Type 1 Diabetes Mellitus Treatment & Management

Patients with type 1 diabetes mellitus (DM) require lifelong insulin therapy. Most require 2 or more injections of insulin daily, with doses adjusted on the basis of self-monitoring of blood glucose levels. Long-term management requires a multidisciplinary approach that includes physicians, nurses, dietitians, and selected specialists.

In some patients, the onset of type 1 DM is marked by an episode of diabetic ketoacidosis (DKA) but is followed by a symptom-free “honeymoon period” in which the symptoms remit and the patient requires little or no insulin. This remission is caused by a partial return of endogenous insulin secretion, and it may last for several weeks or months (sometimes for as long as 1-2 years). Ultimately, however, the disease recurs, and patients require insulin therapy.

Often, the patient with new-onset type 1 DM who presents with mild manifestations and who is judged to be compliant can begin insulin therapy as an outpatient. However, this approach requires close follow-up and the ability to provide immediate and thorough education about the use of insulin; the signs, symptoms, and treatment of hypoglycemia; and the need to self-monitor blood glucose levels.

The American Diabetes Association (ADA) recommends using patient age as one consideration in the establishment of glycemic goals, with different targets for preprandial, bedtime/overnight, and hemoglobin A_1c (HbA_1c) levels in patients aged 0-6, 6-12, and 13-19 years.^[33] Goals should be further individualized on the basis of awareness of hypoglycemia, frequency of hypoglycemia, and response to treatment.

Although patients with type 1 DM have normal incretin response to meals, administration of exogenous glucagonlike peptide 1 (GLP-1) reduces peak postprandial glucose by 45%. Long-term effects of exogenously administered GLP-1 analogues warrant further studies.^[56]

Pancreatic transplantation for patients with type 1 DM isg a possibility in some referral centers. It is performed most commonly with simultaneous kidney transplantation for end-stage renal disease (ESRD).

Tight glycemic control

The association between chronic hyperglycemia and increased risk of microvascular complications in patients with type 1 DM was demonstrated in the Diabetes Control and Complications Trial (DCCT).^[57] In that trial, intensive therapy designed to maintain normal blood glucose levels greatly reduced the development and progression of retinopathy, microalbuminuria, proteinuria, and neuropathy, as assessed over 7 years.

The DCCT ended in 1993. However, the Epidemiology of Diabetes Interventions and Complications Study (EDIC), an observational study that continues to follow the patients previously enrolled in the DCCT, has demonstrated continued benefit from intensive treatment.^{[58, 59]}

Benefits

Benefits of tight glycemic control include not only continued reductions in the rates of microvascular complications but also significant differences in cardiovascular events and overall mortality. These benefits occurred even though subjects in the intensively treated group and those in the standard treatment group maintained similar HbA_1c levels (about 8%), starting 1 year after the DCCT ended. It is postulated that a “metabolic memory” exists and that better early glycemic control sets the stage for outcomes many years in the future.

Increasing HbA_1c levels correlated with increasing risk of developing heart failure in a study of 20,985 patients with type 1 DM. Thus, improved glycemic control should prevent heart failure as well.^[60]

Risks

For many patients, the HbA_1c target should be less than 7%, with a premeal blood glucose level of 80–130 mg/dL. However, targets should be individualized.

Individuals with recurrent episodes of severe hypoglycemia, cardiovascular disease, advanced complications, substance abuse, or untreated mental illness may require higher targets, such as an HbA_1c of less than 8% and preprandial glucose levels of 100-150 mg/dL. The 2011 American Association of Clinical Endocrinologists (AACE) guidelines support the creation of individualized targets that consider these factors as part of a comprehensive treatment plan.^[61]

Although tight glycemic control is beneficial, an increased risk of severe hypoglycemia accompanies lower blood glucose levels. The 2011 AACE guidelines for developing a comprehensive care plan emphasize that hypoglycemia should be avoided.^[48]

In patients with type 1 DM, recurrent and chronic hypoglycemia has been linked to cognitive dysfunction.^[62] This has important implications in the management of children with type 1 DM.^[63]

An 18-year follow-up of the DCCT by Jacobson et al found that HbA_1c levels and retinal and renal complications were independently linked to cognitive declines. No relation with macrovascular risk factors or severe hypoglycemic events was found. A smoking history was modestly associated with decrements in learning, memory, spatial information processing, and psychomotor efficiency. This information is useful in advising patients with type 1 DM interested in preserving cognitive function.

Optimal diabetic control requires frequent self-monitoring of blood glucose levels, which allows rational adjustments in insulin doses. All patients with type 1 DM should learn how to self-monitor and record their blood glucose levels with home analyzers and adjust their insulin doses accordingly.

Insulin-dependent patients ideally should test their plasma glucose daily before meals, in some cases 1-2 hours after meals, and at bedtime. In practice, however, patients often obtain 2-4 measurements each day, including fasting levels and levels checked at various other times (eg, preprandially and at bedtime).

Instruct patients with type 1 DM in the method of testing for urine ketones with commercially available reagent strips. Advise patients to test for urine ketones whenever they develop any of the following:

• Symptoms of a cold, flu, or other intercurrent illness
• Nausea, vomiting, or abdominal pain
• Polyuria
• An unexpectedly high plasma glucose level on self-monitoring
• Persistent, rapid, and marked fluctuation in the degree of hyperglycemia

Continuous glucose monitoring

Continuous glucose monitors (CGMs) contain subcutaneous sensors that measure interstitial glucose levels every 1-5 minutes, providing alarms when glucose levels are too high or too low or are rapidly rising or falling. CGMs transmit to a receiver, which either is a pagerlike device or is integral to an insulin pump. Looking at the continuous glucose graph and responding to the alarms can help patients avoid serious hyperglycemia or hypoglycemia.

CGMs have several drawbacks. First, there is a lag between glucose levels in the interstitial space and levels in capillary blood, so that the levels recorded by the CGM may differ from a fingerstick (capillary) glucose reading. For that reason, the trends (ie, whether the glucose levels are rising or falling) tend to be more helpful.

Second, patients may overtreat hyperglycemia (repeatedly giving insulin because the glucose levels do not fall rapidly enough—a phenomenon known as stacking) as well as overtreat low glucose levels (because the glucose levels rise slowly with ingestion of carbohydrate).

Use of CGMs may help prevent significant glucose variability in patients receiving either multiple daily injection therapy or continuous insulin infusion therapy.^[64] Additionally, continuous glucose monitoring is associated with reduced time spent in hypoglycemia.^[65] Whether glucose variability is detrimental in the absence of hypoglycemia remains an unresolved question; in any event, variability leads to the expense of frequent testing.

Guidelines from the Endocrine Society^[66] recommend the use of real-time CGMs in adult patients with type 1 DM who have demonstrated that they are able to use these devices on a nearly daily basis. The guidelines suggest the intermittent use of CGM systems for short-term retrospective analysis in the following cases^[66] :

• Patients with suspected nocturnal hypoglycemia, dawn phenomenon, or postprandial hyperglycemia
• Patients with hypoglycemic unawareness
• Patients experimenting with important changes to their diabetes regimen (eg, instituting new insulin or switching from multiple daily injections to pump therapy)

Types of insulin

Rapid-, short-, intermediate-, and long-acting insulin preparations are available. Various pork, beef, and beef-pork insulins were previously used; however, in the United States, recombinant human insulin is now used almost exclusively. Commercially prepared mixtures of insulin are also available.

Rapid-acting insulins include lispro, glulisine, and aspart insulin. Lispro insulin is a form of regular insulin that is genetically engineered with the reversal of the amino acids lysine and proline at B28,29 in the B chain. Glulisine insulin substitutes glutamic acid for lysine in position B29. Aspart insulin substitutes aspartic acid for proline in position 28 of the B chain.

These insulins are absorbed more quickly and have a rapid onset of action (5-10 minutes), a short interval to peak action (45-75 minutes), and a short duration of action (2-4 hours). Therefore, they can be administered shortly before eating. In addition, neutral protamine Hagedorn (NPH) insulin will not inhibit the action of insulin lispro when the 2 agents are mixed together right before injection; this is not true of regular insulin.

Short-acting insulin includes regular insulin. Regular insulin is a preparation of zinc insulin crystals in solution. When it is administered subcutaneously, its onset of action occurs in 0.5 hours, its peak activity comes at 2.5-5 hours, and its duration of action is 4-12 hours.

The standard strength of regular insulin is 100 U/mL (U-100), but 500 U/mL (U-500) insulin is increasingly used, albeit mostly in type 2 DM. Accidental prescribing of U-500 rather than U-100 is a potential safety issue.^[67] A study by de la Pena et al found that although the overall insulin exposure and effects of 500 U/mL insulin are similar to those of 100 U/mL insulin, peak concentration was significantly lower with U-500, and the effect after the peak was prolonged; areas under the curve were similar for the 2 strengths.^[68]

Both regular human insulin and rapid-acting insulin analogues are effective at lowering postprandial hyperglycemia in various basal bolus insulin regimens used in type 1 DM. Rapid-acting insulin analogues may be slightly better at lowering HbA_1c and are preferred by most US diabetologists, but the differences are clinically insignificant.^[69]

Semilente insulin is like regular insulin and is a rapid-acting insulin with a slightly slower onset of action. It contains zinc insulin microcrystals in an acetate buffer. It is not readily available in the United States.

Intermediate-acting insulins include NPH insulin, a crystalline suspension of human insulin with protamine and zinc. NPH provides a slower onset of action and longer duration of action than regular insulin does. The onset of action usually occurs at 1-2 hours, the peak effect is noted at 4-12 hours, and the duration of action is normally 14–24 hours.

Lente insulin is a suspension of insulin in buffered water that is modified by the addition of zinc chloride. This insulin zinc suspension is equivalent to a mixture of 30% prompt insulin zinc (Semilente) and 70% extended insulin zinc (Ultralente). It is not used in the United States.

Long-acting insulins used in the United States include insulin glargine and insulin detemir. Insulin glargine has no peak and produces a relatively stable level lasting more than 24 hours. In some cases, it can produce a stable basal serum insulin concentration with a single daily injection, though patients requiring lower doses typically are given twice-daily injections. Insulin detemir has a duration of action that may be substantially shorter than that of insulin glargine but longer than those of intermediate-acting insulins.

A new ultralong-acting basal insulin, insulin degludec, which has a duration of action of up to 42 hours, has been developed. It awaits approval by the US Food and Drug Administration (FDA).^[70]

Mixtures of insulin preparations with different onsets and durations of action frequently are administered in a single injection by drawing measured doses of 2 preparations into the same syringe immediately before use. The exceptions are insulin glargine and insulin detemir, which should not be mixed with any other form of insulin. Preparations that contain a mixture of 70% NPH and 30% regular human insulin (eg, Novolin 70/30 and Humulin 70/30) are available, but the fixed ratios of intermediate-acting to rapid-acting insulin may restrict their use.

Insulin glargine and cancer

Controversy has arisen over a disputed link between insulin glargine and cancer. On July 1, 2009, the FDA issued an early communication regarding a possible increased risk of cancer in patients using insulin glargine (Lantus).^[71] The FDA communication was based on 4 observational studies that evaluated large patient databases and found some association between insulin glargine (and other insulin products) and various types of cancer.

The validity of the link remains in question, however. The duration of these observational studies was shorter than that considered necessary to evaluate for drug-related cancers. Additionally, findings were inconsistent within and across the studies, and patient characteristics differed across treatment groups.

In a study by Suissa et al, insulin glargine use was not associated with an increased risk of breast cancer during the first 5 years of use. The risk tended to increase after 5 years, however, and significantly so for the women who had taken other forms of insulin before starting insulin glargine.^[72]

A study by Johnson et al found the same incidences for all cancers in patients receiving insulin glargine as in those not receiving insulin glargine. Overall, no increase in breast cancer rates was associated with insulin glargine use, although patients who used only insulin glargine had a higher rate of cancer than those who used another type of insulin. This finding was attributed to allocation bias and differences in baseline characteristics.^[73]

Common insulin regimens

The goal of treatment in type 1 DM is to provide insulin in as physiologic a manner as possible. Insulin replacement is accomplished by giving a basal insulin and a preprandial (premeal) insulin. The basal insulin is either long-acting (glargine or detemir) or intermediate-acting (NPH). The preprandial insulin is either rapid-acting (lispro, aspart, or glulisine) or short-acting (regular). Currently, NPH insulin is being used less frequently, whereas insulin glargine and insulin detemir are being used more frequently.

For patients on intensive insulin regimens (multiple daily injections or insulin pumps), the preprandial dose is based on the carbohydrate content of the meal (the carbohydrate ratio) plus a correction dose if their blood glucose level is elevated (eg, an additional 2 U of rapid-acting insulin to correct the blood glucose from a level of 200 mg/dL to a target of 100 mg/dL). This method allows patients more flexibility in caloric intake and activity, but it requires more blood glucose monitoring and closer attention to the control of their diabetes.

Common insulin regimens include the following:

• Split or mixed – NPH with rapid-acting (eg, lispro, aspart, or glulisine) or regular insulin before breakfast and supper
• Split or mixed variant – NPH with rapid-acting or regular insulin before breakfast, rapid-acting or regular insulin before supper, and NPH before bedtime (the idea is to reduce fasting hypoglycemia by giving the NPH later in the evening)
• Multiple daily injections (MDI) – A long-acting insulin (eg, glargine or detemir) once a day in the morning or evening (or twice a day in about 20% of patients) and a rapid-acting insulin before meals or snacks (with the dose adjusted according to the carbohydrate intake and the blood glucose level)
• Continuous subcutaneous insulin infusion (CSII) – Rapid-acting insulin infused continuously 24 hours a day through an insulin pump at 1 or more basal rates, with additional boluses given before each meal and correction doses administered if blood glucose levels exceed target levels

Insulin is sensitive to heat and exposure to oxygen. Once a bottle of insulin is open, it should be used for no more than 28 days and then discarded; even if there is still some insulin in the bottle, it may have lost its clinical effectiveness. Insulin kept in a pump reservoir for longer than 3 days may lose its clinical effectiveness (though insulin aspart has now been approved for use for as long as 6 days in a pump).

Sometimes, insulin distributed from the pharmacy has been exposed to heat or other environmental factors and therefore may be less active. If a patient is experiencing unexplained high blood sugar levels, new insulin vials should be opened and used.

Initiation of insulin therapy

The initial daily insulin dose is calculated on the basis of the patient’s weight. This dose is usually divided so that one half is administered before breakfast, one fourth before dinner, and one fourth at bedtime. After selecting the initial dose, adjust the amounts, types, and timing according to the plasma glucose levels. Adjust the dose to maintain preprandial plasma glucose at 80-150 mg/dL (ie, 4.44-8.33 mmol/L).

The insulin dose is often adjusted in increments of 10% at a time, and the effects are assessed over about 3 days before any further changes are made. More frequent adjustments of regular insulin can be made if a risk of hypoglycemia is present.

Carbohydrate counting may be used to determine the meal-time insulin dose. Because patients may experience hyperglycemic episodes despite strict adherence to carbohydrate counting, particularly after meals that are high in protein or fat, Australian researchers developed an algorithm for estimating the mealtime insulin dose on the basis of measurements of physiologic insulin demand evoked by foods in healthy adults. The researchers showed that use of this algorithm improved glycemic control.^[74]

Initiation of insulin therapy in children

Children with moderate hyperglycemia but without ketonuria or acidosis may be started with a single daily subcutaneous injection of 0.3-0.5 U/kg of intermediate-acting insulin alone. Children with hyperglycemia and ketonuria but without acidosis or dehydration may be started on 0.5-0.7 U/kg of intermediate-acting insulin and subcutaneous injections of 0.1 U/kg of regular insulin at 4- to 6-hour intervals.

Multiple daily injections

Multiple subcutaneous insulin injections are administered to control hyperglycemia after meals and to maintain normal plasma glucose levels throughout the day. This may increase the risks of hypoglycemia. Therefore, patients should be well educated about their disease and about self-monitoring of plasma glucose levels.

About 25% of the total daily dose is administered as intermediate-acting insulin at bedtime, with additional doses of rapid-acting insulin before each meal (4-dose regimen). Where available, a basal insulin such as glargine or detemir is preferred to NPH. These patients may need additional intermediate- or long-acting insulin in the morning for all-day coverage.

Patients should adjust their daily dosage(s) on the basis of their self-monitoring of glucose levels before each meal and at bedtime. Patients should also assess their plasma glucose levels at 2:00-4:00 AM at least once per week during the first few weeks of treatment and thereafter as indicated.

Continuous subcutaneous insulin infusion

A small battery-operated infusion pump that administers a continuous subcutaneous infusion of rapid-acting insulin can provide selected, programmed basal rate(s) of insulin and a manually administered bolus dose before each meal. The patient self-monitors preprandial glucose levels to adjust the bolus dose(s).

The CSII method provides better control than the MDI method does. Initially, hypoglycemia is common with pump therapy, but once metabolic control is achieved, the risk is the same as with MDI. Bergenstal et al determined that sensor-augmented pump therapy led to better glycemic control and that more patients reached targets with this technology than with injection therapy.^[75]

Increased bedtime doses of hypoglycemic agents with nighttime peaks in action may correct early morning hyperglycemia but may be associated with undesirable nocturnal hypoglycemia. Targeted CSII programming can facilitate the prevention of early-morning hyperglycemia in selected patients.

Changes in altitude may affect delivery from insulin pumps. During the flight of a commercial airliner (200 mm Hg pressure decrease), excess insulin delivery of 0.623% of cartridge volume was demonstrated as a result of bubble formation and expansion of preexisting bubbles.^[76]

Local allergic reactions

Generalized insulin allergy is rare. Symptoms occur immediately after the injection and include urticaria, angioedema, pruritus, bronchospasm, and, rarely, circulatory shock. As a rule, allergy may be treated with antihistamines. Some cases may require epinephrine and intravenous (IV) steroids.

Local allergic reactions can occur at the site of insulin injections and can cause pain, burning, local erythema, pruritus, and induration. These complications are less common with the human insulins now in use than with the animal insulins once widely employed. Such reactions usually resolve spontaneously without any intervention.

Local fat atrophy or hypertrophy at injection sites was common with animal insulins but is rare with human insulin and insulin analogues. Patients do not require any specific treatment of local fat hypertrophy, but injection sites should be rotated. Changing to a different insulin preparation may be necessary.^[77]

Hypoglycemia may result from a change in insulin dose, a small or missed meal, or strenuous exercise. Regular insulin doses may cause hypoglycemia if the patient becomes anorectic or has another cause for reduced food intake, has gastroparesis, or is vomiting.

Common symptoms of hypoglycemia are light-headedness, dizziness, confusion, shakiness, sweating, and headache. Patients should be made aware of these symptoms and educated to respond rapidly with sugar intake. They should be advised to carry candy or sugar cubes. Family members can be taught to administer a subcutaneous injection of glucagon. In an emergency situation, initial treatment consists of a bolus injection of 25 mL of 50% glucose solution followed by a continuous glucose infusion.

Repeated hypoglycemia may be an aggravating factor in preclinical atherosclerosis. Thus, in the process of designing treatment plans aimed at reducing the glycemic burden and minimizing vascular complications, hypoglycemic episodes might negate some benefits.^[78]

Controversy surrounds the question of whether severe hypoglycemia in youths with type 1 DM has lasting cognitive consequences. In a follow-up to the DCCT, recurrent and chronic hypoglycemia was linked to cognitive dysfunction.^[62] In another study, however, electroencephalography (EEG) and cognition studies were performed at baseline and 16 years later in patients with type 1 DM, and no association between early severe hypoglycemia and subsequent reduced adult cognition or EEG changes was found.^[79]

Acute hyperglycemia, even when not associated with DKA (or hyperosmolar hyperglycemic state [HHS], which occurs most commonly in type 2 DM), is harmful for a number of reasons. If the blood glucose level exceeds the renal threshold for glucose (which is typically 240 mg/dL in a healthy person but is lower in older patients, those with renal insufficiency, and pregnant women), an osmotic diuresis ensues, with loss of glucose, electrolytes, and water.

In addition, hyperglycemia impairs leukocyte function through a variety of mechanisms. Patients with diabetes have an increased rate of wound infection, and hyperglycemia impairs wound healing.

In patients with known, poorly controlled type 1 DM, no absolute level of blood glucose elevation mandates admission to the hospital or administration of insulin in the emergency department (ED). In general, lowering the patient’s glucose level in the ED does not correct the underlying cause and has no long-term effect on the patient’s glucose levels. Therefore, a plan for lowering and monitoring the patient’s glucose levels is needed.

Adequacy of follow-up is extremely important. Whether insulin is given in the ED is of less consequence and can be decided on an individual basis.

Patients with type 1 DM can have coexisting illnesses that aggravate hyperglycemia, such as infection, coronary artery disease (CAD), or fever. Additionally, certain medications can aggravate the condition.

Diabetic ketoacidosis

DKA involves acute metabolic changes in the body that develop as a result of lack of insulin or poor response to insulin arising from stress or illness. DKA is characterized by hyperglycemia, ketosis, and acidosis, leading to osmotic diuresis and dehydration. Volume repletion, insulin therapy, and specific metabolic corrections are the keys to treatment of DKA. (See Diabetic Ketoacidosis.)

Dawn and Somogyi phenomena

The dawn phenomenon is the normal tendency of the blood glucose to rise in the early morning before breakfast. This rise, which may result from the nocturnal spikes in growth hormone that cause insulin resistance, is probably enhanced by increased hepatic gluconeogenesis secondary to the diurnal rise in serum cortisol.

Augmented hepatic gluconeogenesis and glycogen cycling are known to occur in patients with type 1 DM. However, both abnormalities, regardless of the duration of diabetes, can be corrected with intensified insulin therapy.^[80]

In some patients, however, nocturnal hypoglycemia may be followed by a marked increase in fasting plasma glucose with an increase in plasma ketones (the Somogyi phenomenon). Thus, both the dawn phenomenon and the Somogyi phenomenon are characterized by morning hyperglycemia, but the latter is considered to be rebound (counterregulation) hyperglycemia.

The existence of a true Somogyi phenomenon is a matter of debate. Most endocrinologists now believe this phenomenon reflects waning of insulin action with consequent hyperglycemia.

In cases of the dawn phenomenon, the patient should check blood glucose levels at 2:00-4:00 AM. The dawn and Somogyi phenomena can be ameliorated by administering intermediate insulin at bedtime.

Use of insulin

The insulin coverage, with a sliding scale for insulin administration, should not be the only intervention for correcting hyperglycemia, because it is reactive rather than proactive. Also, insulin may be used inappropriately when hyperglycemia reflects hepatic gluconeogenesis in response to previously uncorrected hypoglycemia.

Continue intermediate-acting (ie, NPH or Lente) insulin at 50-70% of the daily dose divided into 2 or, occasionally, 3-4 daily doses. Administer supplemental regular insulin on a sliding scale. Blood glucose should be monitored before meals and at bedtime.

One of the first steps in managing type 1 DM is diet control. According to ADA policy, dietary treatment is based upon nutritional assessment and treatment goals. Dietary recommendations should take into account the patient’s eating habits and lifestyle.

Diet management includes education about how to adjust the timing, size, frequency, and composition of meals so as to avoid hypoglycemia or postprandial hyperglycemia. All patients on insulin should have a comprehensive diet plan, created with the help of a professional dietitian, that includes the following:

• A daily caloric intake prescription
• Recommendations for amounts of dietary carbohydrate, fat, and protein
• Instructions on how to divide calories between meals and snacks

Caloric distribution is an important aspect of dietary planning in these patients. A recommended distribution consists of 20% of daily calories for breakfast, 35% for lunch, 30% for dinner, and 15% for a late-evening snack.

The minimum protein requirement for good nutrition is 0.9 g/kg/day (usual range, 1-1.5 g/kg/day), but a reduced protein intake is indicated in cases of nephropathy. Fat intake should be limited to no more than 30% of the total calories, and a low-cholesterol diet is recommended. Patients should minimize consumption of sugars and ensure that they have adequate fiber intake. In some cases, midmorning and midafternoon snacks are important to avoid hypoglycemia.

Exercise is an important aspect of diabetes management. Patients should be encouraged to exercise regularly.

Educate the patients about the effects of exercise on the blood glucose level. If patients participate in rigorous exercise for more than 30 minutes, they may develop hypoglycemia unless they either decrease the preceding insulin injection by 10-20% or have an extra snack. Patients must also make sure to maintain their hydration status during exercise.

Infections

Diabetes predisposes patients to a number of infectious diseases (see Infections in Patients with Diabetes Mellitus). These include the following:

• Malignant otitis externa
• Rhinocerebral mucormycosis
• Bacteriuria
• Pyuria
• Cystitis
• Upper urinary tract infection
• Intrarenal bacterial infection
• Skin and soft tissue infections
• Osteomyelitis

Ophthalmologic complications

Patients with preproliferative or proliferative retinopathy must immediately be referred for ophthalmologic evaluation. Laser therapy is effective in this condition, especially if it is provided before hemorrhage occurs.

Often, the first hemorrhage is small and is noted by the patient as a fleeting dark area (or “floater”) in the field of vision. Because subsequent hemorrhages can be larger and more serious, the patient should immediately be referred to an ophthalmologist for possible laser therapy. Patients with retinal hemorrhage should be advised to limit their activity and keep their head upright (even while sleeping), so that the blood settles to the inferior portion of the retina and thus obscures less of the central visual area.

Multifactorial intervention is important for slowing the progression of diabetic retinopathy. Metabolic control, smoking cessation, and blood pressure control are all protective. Patients with active proliferative diabetic retinopathy are at increased risk for retinal hemorrhage if they receive thrombolytic therapy; therefore, this condition is a relative contraindication to the use of thrombolytic agents. (See Diabetic Retinopathy and Macular Edema in Diabetes.)

Diabetic nephropathy

Extreme care should be exercised whenever any nephrotoxic agent is used in a patient with diabetes. Potentially nephrotoxic drugs should be avoided whenever possible. Renally excreted or potentially nephrotoxic drugs should be given at reduced doses appropriate to the patient’s serum creatinine level. (See Diabetic Nephropathy.)

In particular, caution should be exercised when contrast-enhanced radiologic studies are being considered in patients with diabetes who have a creatinine level higher than 2 mg/dL. Indeed, such studies should absolutely be avoided in patients with a creatinine level higher than 3 mg/dL.

Patients with diabetes who must undergo such studies should be well hydrated before, during, and after the study, and their renal function should be carefully monitored.^[81] A better solution is to seek equivalent clinical information by using an alternative modality that does not require the use of contrast material (eg, ultrasonography, noncontrast computed tomography [CT], or magnetic resonance imaging [MRI]).

Current ADA guidelines recommend annual screening for nephropathy.^[33] All adults with diabetes should have serum creatinine measured at least annually. In adults (and children aged 10 years or older) who have had type 1 DM for 5 or more years, annual assessment of urine albumin excretion is appropriate.

Microalbuminuria and macroalbuminuria are not permanent features in most diabetic children and adolescents.^[82] Regression of microalbuminuria is common; female gender, absence of retinopathy, better glucose control, lower blood pressure, and better lipid control favor this outcome.^[83] In patients with persistent microalbuminuria, the use of angiotensin-converting enzyme (ACE) inhibitors and good metabolic control can usually induce remission.

Progression and regression of kidney disease are common even after development of persistent microalbuminuria. Tight glycemic control, lower blood pressure, and a favorable lipid profile are associated with improved outcome.^[83]

When chronic kidney disease is present, reduction of protein intake may improve renal function. If kidney disease is advanced or difficult to manage or its etiology is unclear, consider referral to a physician with experience in kidney disease patient care.

Control of blood pressure is a critical element of care. An ACE inhibitor or an angiotensin II receptor blocker (ARB) should be used because these classes of agents decrease proteinuria and slow the decline in renal function independent of the effect on blood pressure.^[84] ACE inhibitors and ARBs tend to increase the serum potassium levels and therefore should be used with caution in patients with renal insufficiency or elevated serum potassium levels.

Diabetic neuropathy

Autonomic dysfunction can involve any part of the sympathetic or parasympathetic chains and produce myriad manifestations.^{[44, 85]} Patients likely to seek care in the ED are those with diabetic gastroparesis and vomiting, severe diarrhea, bladder dysfunction and urinary retention, or symptomatic orthostatic hypotension. Treatment of gastroparesis is symptomatic, and symptoms tend to wax and wane. Patients with gastroparesis may benefit from metoclopramide or erythromycin.

Before these therapies are started, the degree of dehydration and metabolic imbalance must be assessed, and other serious causes of vomiting must be excluded. In severe cases, gastric pacing has been used. Patients with disabling orthostatic hypotension may be treated with salt tablets, support stockings, or fludrocortisone. Alleviating the functional abnormalities associated with the autonomic neuropathy is often difficult and frustrating for both doctor and patient. (See Diabetic Neuropathy and Diabetic Lumbosacral Plexopathy.)

Diabetic foot disease

Patients with diabetes who present with wounds, infections, or ulcers of the foot should be treated intensively.^[86] In addition to appropriate use of antibiotics, the use of crutches, wheelchairs, or bed rest is mandatory for preventing further trauma to the healing foot. Patients should be treated by a podiatrist or an orthopedist with experience in the care of diabetic foot disease. (See Diabetic Foot and Diabetic Foot Infections.)

If bone or tendon is visible, osteomyelitis is present, and hospitalization for IV antibiotic therapy is often necessary. Many patients need a vascular evaluation in conjunction with local treatment of the foot ulcer because a revascularization procedure may be required to provide adequate blood flow for wound healing.

Because ulcers and foot infections are difficult to cure, their prevention is extremely important.^[87] At one clinic, the rate of amputation was halved after patients were required to remove their shoes and socks at every visit. The emergency physician can facilitate this practice by briefly inspecting the feet of patients with diabetes and by educating them about the need for proper foot care.

Referral to a podiatrist is indicated for diabetic patients with any of the following:

• Distal sensory neuropathy with inability to feel a pinprick or light touch
• Decreased peripheral pulses
• Moderate-to-severe onychomycosis
• Impending skin breakdown

Charcot joint, a type of arthropathy observed in people with diabetes, is a progressive deterioration of foot joints caused by underlying neuropathy. Tarsometatarsal and midtarsal joints are affected most commonly. Other neuromuscular foot deformities also may be present. Early diagnosis and treatment are important for preventing further joint degeneration.

Macrovascular disease

Hypercholesterolemia and hypertension increase the risk of specific late complications and require special attention and appropriate treatment. Although physicians can safely use beta blockers (eg, propranolol) in most patients, these agents can mask the adrenergic symptoms of insulin-induced hypoglycemia and can impair the normal counterregulatory response. ACE inhibitors are the drugs of choice for hypertension because of their renal protective action, especially early in the course of the disease.

The ADA advises that a systolic blood pressure below 130 mm Hg is an appropriate goal for most patients with diabetes and hypertension, but it also recommends modifying systolic blood pressure targets in accordance with individual patient characteristics. Diastolic blood pressure should be less than 80 mm Hg.^[33]

Subtle differences in the pathophysiology of atherosclerosis in patients with diabetes result in both earlier development and a more malignant course. Therefore, lipid abnormalities must be treated aggressively to reduce the risk of serious atherosclerosis.^[88] This is important from an epidemiologic point of view and has a bearing on the treatment strategies that must be used to mitigate the risk.^[89]

Prediction of cardiovascular risk in diabetic patients on the basis of the lipid profile is not affected by the timing of blood specimen. Therefore, it may be unnecessary to insist on using fasting blood samples to determine the lipid profile.^[90]

In a study involving diabetic adolescents and children, nocturnal hypertension was significantly associated with higher daytime blood pressure and carotid intima-media thickness, which could be precursors of atherosclerotic cardiovascular disease later in life; these findings warrant confirmation and longitudinal follow-up.^[91]

Patients with diabetes may have increased incidence of silent ischemia.^[47] However, silent ischemia is common in many patients with CAD, and the apparent increase in its incidence may come about because patients with diabetes are more likely than others to have CAD to begin with. Nevertheless, it is prudent to perform electrocardiography (ECG) in patients who have diabetes and a serious illness or who present with generalized weakness, malaise, or other nonspecific symptoms that are not usually expected to result from myocardial ischemia.

Persistent lipid abnormalities remain in patients with diabetes, despite evidence supporting the benefits of lipid-modifying drugs. Up-titration of the statin dose and addition of other lipid-modifying agents are needed.^[92] Although metformin is used principally in type 2 DM because of its lipid-lowering effect, a placebo-controlled study by Lund et al found that metformin (1000 mg orally twice daily) significantly reduced total cholesterol and low-density lipoprotein (LDL) cholesterol in patients with type 1 DM.^[93]

Serious medical illness and surgery produce a state of increased insulin resistance and relative insulin deficiency. Hyperglycemia can occur even in patients without diabetes as a consequence of stress-induced insulin resistance coupled with the use of dextrose-containing IV fluids. Increases in glucagon, catecholamines, cortisol, and growth hormone levels antagonize the effects of insulin, and the alpha-adrenergic effect of increased catecholamine levels inhibits insulin secretion. Counterregulatory hormones also directly increase hepatic gluconeogenesis.

Much less is known about optimal blood glucose levels in hospitalized patients with preexisting diabetes whose hyperglycemia reflects both their diabetes and a stress response to illness. Nonetheless, it is clear that management of hospitalized patients with preexisting diabetes requires modification of treatment regimens to compensate for both the decreased caloric intake and the increased physiologic stress. Near-normal blood glucose levels should be maintained in medical and surgical patients with diabetes, for the following reasons:

• To prevent the development of ketosis
• To prevent electrolyte abnormalities and volume depletion secondary to osmotic diuresis
• To prevent the impairment of leukocyte function that occurs when blood glucose levels are elevated
• To prevent the impairment of wound healing that occurs when glucose levels are elevated

Patients with type 1 DM must take in insulin and carbohydrate at all times to prevent ketosis. It is strongly recommended that continuous IV infusions of dextrose and insulin be used in patients who are undergoing general anesthesia or who are critically ill.

Blood glucose levels must be measured with a glucose meter every hour, and the rates of insulin and dextrose infusion must be adjusted accordingly to prevent hypoglycemia or persistent hyperglycemia.^[94] Algorithms are available for insulin infusions, and the use of a preprinted order facilitates administration and reduces dosing errors.

For patients who are less seriously ill or are undergoing minor surgery, frequent blood glucose monitoring is not always possible. These patients may do as well with subcutaneously injected insulin. A basal bolus insulin regimen, rather than a sliding-scale regular insulin regimen, should be used in these patients.

The same principles of providing a constant source of insulin and carbohydrate apply to patients with type 1 DM who must also take nothing by mouth for medical reasons. Patients should receive a basal insulin (eg, glargine or detemir insulin) with additional correction doses of regular insulin or a rapid-acting insulin. In many localities, regular insulin has been replaced by rapid-acting insulin (eg, lispro, aspart, or glulisine)

To prevent hypoglycemia, regular insulin should not be given more often than every 3-4 hours, because a dose is effective for up to 6 hours. Rapid-acting insulins may be given every 3 hours. Once the patient is eating, a preprandial insulin dose can be added.

Cardiovascular disease or renal dysfunction increases surgical morbidity and mortality, and diabetic autonomic neuropathy increases the risk of cardiovascular instability. The emergency physician caring for patients with diabetes who require emergency surgery must notify the surgeon and the anesthesiologist of the patient’s condition, consult medical specialists when appropriate, and promptly initiate a thorough medical evaluation.

Recent guidelines have trended away from stressing intensive glucose control in ill patients with diabetes. The ADA recommends that in critically ill patients, insulin therapy should be initiated if the glucose level exceeds 180 mg/dL (10 mmol/L), with a target range of 140-180 mg/dL (7.8-10 mmol/L) for the majority of critically ill patients.^[33] More stringent goals, such as 110-140 mg/dL (6.1-7.8 mmol/L), may be appropriate for selected patients, provided that significant hypoglycemia can be avoided.

In the absence of clear evidence for specific blood glucose goals in non–critically ill patients, the ADA suggests that reasonable targets are premeal blood glucose levels lower than 140 mg/dL (7.8 mmol/L) with random blood glucose levels below 180 mg/dL (10.0 mmol/L), provided that these targets can be safely achieved.^[33] It may be appropriate to use more stringent targets in stable patients with previous tight glycemic control and less stringent targets in patients with severe comorbidities.

The guidelines on glycemic control in hospitalized patients formulated by the American College of Physicians (ACP) recommend a target blood glucose level of 140-200 mg/dL if insulin therapy is used to manage patients with diabetes in nonsurgical (medical) intensive care units (ICUs).^[95] These guidelines were based on a review of 21 trials in intensive care, perioperative care, myocardial infarction, stroke, or brain injury settings.^[96]

The ACP found no convincing evidence that intensive insulin therapy reduced short-term or long-term mortality, infection rates, length of hospital stay, or the need for renal replacement therapy. In recommending 200 mg/dL as the upper target, the ACP guidelines depart from the 2009 AACE/ADA consensus statement on inpatient glycemic control, which recommended a target range of 140-180 mg/dL in critically ill patients.^[97]

Nevertheless, in certain circumstances, such as after cardiovascular surgery and during treatment in a surgical ICU, it is very important to maintain near-normal blood glucose levels in patients with acute hyperglycemia of illness. These patients should receive sufficient insulin to maintain glucose levels around 100 mg/dL.^[98]

Perioperative blood glucose management

Surgical procedures—including the preoperative emotional stress and the effects of general anesthesia as well as the trauma of the procedure itself—can markedly increase plasma glucose levels and induce DKA in patients with type 1 DM. (See Perioperative Management of the Diabetic Patient.) In patients going to surgery who have not received a dose of intermediate-acting insulin that day, injection of one third to one half of the total daily dose as NPH insulin or 80% of the dose as glargine or detemir insulin before surgery is often effective.

At the same time, an IV infusion containing 5% glucose in either 0.9% saline solution or water should be started at a rate of 1 L (50 g glucose) over 6-8 hours (or 125-150 mL/h). Blood glucose levels should be checked every 2 hours during the surgical procedure, and small doses of regular or rapid-acting insulin (eg, lispro, aspart, or glulisine) should be given if values exceed 140 mg/dL.

After the operation, check plasma glucose levels and assess for a reaction to ketones. Unless a change in dosage is indicated, repeat the preoperative dose of insulin when the patient recovers from the anesthesia, and continue the glucose infusion.

Monitor plasma glucose and ketones at 2- to 4-hour intervals, and administer regular insulin every 4-6 hours as needed to maintain the plasma glucose level in the range of 100-250 mg/dL (ie, 5.55-13.88 mmol/L). Continue until the patient can be switched to oral feedings and a 2- or 3-dose insulin schedule.

Some physicians prefer to withhold subcutaneous insulin on the day of the operation and to add 6-10 units of regular insulin to 1 L of 5% glucose in normal saline or water infused at 150 mL/h on the morning of the operation, depending on the plasma glucose level. The infusion is continued through recovery, with insulin adjustments depending on the plasma glucose levels obtained in the recovery room and at 2- to 4-hour intervals thereafter.

Postoperative IV insulin infusion after major surgical procedures is currently considered the standard of care in most hospitals.

Because pregnant patients with type 1 DM are at risk for multiple poor maternal and fetal outcomes, it is essential to provide these patients with prepregnancy counseling, good glycemic control before and during pregnancy, and a complete medical evaluation. (See Diabetes Mellitus and Pregnancy.) High-risk possibilities include exacerbation of existing hypertension, renal insufficiency, retinopathy, and more frequent congenital anomalies. These patients should be referred to obstetricians specializing in high-risk pregnancies.

Despite advanced age, multiparity, obesity, and social disadvantage, patients with type 2 DM were found to have better glycemic control, fewer large-for-gestational-age infants, fewer preterm deliveries, and fewer neonatal care admissions than patients with type 1 DM.^[99] This finding suggests that better tools are needed to improve glycemic control in patients with type 1 DM.

Significant improvements in the prediction of type 1 DM have led to several trials of prevention. These include the Diabetes Prevention Trial–Type 1 (DPT-1) in the United States and the European Nicotinamide Diabetes Intervention Trial (ENDIT) in Europe and North America. Both trials have reported disappointing results.

In DPT-1, parenteral insulin failed to delay or prevent type 1 DM in subjects at elevated risk (as indicated by family history and the presence of islet cell antibodies). These subjects received low-dose subcutaneous Ultralente insulin twice daily, plus annual 4-day continuous IV infusions of insulin.^[100] DPT-1 subjects who received oral insulin experienced considerable delays in the onset of diabetes, but once therapy was stopped, their rate of developing diabetes increased to a rate similar to that seen in the placebo group.^[101]

In the ENDIT study, nicotinamide (which prevents autoimmune diabetes in animal models) did not prevent or delay the clinical onset of diabetes in people with a first-degree family history of type 1 DM. Subjects in the treatment arm received oral modified-release nicotinamide in a dose of 1.2 g/m².^[102]

Slowing progress of recent-onset type 1 DM

In animal models of autoimmunity, treatment with a target antigen can modulate aggressive autoimmunity. However, a trial of antigen-based immunotherapy with 2 or 3 doses of glutamic acid decarboxylase formulated with aluminum hydroxide (GAD-alum) vaccine for 4-12 weeks in patients with newly diagnosed type 1 DM did not alter the course of loss of insulin secretion during the first year.^[103]

A phase 3 trial using an anti-CD3 monoclonal antibody, teplizumab, found an encouraging trend toward preservation of beta-cell function with reduction in daily insulin requirements in patients with recently diagnosed type 1 DM. However, rash was almost 3 times more common in treated patients than in those receiving placebo.^[104]

A study by Orban et al found that costimulation modulation of activated T cells with abatacept slowed reduction in beta-cell function over a 2-year period of administration. However, this effect was reduced after 6 months of treatment, suggesting that T-cell activation lessens over time. Further studies are needed.^[105]

Patients with type 1 DM should be referred to an endocrinologist for multidisciplinary management. They should also undergo a complete retinal examination by an ophthalmologist at least once a year. Those patients with significant proteinuria or a reduced creatinine clearance should be referred to a nephrologist. Patients with significant foot involvement should see a podiatrist.