater degree severe hypoglycemia,” Dr. Hersh and his coauthors noted in their report.

 

“Given improvements in insulin delivery and glucose monitoring, a reassessment of potential benefits of [intensive insulin therapy] should once again be evaluated in a prospective randomized trial,” they wrote.

Dr. Hersh and his coauthors declared no financial or nonfinancial disclosures related to the study.

SOURCE: Hersh AM et al. CHEST 2018. doi: 10.1016/j.chest.2018.04.025.

 

In critically ill patients, treating blood glucose with a low target of 80-110 mg/dL was associated with a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL, according to results of a retrospective cohort analysis.

With the computerized intravenous insulin protocol used in the study, the strict target could be achieved with a low rate of hypoglycemia, the authors wrote. The analysis was published in the journal CHEST^®.

monkeybusinessimages/Thinkstock

These findings do not suggest that clinicians should practice counter to current guidelines, which recommend against intensive insulin therapy, noted Andrew M. Hersh, MD, of the division of pulmonary and critical care at San Antonio Military Medical Center, and his coauthors.

However, it does raise the possibility that earlier investigations finding an association between intensive insulin therapy and excess mortality “may have been accurate only in the setting of technologies which led to high rates of severe hypoglycemia,” they wrote.

The retrospective cohort analysis by Dr. Hersh and his colleagues included 1,809 adult patients treated at three different ICUs in two hospitals between January 2010 and December 2015. Treatment was delivered with a computerized ICU insulin infusion protocol that allows clinicians to choose between two blood glucose targets: 80-110 mg/dL or 90-140 mg/dL. The lower target was chosen for 951 patients, and the moderate target for 858 patients.

The most common primary admission diagnoses in the cohort included chest pain or acute coronary syndrome in 43.3%, cardiothoracic surgery in 31.9%, heart failure (including cardiogenic shock) in 6.8%, and vascular surgery in 6.0%.

While patients in the low blood glucose target group had a higher rate of moderate hypoglycemia, both groups had a low rate of severe hypoglycemia, at 1.16% in the low target group and 0.35% in the moderate target group (P = .051).

 

Unadjusted 30-day mortality was significantly lower in the 80-110–mg/dL group compared with the 90-140–mg/dL group (4.3% vs. 9.2%, respectively; P less than .001), according to the investigators.

Furthermore, logistic regression analysis showed that patients treated with a target of 80-110 mg/dL had a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL (odds ratio 0.65; 95% confidence interval, 0.43-0.98; P = .04).

These results advance the debate over appropriate blood glucose targets in critically ill patients, as they suggest that the effects of targeting blood glucose and the effects of severe hypoglycemia “can be separated,” the investigators wrote.

Current guidelines on intensive insulin therapy are based in part on findings of the NICE-SUGAR trial, which found that among adults treated in the ICU, intensive glucose control increased mortality. However, a post hoc analysis suggested the mortality increase in NICE-SUGAR was “largely driven by a significant incidence of moderate hypoglycemia, and to a greater degree severe hypoglycemia,” Dr. Hersh and his coauthors noted in their report.

 

“Given improvements in insulin delivery and glucose monitoring, a reassessment of potential benefits of [intensive insulin therapy] should once again be evaluated in a prospective randomized trial,” they wrote.

Dr. Hersh and his coauthors declared no financial or nonfinancial disclosures related to the study.

SOURCE: Hersh AM et al. CHEST 2018. doi: 10.1016/j.chest.2018.04.025.

 

In critically ill patients, treating blood glucose with a low target of 80-110 mg/dL was associated with a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL, according to results of a retrospective cohort analysis.

With the computerized intravenous insulin protocol used in the study, the strict target could be achieved with a low rate of hypoglycemia, the authors wrote. The analysis was published in the journal CHEST^®.

monkeybusinessimages/Thinkstock

These findings do not suggest that clinicians should practice counter to current guidelines, which recommend against intensive insulin therapy, noted Andrew M. Hersh, MD, of the division of pulmonary and critical care at San Antonio Military Medical Center, and his coauthors.

However, it does raise the possibility that earlier investigations finding an association between intensive insulin therapy and excess mortality “may have been accurate only in the setting of technologies which led to high rates of severe hypoglycemia,” they wrote.

The retrospective cohort analysis by Dr. Hersh and his colleagues included 1,809 adult patients treated at three different ICUs in two hospitals between January 2010 and December 2015. Treatment was delivered with a computerized ICU insulin infusion protocol that allows clinicians to choose between two blood glucose targets: 80-110 mg/dL or 90-140 mg/dL. The lower target was chosen for 951 patients, and the moderate target for 858 patients.

The most common primary admission diagnoses in the cohort included chest pain or acute coronary syndrome in 43.3%, cardiothoracic surgery in 31.9%, heart failure (including cardiogenic shock) in 6.8%, and vascular surgery in 6.0%.

While patients in the low blood glucose target group had a higher rate of moderate hypoglycemia, both groups had a low rate of severe hypoglycemia, at 1.16% in the low target group and 0.35% in the moderate target group (P = .051).

 

Unadjusted 30-day mortality was significantly lower in the 80-110–mg/dL group compared with the 90-140–mg/dL group (4.3% vs. 9.2%, respectively; P less than .001), according to the investigators.

Furthermore, logistic regression analysis showed that patients treated with a target of 80-110 mg/dL had a lower risk of 30-day mortality compared with patients with a target of 90-140 mg/dL (odds ratio 0.65; 95% confidence interval, 0.43-0.98; P = .04).

These results advance the debate over appropriate blood glucose targets in critically ill patients, as they suggest that the effects of targeting blood glucose and the effects of severe hypoglycemia “can be separated,” the investigators wrote.

Current guidelines on intensive insulin therapy are based in part on findings of the NICE-SUGAR trial, which found that among adults treated in the ICU, intensive glucose control increased mortality. However, a post hoc analysis suggested the mortality increase in NICE-SUGAR was “largely driven by a significant incidence of moderate hypoglycemia, and to a greater degree severe hypoglycemia,” Dr. Hersh and his coauthors noted in their report.

 

“Given improvements in insulin delivery and glucose monitoring, a reassessment of potential benefits of [intensive insulin therapy] should once again be evaluated in a prospective randomized trial,” they wrote.

Dr. Hersh and his coauthors declared no financial or nonfinancial disclosures related to the study.

SOURCE: Hersh AM et al. CHEST 2018. doi: 10.1016/j.chest.2018.04.025.