Diabetes Care in Hospital Settings
- Recommendations
▪️ A1C test should be performed for any patient with random blood glucose sample > 140 mg/dl during hospitalization or for any patient with a known history of diabetes. (strong recommendation, moderate certainty evidence) 3
▪️ Health care facilities should implement validated protocols (written or computerized) and delivered to all departments including: emergency units, ICU units, general wards, obstetrician/delivery units and dialysis suits. (strong recommendation, moderate certainty evidence)4-11
Remarks
It is reasonable to check A1C level in any hospitalized patient with a random blood glucose sample > 140 mg/dl. This is the cutoff point for further screening in hospitalized patients. It is also very important to have validated protocols (written or computerized) and delivered to all departments to ensure high quality management for all the hospitalized patients with dysglycemia.
The National Academy of Medicine advocates for the implementation of computerized provider order entry (CPOE) to mitigate medication-related errors and streamline medication administration processes 10
Summary of evidence
Reviews of randomized controlled trials have revealed notable improvements in glycemic outcomes within hospital settings through the utilization of computerized advice systems. These improvements include increased time spent within the desired glycemic range, reduced average blood glucose levels, and no rise in hypoglycemia rates11. Whenever possible, hospitals should integrate structured order sets to offer computerized guidance for managing glycemic levels effectively. Additionally, there is promising potential in utilizing machine learning algorithms and data from electronic health records (EHRs) to forecast insulin requirements during hospitalization 12.
Evidence supports the preadmission treatment of hyperglycemia in individuals scheduled for elective surgery as an effective strategy for reducing adverse outcomes. 4-12
Rationale for the recommendation
Delivering high-quality hospital care for diabetes necessitates the establishment of standards for care delivery, preferably implemented through structured order sets and quality improvement strategies. However, the consistent implementation of "best practice" protocols, reviews, and guidelines within hospitals remains a challenge. To address this issue, medical centers aiming for optimal inpatient diabetes treatment should develop protocols and structured order sets, incorporating computerized provider order entry (CPOE). Regular audits are recommended to monitor adherence to these protocols, alongside the implementation of educational and training programs to ensure staff remain updated. 3
During initial evaluation, it is essential to specify the type of diabetes, including type 1, type 2, gestational, pancreatogenic, drug-related, or nutrition-related, if known. Given that inpatient treatment and discharge planning are more effective when considering preadmission glycemia, A1C should be assessed for all individuals with diabetes or dysglycemia admitted to the hospital, particularly if no A1C test result is available from the previous three months 4-7. Additionally, admission assessments should include an evaluation of diabetes self-management knowledge and behaviors, with provision for diabetes self-management education, especially if a new treatment plan is being considered. This education should encompass knowledge and skills necessary for post-discharge management, including medication dosing and administration, glucose monitoring, and the recognition and treatment of hypoglycemia 8
To ensure lower incidence of complications during hospitalization, medical centers aiming for optimal inpatient diabetes treatment should develop protocols and structured order sets, incorporating computerized provider order entry (CPOE). Regular audits are recommended to monitor adherence to these protocols, alongside the implementation of educational and training programs to ensure staff remain updated 3.
l When caring of hospitalized patients with existing diabetes or stress induced hyperglycemia, consult with a diabetes specialized team whenever possible. (Conditional recommendation, moderate certainty evidence).13
Remarks
Specialized diabetes team should be responsible for caring with people with diabetes during hospitalization.
Summary of evidence
In a cross-sectional study that compared standard care to cases where diabetes specialists reviewed patient records and provided recommendations remotely through the Electronic Health Record (EHR), both hyperglycemia and hypoglycemia rates saw a reduction of 30-40% 13.
Rationale for the recommendation
The provision of care by adequately
trained specialists or specialized teams could potentially shorten hospital
stays and enhance outcomes related to glycemic control
14-15. Furthermore, the heightened risk of readmission within 30
days post-hospitalization associated with diabetes can be mitigated, resulting
in cost savings, through the delivery of inpatient care by a specialized team
focused on diabetes management 16-17.Diabetes specialized teams
could potentially shorten hospital stays and enhance outcomes related to
glycemic control.13
l Insulin should be initiated and intensified for the treatment of persistent hyperglycemia (blood glucose level ≥ 180 mg/dl on two occasions within 24 hours). (strong recommendation, high certainty evidence)18-22
l Once insulin therapy started, a glycemic goal of 140 – 180 mg/dl is recommended for most critically ill and noncritically ill hospitalized patients. (strong recommendation, high certainty evidence)18-22
l More stringent goals such as 100 – 140 mg/dl may be acceptable in some patients if these goals could be reached without significant hypoglycemia. (conditional recommendation, moderate certainty evidence)18-22
Remarks
Hyperglycemia among hospitalized patients is defined as blood glucose level ≥ 180 mg/dl on two occasions within 24 hours. A glycemic goal of 140 – 180 mg/dl is recommended for most critically ill and noncritically ill hospitalized patients.3
Summary of evidence
In a large clinical study carried out in a surgical intensive care unit (ICU), Van den Berghe et al. revealed that employing an intensive intravenous insulin regimen targeting a glycemic goal of 80-110 mg/dL resulted in a 40% reduction in mortality compared to a conventional approach aiming for a glycemic goal of 180-215 mg/dL in critically ill patients recently subjected to surgery. This investigation offered compelling evidence that actively managing blood glucose levels in hospitalized individuals could yield immediate advantages .18 Nevertheless, a significant multicenter follow-up investigation in critically ill hospitalized individuals, known as the Normoglycemia in Intensive Care Evaluation and Survival Using Glucose Algorithm Regulation (NICE-SUGAR) trial, prompted a reassessment of the optimal glucose-lowering target in critical illness. In this trial, critically ill patients randomly assigned to intensive glycemic control (aiming for 80-110 mg/dL) did not exhibit any substantial treatment advantage compared to a cohort with more moderate glycemic targets (140-180 mg/dL) and experienced slightly but notably higher mortality rates (27.5% vs. 25%). The group subjected to intensive management had markedly elevated rates of hypoglycemia, which likely contributed to the observed adverse outcomes.19
The findings from the NICE-SUGAR trial, corroborated by numerous meta-analyses and a randomized controlled trial, revealed heightened hypoglycemia rates and a rise in mortality associated with more aggressive glycemic control objectives compared to moderate glycemic targets 20-22. Based on these results, insulin should be initiated for the treatment of persistent hyperglycemia > 180 mg/dL. Once therapy is initiated, a glycemic goal of 140-180 mg/dL is recommended for most critically and noncritically ill individuals with hyperglycemia 3. In noncritical care settings, managing hyperglycemia in hospitalized patients typically involves aiming for a glycemic target of 100-180 mg/dL, whether the hyperglycemia is newly diagnosed (e.g., stress hyperglycemia or newly diagnosed diabetes) or is related to diabetes diagnosed prior to admission 3.
l For hospitalized individuals with diabetes who are consuming meals, it's recommended to conduct point-of-care (POC) blood glucose monitoring before meals. For those who are not eating, glucose monitoring every 4-6 hours is advised. However, when using intravenous insulin therapy, more frequent POC blood glucose monitoring, ranging from every 30 minutes to every 2 hours, is necessary to ensure safe implementation. (Good clinical practice, low certainty evidence)3, 23
l Basal insulin or a basal plus bolus correction insulin is the preferred treatment for noncritically ill hospitalized patients with poor oral intake or who are fasting. (strong recommendation, high certainty evidence )(29-35)
l An insulin regimen consisting of basal, prandial and correction components is the preferred treatment of for noncritically ill hospitalized patients with adequate nutritional intake. (strong recommendation, high certainty evidence)(24-30)
l The sole use of sliding scale insulin regimen in the inpatient settings is strongly discouraged. (strong recommendation, high certainty evidence )(24-30)
Remarks
In most cases, insulin is the preferred treatment for hyperglycemia in hospitalized individuals. Continuous intravenous insulin infusion stands out as the most effective approach for attaining specific glycemic targets and averting hypoglycemia in critical care settings. Beyond critical care units, scheduled subcutaneous insulin regimens are recommended for managing hyperglycemia in individuals with diabetes. For noncritically ill hospitalized patients with insufficient or limited oral intake, basal insulin or a basal plus bolus correction regimen is favored. Meanwhile, for most noncritically ill hospitalized individuals with diabetes who have adequate nutritional intake, an insulin regimen incorporating basal, prandial, and correction components is preferred. (31-35)
Summary of evidence
A randomized controlled trial has shown that basal plus bolus treatment improved glycemic outcomes and reduced hospital complications compared with a correction or supplemental insulin without basal insulin (formerly known as sliding scale) for people with type 2 diabetes admitted for general surgery.36
A prospective randomized inpatient study of 70/30 intermediate-acting (NPH)/ regular insulin mixture versus basal-bolus therapy showed comparable glycemic outcomes but significantly increased hypoglycemia in the group receiving insulin mixture.37Therefore, insulin mixtures such as 75/ 25, 70/30, or 50/50 insulins are not routinely recommended for in-hospital use. Data on the use of glargine U-300 and degludec U-100 or U-200 in the inpatient and perioperative settings are limited. A few studies have shown that they demonstrated similar efficacy and safety compared with glargine U-100.(38-42) At this time, there is no available evidence for weekly insulin use in hospital or surgical settings.3
· Top of Form
l For patients with type 2 diabetes who are admitted with heart failure (after recovery from the acute illness), it is recommended to initiate or continue SGLT 2 inhibitors during hospitalization and upon discharge, if there are no contraindications to this group. (Strong recommendation, high certainty evidence).36-41
Remarks
Individuals with type 2 diabetes hospitalized due to heart failure are advised to consider initiating or continuing the use of a sodium-glucose cotransporter 2 (SGLT2) inhibitor during hospitalization and upon discharge, provided there are no contraindications and after recovery from the acute illness. However, SGLT2 inhibitors should be avoided in cases of severe illness, ketonemia or ketonuria, and during prolonged fasting or surgical procedures.3
Proactive adjustment of diuretic dosing is advised during hospitalization and/or discharge, particularly in collaboration with a cardiology/heart failure consult team. Additionally, the FDA cautions that SGLT2 inhibitors should be stopped 3 days before scheduled surgeries (4 days in the case of ertugliflozin). 42
l In adult patients who are hospitalized for noncritical illness and experience hyperglycemia while receiving glucocorticoids (GCs) glycemic management should be pursued with either neutral protamine Hagedorn (NPH)-based insulin or basal bolus insulin (BBI) regimens (conditional recommendation, low certainty evidence) 43,44
Remarks
An NPH-based regimen may consist of NPH (with or without prandial insulin) given in divided doses depending on the timing, pharmacokinetics, and frequency of the specific GC being administered. NPH insulin may be added to BBI if the patient is already ont this regimen.45
Management of patients with GC-associated hyperglycemia requires ongoing BG monitoring with adjustment of insulin dosing. All therapies require safeguards to avoid hypoglycemia when doses of GCs are tapered or abruptly discontinued.45
The feasibility of implementing complex insulin regimens may be difficult for nursing personnel, placing additional burdens that have potential to affect patient safety. However, from a patient perspective, a once-daily morning NPH regimen may be easier to learn than multiple daily injections, particularly for patients who will be discharged home on GC therapy .45
patient perspective, a once-daily morning NPH regimen may be easier to learn than multiple daily injections, particularly for patients who will be discharged home on GC therapy.46
Summary of evidence
The systematic review identified 6 RCTs and 1 non-RCT to address this question (1. Much variability occurred among the studies regarding the insulin regimens used in both the NPH-based groups and the comparator groups. 47
NPH showed a lower BG (blood glucose ) in 4 RCTs while 2 RCTs found no difference and one observational study . No differences occurred in rates of hypoglycemia measured as the number of patients with an event or the number of events per patient in the 5 other studies that reported these data (all with very low certainty of evidence). In 2 RCTs, no differences occurred in hospital length of stay (low certainty evidence).48
Comments
These recommendations has been based on low-certainty evidence demonstrating similar glycemic outcomes for mean BG, hyperglycemia, hypoglycemia, and hospital LOS with NPH and BBI-based regimens for patients with GC-associated hyperglycemia in the hospital. Therefore it has been suggested that either NPH- or BBI-based regimens for glycemic management of GC-associated hyperglycemia. Neither regimen demonstrated cost, feasibility, acceptability, or equity advantages.49
Barriers to addressing GC-associated hyperglycemia may be bridged by establishing protocols and guidelines that outline best practices for achieving and maintaining glycemic control, such as administration of NPH at the same time as intermediate-acting GCs, such as prednisone or methylprednisolone. 49
l In select adult patients with mild hyperglycemia and T2D hospitalized for a noncritical illness, we suggest using either DPP4i with correction insulin or scheduled insulin therapy. (Conditional recommendation, low certainty evidence) 47
Remarks
Select patients include those with T2D that is moderately well-managed as reflected by a recent HbA1c < 7.5% (9.4 mmol/L), BG < 180 mg/dL (10 mmol/L), and, if on insulin therapy before hospitalization, to have a total daily insulin dose < 0.6 units/kg/day; this recommendation applies both to patients taking the DPP4i before admission and those who are not.34
Patients who develop persistently elevated BG [eg, >180 mg/dL (10 mmol/L)] on DPP4i therapy should be managed with scheduled insulin therapy; this recommendation does not apply to patients with T1D or other forms of insulin-dependent diabetes.34
Summary of evidence
Based on a metanalysis of 3 RCTs performed in individuals with established T2D prior to hospitalization, DPP4i dosed once daily compared with BBI therapy may provide no benefit on glycemic management . In select patients, there may be a reduced insulin requirement and lower frequency of hypoglycemic events [RR 0.27 (95% CI 0.09 to 0.84); low certainty evidence]. 47
The incidence of hypoglycemia was reduced with use of DPP4i in several trials; however, patients with impaired renal function and those considered to be at higher risk of hypoglycemia and hyperglycemia were excluded from enrollment. DPP4is are approved for use and considered safe in patients with any degree of kidney disease (note that dose adjustment for renal dysfunction is required for select DPP4is; eg, sitagliptin and alogliptin). Therefore, while patients with advanced kidney disease may benefit from reduced hypoglycemia, this remains unknown.1, 50
Of importance, a meta-analysis excluded those studies in which the intervention was a combination of DPP4i and scheduled insulin. However, all RCTs except 1 comparing DPP4i to BBI allowed the use of correction insulin for intermittent hyperglycemia. Finally, all RCTs included criteria for conversion to scheduled insulin therapy in the case of persistent hyperglycemia.1, 50
l For adult patients with diabetes undergoing elective surgical procedures, we suggest targeting preoperative hemoglobin A1c (HbA1c) levels < 8% (63.9 mmol/mol) and blood glucose (BG) concentrations 100 to 180 mg/dL (5.6 to10 mmol/L). (Conditional recommendation very low certainty evidence)(47,51, 52-62)
l For adult patients with diabetes undergoing elective surgical procedures, when targeting hemoglobin A1c (HbA1c) to < 8% (63.9 mmol/mol) is not feasible, we suggest targeting preoperative blood glucose (BG) concentrations 100 to 180 mg/dL (5.6 to 10 mmol/L). (Conditional recommendation, very low certainty evidence)(63-76)
Summary of evidence
A systematic review identified 44 observational (non-RCT) studies to address this question 47. A minority of patients with T1D were included, a single study recruited only patients with T1D 51, and many studies did not specify the type of diabetes. The majority of studies were performed in patients undergoing cardiac and orthopedic surgery, but other surgeries were included. Studies also used different cutoff values for HbA1c and BG concentrations and different strategies and interventions for peri- and intraoperative glucose control, which are likely to have affected outcomes. A meta-analysis of 11 non-RCTs that measured hospital length of stay comparing patients with a preoperative HbA1c < 7% vs ≥ 7% (<53 mmol/mol vs ≥53 mmol/mol) reported a shorter length of stay [MD −0.45 days (95% CI –0.89 to 0.00); very low level of certainty].(52-62)
In 10 non-RCTs, postoperative infections were less frequent in patients with a preoperative HbA1c < 7% vs ≥ 7% [<53 mmol/mol vs ≥53 mmol/mol; odds ratio 0.54 (95% CI 0.40 to 0.73); very low level of certainty] (52, 53, 58, 63-69). Similar findings were observed in 2 non-RCTs that compared patients with a preoperative HbA1c < 8% vs ≥ 8% [<63.9 mmol/mol vs ≥63.9 mmol/mol; odds ratio 0.83 (95% CI 0.15 to 4.63); very low level of certainty] (70,71). One study reported a HbA1c ≥ 7.8% (61.7 mmol/mol) as the threshold above which a significantly higher rate of wound complications occurred (70). The incidence of postoperative infections may also be reduced in patients with better glycemic measures in studies using different HbA1c cutoff levels (<6.5% vs ≥6.5%, <7.5% vs ≥7.5%, and < 8% vs ≥8%; <48.6 mmol/mol vs ≥48.6 mmol/mol, <58.5 mmol/mol vs ≥58.5 mmol/mol, and < 63.9 mmol/mol vs ≥63.9 mmol/mol, respectively) (72-76).
l In adult patients hospitalized for noncritical illness who are receiving enteral nutrition with diabetes-specific and nonspecific formulations, use neutral protamine Hagedorn (NPH)-based or basal bolus regimens. (conditional recommendation , low certainty evidence)(77-81)
Summary of evidence
The systematic review identified 2 systematic reviews, 1 RCT, and 3 non-RCTs that address this question. Studies found little to no difference in mean daily blood glucose between basal or basal bolus insulin vs NPH-based regimens with correctional (sliding scale) insulin. One observational (non- RCT) study found that the average hospital length of stay may be reduced by 1.57 days with NPH regimens compared to basal bolus (95% CI −1.71 to 4.85), and another found that the number of hypoglycemic events may result in an increase of 41% with basal-bolus compared to 70/30-biphasic insulin [IRR 2.92 (95% CI 0.70 to 12.20)], but these results are uncertain. No studies reported outcomes related to nurse time and effort. (77-81)
l Metformin administration should be withheld on the day of surgery. Other oral glucose-lowering agents should be withheld on the morning of the surgery or procedure, and instead, administer half of the NPH dose or 75-80% doses of long-acting analog insulin. (Good clinical practice )3
l SGLT2 inhibitors should be discontinued 3–4 days before surgery. (Good clinical practice )3
l Reducing basal insulin by 25% the evening before surgery, compared to usual dosing, is more likely to help achieve perioperative blood glucose goals with a reduced risk of hypoglycemia. (Good clinical practice, low certainty evidence )3
l In individuals undergoing noncardiac general surgery, utilizing basal insulin in combination with premeal short or rapid-acting insulin (basal-bolus) coverage has been linked to enhanced glycemic outcomes and reduced rates of perioperative complications compared to relying solely on reactive, correction-only short- or rapid-acting insulin coverage without basal insulin dosing.(Good clinical practice, low certainty evidence )3
l Treatment plans should be reviewed and changed as necessary to prevent hypoglycemia and recurrent hypoglycemia when a blood glucose value of <70 mg/dL (<3.9 mmol/L) is documented. (Conditional recommendation, low certainty evidence)3
Remarks
All health care professionals should be aware of hypoglycemia management protocol
Rationale for the recommendation
Numerous instances of inpatient hypoglycemia can be prevented. Therefore, it's essential for each hospital or hospital system to adopt and implement a hypoglycemia prevention and management protocol.3
l A structured discharge plan discharge plan should be tailored to the individual with diabetes. (strong recommendation, moderate certainty evidence )82
Remarks
Implementing a structured discharge plan customized to each individual can potentially decrease the length of hospital stay and reduce readmission rates while enhancing satisfaction with the hospital experience. Key strategies include providing diabetes self-management education before discharge, reconciling diabetes medications with a focus on access, and scheduling virtual and/or face-to-face follow-up visits post-discharge. It's crucial to initiate discharge planning upon admission and to continually update it as the individual's needs evolve 82.
It is recommended that an outpatient follow-up visit with the primary care clinician, endocrinologist, or diabetes care and education specialist within one month of discharge. If there have been changes to glycemic medications or if glucose management is not optimal at discharge, an earlier appointment (within 1-2 weeks) is preferred. In some cases, frequent contact may be necessary to prevent hyperglycemia and hypoglycemia.A discharge algorithm for adjusting glycemic medication, which takes into account admission A1C levels, diabetes medications before admission, and insulin usage during hospitalization, has been found to be useful in guiding treatment decisions and significantly improving A1C levels after discharge 5.
Rationale for the recommendation
Implementing a structured discharge plan tailored to each individual has the potential to reduce the length of hospital stay and lower readmission rates, all while enhancing satisfaction with the hospital experience.