Intensive insulin therapy in sepsis patients: Better data enables better intervention

2.1. Benefit of intensive insulin therapy

Insulin lowers blood sugar, on the one hand, by promoting more glucose to enter hepatocytes, muscle cells, adipocytes and other tissue cells to synthesize glycogen. On the other hand, it can inhibit fat decomposition and normalize dyslipidemia. What's more, insulin has immunoregulation effects such as inhibiting the overexpression of inflammatory factors and improving the macrophage function [29]. Strictly maintaining glucose levels between 4.4 and 6.1 mmol/L through insulin infusion can significantly reduce mortality and morbidity in critically ill patients [30]. For long-term hospitalized patients who need intensive care, intensive insulin therapy can not only exhibit a lower mortality rate, but also significantly reduce the incidence of sepsis, renal failure and nervous system diseases when compared to conventional treatment [31].

Some clinical trials have shown that intensive insulin therapy is beneficial for patients with the diagnosis of diabetes in terms of preventing complications associated with diabetes mellitus [32]. In a cohort study, the authors observed a significant decrease in the proportion of patients with retinopathy deterioration and increased urinary albumin excretion in the intensive treatment group, as well as an 86% reduction in the risk of new albuminuria in that group [33]. Two other randomized controlled trials have also demonstrated that intensive blood glucose control can reduce the prevalence of neuropathy [34] and the frequency or severity of diabetic microvascular complications (retinopathy and nephropathy) [35]. What is more noteworthy is that the most landmark studies supporting the idea that intensive care patients receiving intensive insulin therapy can reduce morbidity and mortality are two large single-center prospective randomized controlled trials. One of these trials evaluated the effects in surgical patients, which found that the intensive treatment group had a lower hospital mortality rate than the conventional treatment group (37.3% vs 40.0%) [23]. Another study concluded that intensive insulin therapy reduced the mortality of critically ill patients from 8.0% of conventional treatment to 4.6% (P < 0.04) [6]. This reduction is mainly due to a decrease in deaths from sepsis-induced multisystem organ failure. The authors also reported that intensive insulin therapy reduced overall hospital mortality by 34%, decreased bloodstream infections by 46% and acute renal failure requiring dialysis or hemofiltration by 41% [6].

All of the research cited above showed the effectiveness of intensive insulin therapy. Meanwhile, those studies have also mentioned the risk of intensive insulin therapy but have tended to suggest that the clinical benefit of intensive insulin therapy outweighs the risk.

2.2. Risk of intensive insulin therapy

Even though most studies believed that intensive insulin therapy has a significant effect on blood glucose control in patients with sepsis, it does not mean that there is no risk. On the contrary, some clinical research has suggested that intensive insulin therapy has limited clinical benefit in critically ill patients and increases the risk of hypoglycemia.

A meta-analysis of 8432 patients from 29 randomized controlled trials conducted by Wiener et al. [36] reported that intensive insulin therapy was associated with an increased risk of hypoglycemia (13.7% vs 2.5%; RR, 5.13; 95% confidence interval [CI], 4.09–6.43). Additionally, it revealed that there was no difference in overall hospital mortality between intensive insulin therapy and conventional therapy (21.6% vs 23.3%; RR, 0.93; 95% CI, 0.85–1.03), nor did it significantly reduce the risk of new need for dialysis (11.2% vs 12.1%; RR, 0.96; 95% CI, 0.76–1.20). Two other randomized controlled trials described results that intensive treatment group had a higher incidence of severe hypoglycemia (6.8% vs 0.5%, P < 0.001; 17.0% vs 4.1%, P < 0.001) [24,27], which is comparable to the findings of the aforementioned study. However, the difference is that the mortality rate in the intensive treatment group is higher than that in the conventional treatment group (27.5% vs 24.9%, P = 0.02) [27]. The authors also reported that the median number of days in the hospital (P = 0.86) or ICU (P = 0.84) or the median number of days of renal-replacement therapy (P = 0.39) or mechanical ventilation (P = 0.56) between the two treatment groups had no significant difference [27]. Moreover, the study by Brunkhorst et al. showed there was no significant difference in mortality or the mean score for organ failure between the two groups but did note the incidence of serious adverse events in the intensive treatment group was higher than that in the conventional treatment group (10.9% vs 5.2%, P = 0.01) [24].

Since the prevalence of intensive insulin therapy in ICUs is increasing [37], the “Surviving Sepsis Campaign Guidelines” recommended that we should be on guard against a series of problems during intensive insulin therapy. Especially the signs and symptoms of hypoglycemia in a critically ill patient can be misconstrued as another diagnosis, leading to a delay in recognizing the hypoglycemia [[38], [39], [40]]. A large multicenter analysis of evaluating the trend of early BG (that is, within 24 h of ICU admission) and the relationship between blood glucose control and hospital mortality found that both hypoglycemia and hyperglycemia were harmful to critically ill patients [41]. Although both high and low blood sugar levels increase the risk of death [12], hypoglycemia may sometimes be more dangerous. For example, it may be associated with increased severity and higher mortality (<3.89 mmol/L, 38.2% vs 3.89–7.72 mmol/L, 19.5%) in patients with severe sepsis [42]. Therefore, it is necessary to analyze the risk factors for hypoglycemia in sepsis patients receiving intensive insulin therapy, so as to determine a blood glucose range for each ICU that does not cause a significant increase in severe hypoglycemia and is in line with the limitations of nursing and economic resources [43].

To begin with, when eating irregularly or delaying meals, the insulin intensive treatment group is more likely to develop hypoglycemia than the conventional treatment group. Secondly, different patients with different diseases, treatment measures and other factors may result in varying curative effects. Sepsis, hemodialysis, and organ failure have been reported to be risk factors for developing hypoglycemia with intensive insulin therapy [44]. Berghe et al. [45] summed up the data from two previously reported randomized controlled trials [6,23] and discovered that most patients, including those with sepsis at the time of admission, benefited from intensive insulin therapy, but there was no survival benefit for patients with the diagnosis of diabetes [46]. Furthermore, a study comparing the clinical characteristics of patients with and without the diagnosis of diabetes found that up to 84.4% of those with the diagnosis of diabetes died from sepsis [47]. Sathananthan et al. [48] conducted a retrospective study of 1698 patients admitted to ICUs with sepsis secondary to bacterial pneumonia from 2001 to 2012, and also found an increase in mortality in patients with the diagnosis of type 2 diabetes who had lower admission blood glucose.

In addition, rapid normalization of blood sugar levels can be harmful for patients with the diagnosis of diabetes [45]. The mechanism of insulin injury may be related to the anabolic effect that is similar to growth hormone [[49], [50], [51]]. Patients with the diagnosis of diabetes can adapt to chronic hyperglycemia by reducing the expression of glucose transporters (GLUT) in some cell types [52] in order to achieve rapid metabolic control and aggravate complications [53], which may also explain the increased risk of death during insulin therapy in septic patients with the diagnosis of diabetes.

We summarize the literature information involved in analyzing the benefit and risk of intensive insulin therapy, such as type, quality of evidence and number of cases, as shown in Table 1.

Despite the clinical controversy surrounding intensive insulin therapy, we believe that appropriate insulin therapy is crucial for septic patients with elevated blood sugar. Firstly, long-term hyperglycemia will cause pathological changes in various organs of the body, resulting in a series of chronic complications. Regardless of the hypoglycemia risks caused by intensive insulin therapy, the strictest blood sugar control does benefit the most. Secondly, the risk of hypoglycemia is not inevitable. While there is no expert consensus on a standard procedure for risk avoidance of intensive insulin scheme, severe hypoglycemia can be avoided by adjusting the monitoring frequency, administration method, infusion speed and insulin dosage based on the monitoring results of blood glucose level [54].

3.1. Indication and contraindication of intensive insulin scheme

Through the aforementioned analysis of the clinical risk factors of intensive insulin therapy, we can know that not all patients can benefit from intensive insulin therapy. The individual differences may to some extent determine whether patients should receive the treatment, which requires consideration of the indications and contraindications of the treatment.

The most common indication of insulin therapy is the inability to control blood sugar levels through diet, exercise and oral drugs [55]. As all type 1 diabetics require subcutaneous or intravenous insulin therapy whether in the outpatient or inpatient setting, especially in the inpatient setting. Here we mainly discuss indications for intensive insulin therapy in patients with a diagnosis of type 2 diabetes, which include acute illness or surgery, glucose toxicity, contraindications (such as renal or liver failure) to or failure to achieve goals with oral antidiabetic medications, the need for flexible treatment [55], patients with cardiovascular lesions or retinal macular edema, failure to achieve satisfactory blood glucose control on multiple daily injections (MDI) [56], and extreme insulin resistance syndromes such as pregnancy, insulin allergy, and lipodystrophy syndromes [57]. Other indications for intensive insulin therapy include: ketoacidosis, acute infections, severely decompensated patients with clinical symptoms, and dehydration [58]. In addition, we know that a particular concern with septic patients with sepsis in ICUs with regards to subcutaneously administered medications is the risk of malabsorption, particularly for patients receiving vasopressors for hemodynamic support in their shock state. Ideally, patients who are in multi-pressor shock would be receiving an intravenous insulin infusion to avoid the issues associated with malabsorption of subcutaneous medications. And those patients requiring multiple vasopressors or high infusion rates of a single vasopressors would likely still benefit from intravenous insulin infusion rather than the subcutaneous infusion protocol.

Intensive insulin therapy is a common and effective therapeutic method, but there are also situations where the treatment is ineffective and/or dangerous to the patient. Consequently, it is necessary to understand the absolute and relative contraindications of intensive insulin therapy. Absolute contraindications [59] include: severe psychiatric illness; severe, rapidly progressive or proliferative retinopathy; regular exposure to strong magnetic fields. Relative contraindications [59] include: suboptimal adherence to diabetes treatment; suboptimal acceptance of treatment by the patient; poor hygiene and participation in violent sports; sensory (particularly visual) or gestural impairment (physically handicapped); end-stage renal failure and the risk of acidosis; living in extremely cold or heated environments for professional or personal reasons; patients taking underwater diving as a sport or profession; participation in extreme sports.

3.2. Suggestion of intensive insulin regimen in patients with sepsis

In the case of septic patients, while intensive insulin therapy improves clinical outcomes, there is indeed a risk of hypoglycemia. Although hypoglycemia is not an extremely difficult complication to treat, if it is not detected in time and appropriate measures are not taken, it will also bring serious consequences and even lead to the death of patients. Therefore, it is an essential step to monitor the blood glucose levels of septic patients treated with intensive insulin therapy.

The first “Surviving Sepsis Campaign Guidelines” published in 2004 recommended that blood glucose be monitored frequently (every 30–60 min) after the project start and on a regular basis (every 4 h) once blood sugar levels stabilized [37]. The three subsequent published “Surviving Sepsis Campaign Guidelines” varied slightly in terms of recommendations regarding blood glucose monitoring frequency. They all recommended monitoring blood sugar every 1–2 h until blood glucose levels and the rate of insulin infusion are stable, and then every 4 h [[60], [61], [62]]. However, the mode of administration, the speed of administration and the frequency of monitoring for different blood glucose levels are not fully and specifically reflected in these guidelines. Currently, with reference to these guidelines and clinical practice, our hospital has proposed the following intensive insulin therapy process implemented in ICUs, and it has been applied in clinic. We made the flow chart, as shown in Fig. 2.

The target blood glucose level of this process is 6.1–10.0 mmol/L. We divide the patients into three different situations according to their blood glucose levels and take the appropriate measures. (1) Blood glucose (BG) < 4.0 mmol/L: 25 g glucose was injected intravenously and blood glucose levels were retested 30 min later. If the blood glucose level does not reach the target, the above administration method should be continued. (2) 10.0 < BG < 11.1 mmol/L: 4U INS was injected hypodermically and blood glucose was retested after 2 h. If the blood glucose level does not reach the standard, the above administration method should be continued. (3) BG ≥ 11.1 mmol/L: ①11.1 ≤ BG < 14.1 mmol/L: Hypodermic injection of 6U INS or pump insulin at a rate of 4U/h; ②14.1 = BG < 17.1 mmol/L: Hypodermic injection of 8U INS or pump insulin at a rate of 6U/h; ③17.1 = BG ≤ 22.0 mmol/L: Hypodermic injection of 10U INS or pump insulin at a rate of 8U/h; ④BG > 22.0 mmol/L: Pump insulin at a rate of 10U/h. In the third case, checking the blood glucose once every hour after the correction with pump. If the blood glucose level is not up to the standard and increases or decreases by 1 mmol/L, the pump speed is up or down 0.5U. And the blood glucose level was monitored again an hour later for subsequent adjustment of drug administration.

3.3. Matter needing attention and monitoring principle in the implementation of intensive insulin therapy in patients with sepsis

Sepsis is a clinically critical disease that needs special attention. From the above risk analysis, it is known that the risk factors of hypoglycemia in septic patients receiving intensive insulin therapy include: (1) Irregular diet or delayed meals; (2) Organ failure; (3) Patients who need hemodialysis; (4) Patients with the diagnosis of diabetes; (5) Compliance with the prescribed scheme, frequency of blood glucose measurement, use of care point equipment and continuous blood glucose monitoring [63]. The above situation should be closely monitored and evaluated to avoid serious adverse outcomes.

Most critically ill patients have disturbance of consciousness or treatment with mechanical ventilation and sedative drugs, which would make the symptoms be masked and go unnoticed. As a consequence, we should follow the principle of early detection and early treatment to take appropriate monitoring measures for septic patients treated with intensive insulin therapy. Fast blood glucose meter is often used to monitor blood glucose in clinic, and this method is the most common and convenient. Berghe et al. [6] suggested that newly admitted patients be measured once every 1–2 h until their blood glucose levels stabilize at 4.4–6.1 mmol/L, and then once every 4 h. If the blood glucose level is less than 4.4 mmol/L, it should be measured once every hour. In the blood glucose control scheme determined by Kanji et al. [64], it is required to measure blood glucose once every 1–2 h, but when the blood glucose value is ≤ 3.9 mmol/L or ≥ 14.1 mmol/L, the measurement frequency should be increased to once every 30 min. Although there are slight differences in the frequency of blood glucose monitoring proposed in different literatures, the overall trend is consistent with the recommendations given in guidelines [60,61]. That is, the interval time can be extended when the blood glucose is stable within the target range, and the monitoring frequency should be increased when the blood sugar is too high or too low.

Nevertheless, it is not sufficient to use the biochemical criterion of blood glucose value alone, because blood sugar values below the normal range do not always distinguish between normal and pathological hypoglycemia [65]. Pathological hypoglycemia is defined as a triad of hypoglycemia, symptoms of hypoglycemia and disappearance of symptoms after correction of blood sugar [66]. The symptoms caused by a sudden drop in blood sugar are related to the autonomic nervous system and the central nervous system. The former includes anxiety, palpitation, tremulousness, nausea, sweating and hunger [67,68]. The latter includes fatigue, weakness, seizures, confusion, focal neurological dysfunction and coma [69]. By observing whether patients have these clinical manifestations, we can detect the risk of hypoglycemia in time and treat it as soon as possible to prevent more serious consequences.

What's more, we should also pay attention to the replacement time of insulin configuration solution. Generally speaking, changing the insulin solution every 24 h not only ensures clinical efficacy, but also reduces drug waste and infection risk.