The Goldilocks Zone for Glucose

INFO

In a critically ill patient what blood glucose should we start IV insulin and what is our target range?

This is a topic that has been extensively debated in the literature over the years and whilst i've aimed to summarise some of the key points, trials and artices here, I would urge the reader interested in the detail to read published literature reviews such as this one in the Lancet Diabetes & Endocringology from 2024. For a non peer reviewed summary of where i believe we are at please continue...

My conclusion from the last 30 years of major studies of glucose in critical care can be crystalised into the following 4 principles:

Key Principles

• Avoidance of hypoglycaemia is of paramount importance as this has been associated with worse outcomes in critically ill patients.

• Hyperglycaemia in critically ill patients is very common and may be associated with a pre-existing diagnosis of Diabetes Mellitus or may represent a stress response to critical illness.

• Avoidance of hyperglycaemia is desirable and the treatment of blood glucose above which insulin should commence is 10 mmol/L with a target range of 6.0-10 mmol/L.

• Minimising glucose variability and increasing time in target range are important secondary targets.

Why do we start treatment above 10mmol/l?

"Clinicians should initiate glycemic management protocols and procedures to treat persistent hyperglycemia, ≥ 10 mmol/L (180 mg/dL) in critically ill adults" Society of Critical Care Medicine Guidelines on Glycemic Control for Critically Ill Children and Adults 2024

Most intensive care nurses and doctors understand that we should start treatment with insulin when a patient's blood glucose gets high enough but in my experience there's a lot of variation in when we actually start treating. Many adopt a "watch and wait" strategy when presented with a new admission with a blood glucose of eg 11.4mmol/l.

We know from many large cohort studies that uncontrolled hyperglycaemia is associated with harm. This mostly comes from increased risk of infections (particularly surgical site and pneumonia) as well as acute kidney injury, increased length of stay and mortality.

Early studies such as the often cited Leuwen trial showed that it's possible to reduce these risks by adopting strict control of blood glucose using IV insulin and whist concerns were raised about the risks of using very intensive strategies, it's accepted that avoiding hyperglycaemia is desirable.

When to pull the trigger and start treatment for each individual patient is a very complicated balance of risks and benefits. International consensus guidelines for critically ill adults recommend starting at 10mmol/l. The Centre for Perioperative Care Guidelines for patients undergoing surgery in the UK quote a target range of 6-12mmol/l for patients on insulin whilst the (now quite old) medical inpatient guidelines from JBDS recommend 6-10mmol/l for medical inpatients and advise treatment above 10mmol/l.

The more recent surgical guidelines may reflect the tendency for us to be more wary of over treating hyperglycaemia and causing hypoglycaemia for inpatients on general wards. In ICU we have better resources to provide frequent monitoring so it is considered safe to start treatment at 10mmol/l.

In conclusion, for patients who are critically ill and have a blood glucose above 10mmol/l, starting IV insulin is evidence-based and may reduce complications such as infection and kidney injury. Delaying for many hours to "see if it settles by itself" can often result in requiring more insulin to acheive the target range, particularly if glucose has continued to climb and this is why our guideliens advise starting IV insulin after 2 readings within 2 hours.

What range do we aim for?

On our unit we aim for 6-10mmol/l.

There have been a few studies looking at the optimal range for glucose in critically ill patients. For a concise summary see this article by von Loeffelholz et al. I'll do my best to avoid controversy in summarising them here:

2001 LEUVEN Trial, van den Berghe et al

This was the first major trial looking at intensive glucose control (4.4-6.1mmol/l) in patients in a surgical ICU and showed a reduction in mortality from 8 to 4.6% with reductions in all the morbidity that is commonly associated with hyperglycaemia (infections, AKI etc). This compelling trial led to a widespread worldwide roll out of intensive glucose protocols.

2006 LEUVEN II, van den Berghe et al

This follow up study by the same authors in the medical ICU wasn't able to demonstrate as great a mortality difference (40% to 37.3% which did not meet statistical significance) but did show reduction in acute kidney injury and length of stay. The authors concluded that intensive glucose control reduced morbidity but not mortality in medical patients admitted to the ICU.

2009 NICE-SUGAR, Finfer et al

This large multicentre study concluded that intensive glucose control (4.5-6.0mmol/l) increased mortality in intensive care compared with conventional control (less than 10mmol/l). A post hoc analysis of the trial reported that the increased mortality in the intensive arm of the study may be attributed to hypoglycaemia and that mortality was associated with the number and severity of episodes of hypoglycamia. A further trial looking at intensive vs conventional control was terminated early due to reports of hypoglycaemia.

Following the publication of the NICE-SUGAR trial there was a very slow de-adoption of intensive glucose control in ICU's across the world.

2013 Diabetic status and the relation of the three domains of glycemic control to mortality in critically ill patients: an international multicenter cohort study, Krinsely et al

This cohort study from Northern Europe demonstrated a theme that was representative of a number of publications following NICE-SUGAR which ask the question whether diabetes status pre admission may influence how hyperglycaemia may affect mortality. Whilst they found that hyperglycaemia is associated with increased mortality in persons with and without diabetes, this affect is diminished in diabetics who are poorly controlled. There remains a strong association between hypolgycaemia and mortality regardless of diabetic status. Interestingly they found that glycaemic variability is associated with a worse outcome.

This strongly suggested that there may be a glucose target for each patient that should be individualised to that persons diabetes status and level of control and paved the way for further studies to answer this important question.

2021 CONTROLING, Bohe et al

This was a very elegant study done in French ICU's where the researchers used admission HbA1c to calculate patients average blood glucose pre admission. They then randomised patients to either less than 10mmol/l or to a glucose target range that matched their preadmission average glucose ("individualised control"). The study was stopped early due to an increased risk of hypoglycaemia in the individualised control group and failed to show a statistically significant difference in mortality (67.2% vs 69.6% in the conventional group). Author's conclusions: "Targeting an ICU patient’s pre-admission usual glycaemia using a dynamic sliding-scale insulin protocol did not demonstrate a survival benefit compared to maintaining glycaemia below 180 mg/dL."

2022 LUCID, Bohe et al

This study was eagerly awaited amongst those who care about this topic. Researchers randomised patients with type 2 diabetes to a conentional (<10mmol/l) vs liberal (<14mmol/l) strategy. The primary outcome was hypoglycaemia episodes which were significantly reduced in the liberal group compared to the conventional group (5% vs 18%). Whilst mortality was a secondary outcome, it failed to show benefit in the liberal group compared to conventional (29.5% vs 24.9%, P=0.29) and failed to reach statistical significance. The author's conclusions: "A liberal approach to blood glucose targets reduced incident hypoglycemia but did not improve patient-centered outcomes."

2023 TGC-FAST, Gunst et al

This was another look at intensive glucose control but this time also limiting the use of early parental nutrition. The author's cite the use of a computerised model for controlling blood glucose and randomised to intensive versus liberal (<11.9%). Hypoglycaemia was kept at 1% incidence in the intensive group using a computerised dynamic VRII but the patient centred outcome of mortality was not statistically different (10.1% in liberal versus 10.5% in intensive). Author's conclusion: In critically ill patients who were not receiving early parenteral nutrition, tight glucose control did not affect the length of time that ICU care was needed or mortality

Summary

When to start insulin and what range to aim for has been the subject of a number of good quality trials over the last 20 years and, like with all things in critical care, the answer is likely to be specific to the patient depending on many known and unknown variables. The bottom line is that the most recent trials and studies support starting insulin above 10mmol/l and targeting a conventional range of 6-10mmol/l (or somewhere near this). Targetting a range depending on diabetes status may have benefits but hasn't demonstrated any improvement in patient centred outcomes yet. Use of technology to achieve the goals of minimising hypoglycaemia, controlling hyperglycaemia and minimising glycaemic variability is a sound approach.

Tight versus liberal blood-glucose control in the intensive care unit: special considerations for patients with diabetes von Loeffelholz, Christian et al. The Lancet Diabetes & Endocrinology, Volume 12, Issue 4, 277 - 284

Krinsley JS, Egi M, Kiss A, Devendra AN, Schuetz P, Maurer PM, Schultz MJ, van Hooijdonk RT, Kiyoshi M, Mackenzie IM, Annane D, Stow P, Nasraway SA, Holewinski S, Holzinger U, Preiser JC, Vincent JL, Bellomo R. Diabetic status and the relation of the three domains of glycemic control to mortality in critically ill patients: an international multicenter cohort study. Crit Care. 2013 Mar 1;17(2):R37. doi: 10.1186/cc12547. PMID: 23452622; PMCID: PMC3733432.

Bohé J, Abidi H, Brunot V, Klich A, Klouche K, Sedillot N, Tchenio X, Quenot JP, Roudaut JB, Mottard N, Thiollière F, Dellamonica J, Wallet F, Souweine B, Lautrette A, Preiser JC, Timsit JF, Vacheron CH, Ait Hssain A, Maucort-Boulch D; CONTROLe INdividualisé de la Glycémie (CONTROLING) Study Group. Individualised versus conventional glucose control in critically-ill patients: the CONTROLING study-a randomized clinical trial. Intensive Care Med. 2021 Nov;47(11):1271-1283. doi: 10.1007/s00134-021-06526-8. Epub 2021 Sep 29. PMID: 34590159; PMCID: PMC8550173.