The ‘Call to Arms’
The ‘Call to Arms’
The 2007 Centers for Disease Control (CDC) National Diabetes Fact Sheet reports that 23.6 million persons in the US have diabetes. Of these, 17.9 million have been diagnosed with type 2 diabetes and unfortunately 5.7 million of them are unaware that they have the disease. Of those with known diabetes, it is estimated that 5–10% have type 1 diabetes. The annual economic cost of diabetes in 2007 was estimated to be $174 billion. Of this total bill, medical expenditures accounted for $116 billion, diabetes care amounted to $27 billion, chronic diabetes-related complications contributed $58 billion, and excess general medical costs were $31 billion.1 Thus, decisions on treatment options for diabetes will have a huge effect on healthcare budgets. Clearly, more emphasis must be placed on treating diabetes with insulin because data from the National Health and Nutrition Examination Survey (NHANES) showed that only 36% of patients with type 2 diabetes are treated to a goal of glycated hemoglobin (HbA1c) <7%. Overall, only 16% of patients were on an insulin-based regimen, despite the fact that insulin therapy has the capacity to control any degree of hyperglycemia.2
There are many reasons for suboptimal use of insulin, and these include fears of hypoglycemia and weight gain and both physician and patient resistance to implementing an aggressive management regimen. Peyrot et al. demonstrated in the multinational DAWN study that patient and provider attitudes play a major role in starting insulin therapy to control type 2 diabetes.3 More than half of the 3,790 diabetes care providers (1,109 nurses and 2,681 physicians) surveyed reported delaying insulin until it was absolutely necessary. Other data from the Kaiser Permanente Northwest Region demonstrated that from the time of diagnosis of diabetes until insulin was started, the HbA>sub>1c level had been >7% for 10 years and >8% for at least five years.4 Fewer data are available about the degree of glycemic control in people with type 1 diabetes. Some difficulty arises because the HbA1c goals differ depending on the age group. The American Diabetes Association (ADA) recommends an HbA1c level of between 7.5 and 8.5% for those aged zero to six years, an HbA1c level <8% for those aged between six and 12 years, and an HbA1c level <7.5% for those aged between 13 and 19 years.5 Overall, the consensus is that only a minority of people with diabetes are managed to the recommended goals.
Deciding Between Pumps and Multidose Regimens
The role of insulin therapy is clear for those with type 1 diabetes. Insulin is an absolute life-long requirement from the time of diagnosis. However, almost 50% of physicians do not appreciate the need for insulin in the management of type 2 diabetes.2 The consequences of uncontrolled diabetes are well-known and the capacity to reduce the risk of micro- and macrovascular disease with aggressive insulin therapy treatment has been supported by major world studies such as the Diabetes Control and Complications Trial (DCCT), the Kumamoto study, the UK Prospective Diabetes Study (UKPDS), and the newer Action to Control Cardiovascular Risk in Diabetes (ACCORD) and Action in Diabetes and Vascular disease: Preterax and Diamicron MR Controlled Evaluation (ADVANCE) trials.6–10 Thus, the question arises as to the best approach for insulin therapy. Should a patient use an insulin pump or a multidose injection protocol? Insulin pump therapy is a well established approach for intensive insulin therapy. However, modern basal–prandial insulin regimens are close to mimicking physiological insulin delivery. In this article we will review the current evidence on the efficacy of external insulin pumps versus basal–prandial insulin therapy.
Decisions about prescribing a pump or multidose insulin schedule are not straightforward. Patient and physician attitudes, capabilities, experience, and resources play a major role in the decision to go ahead with insulin pump management of type 1 diabetes. These same factors must be weighed for persons with other types of diabetes, such as type 2 diabetes and late-stage type 2 diabetes evolving into functional type 1 diabetes. Scientific criteria that affect decisions concerning pump versus multidose therapy are scarce because there are incomplete data comparing glycemic control with the two technologies. Another factor that weighs heavily is the cost of pump technology for the device and ongoing expenses for supplies compared with a multidose regimen.11 The costs for education, insulin, and glucose testing are not significantly different for pump management and basal–prandial therapy because the training and strategies are essentially the same. However, cost must be balanced against long-term patient outcomes and quality of life issues.
In one report, Cerosimo and colleagues reported fewer emergency room visits and hospitalizations for patients with type 1 diabetes treated with pumps than for those using multidose insulin regimens.12
The introduction of insulin analogs into new pen devices or modern pumps and continuous glucose monitoring devices has enabled us to closely mimic physiological insulin secretion. Insulin analogs aside, glulisine or lispro afford excellent glycemic control with fewer hypoglycemic excursions than the older neutral protamine Hagedorn (NPH) and regular insulin formulations.13
Furthermore, insulin pumps may offer additional benefits such as better insulin absorption and decreased glycemic variability, leading to a 14–17% reduction in total daily insulin requirements compared with multiple daily injections (MDIs).14,15 The idea of a pump for continuous insulin delivery was first conceived about four decades ago. After early experiments focusing on intravenous insulin therapy failed, Pickup et al.16 (in 1978) and Tamborlane et al.17 (in 1979) were successful in using pumps to supply insulin by the subcutaneous route. They also demonstrated that continuous subcutaneous insulin infusion with patients performing self-monitoring of blood glucose could achieve near-normoglycemia. Much interest in pump therapy was generated after the results of the DCCT were published, and this led to many technological advances, including miniaturization and the development of powerful microprocessor controls with multiple safety features.18 With these improvements, acceptance of insulin pump therapy broadened and currently about 20% of patients with type 1 diabetes in the US use pumps.19 To date, the US Food and Drug Administration (FDA) has approved more than 50 pump devices.
The insulin pump is a battery-driven device about the size of a pager. A tiny electric motor controlled by a microprocessor slowly drives a plunger into an insulin reservoir holding 2–3ml of insulin. The plunger and reservoir system is similar in concept to the functioning of a syringe. Insulin from the reservoir is delivered to the patient by a thin tube that terminates in a 1⁄2-inch needle for insertion into their abdomen or leg. Most people use aspart, glulisine, or lispro insulin in the pump reservoir. These fast analogs are used for both basal and prandial insulin needs. The continuous basal infusion disposes of basal glucose produced by the liver in the fasting state. Bolus doses of the fast analog dispose of the glucose load derived from a meal. One advantage of using a fast analog is that its rapid onset of action and short half-life better approximate the normal first- and second-phase insulin responses that occur when a person consumes a meal. As pumps are programmable, basal and bolus doses of insulin can be finely tailored to achieve tight glycemic control in the fed and fasting states.20,21 The current systems are described as ‘open-loop’ because the user must make final decisions on the amount of insulin to be delivered based on blood glucose monitoring. Research is ongoing for ‘closed-loop systems,’ where next-generation sensors will directly communicate with the pump and regulate insulin delivery without input from the patient and, in effect, will function as an artificial pancreas.
Are Pumps Better Than Multidose Regimens?
Is insulin pump therapy superior to multidose insulin regimens? Pickup et al. recently reviewed the data and the conclusion was that better control can be achieved using insulin pumps in patients with type 1 diabetes.22 It appears that overall HbA1c reductions were about 0.5% better than NPH multidose regimens; however, similar HbA1c reductions could be achieved using basal insulin glargine and prandial insulin aspart, glulisine, or lispro. Perhaps more importantly, regardless of whether the HbA1c was lower or the same, type 1 diabetes patients using insulin pumps experienced fewer episodes of hypoglycemia. Thus, pump therapy may be safer than multidose regimens for those with type 1 diabetes.14,15,23,24
The data on patients with type 2 diabetes using pumps versus multidose therapy are scarce. The study by Herman et al. compared pump management using lispro with a multidose protocol with basal glargine and prandial lispro.25 Both regimens resulted in the same HbA1c reduction (1.7%) and the incidence of hypoglycemia and weight gain was the same. The study by Raskin et al. evaluated pump therapy with aspart compared with once- or twice-daily NPH plus prandial aspart. The average baseline HbA1c was 8% and both groups realized an HbA1c reduction of about 1%. Furthermore, daytime glycemic excursions were smaller and patient satisfaction was higher in the insulin pump group.26 Other studies exist, but the relevance of using pumps or multidose regimens to reduce the HbA1c from 10 to 8% misses all current targets for glycemic control. Although limited data exist, it appears that pumps and multidose regimens are probably equally effective in patients with type 2 diabetes.
Who Is an Insulin Pump Candidate?
One of the most important factors when considering insulin pump management of diabetes concerns patient selection. The ADA recommends that patients should be strongly motivated to achieve glycemic control and able to understand and demonstrate the psychomotor skills necessary for pump management. In addition, the patient must be willing to monitor blood glucose on a frequent basis (four to six times each day) and have the ability to interpret and manage the data.27 Bode has elaborated more specific criteria. Considerations should be given to patients with inadequate glycemic control defined as HbA1c >7%, dawn phenomenon with morning blood glucose >144–162mg/dl, marked daily variations in blood glucose, a history of hypoglycemia unawareness, a need for lifestyle flexibility, pregnancy or intention to become pregnant, and/or low insulin requirements (<20 units/day).28 In our clinics, we also believe other factors such as physician experience and the availability of certified diabetes educators (CDEs) with insulin pump expertise are critical. We believe this last point cannot be overemphasized. An experienced team is essential for complete patient education, follow-up care, and problem-solving. We consider insulin pumps to be great tools to achieve glycemic control, but the user must be actively participating in management of their diabetes. Too often patients are referred to us who were simply prescribed a pump elsewhere, given two hours of instruction, and sent away with poor control and essentially zero working skills. Blind use of a pump can be an expensive proposition if good control is not achieved, and can result in tragic outcomes if hypoglycemia occurs and the pump continues to supply yet more insulin.
Are there other factors to be considered when deciding on insulin pump therapy or multidose regimens? We believe the answer is yes because there are nebulous ‘gray areas’ in patient selection. Schade and colleagues suggest a ‘very cautious’ approach to which patients should be recommended for pump therapy.29 Wysocki recently wrote an editorial discussing the clinical, ethical, and scientific problems associated with defining the inclusion criteria.30 Many of us require that prior to pump initiation the patient is seen by a CDE to determine readiness to start pump therapy, identification of gaps in skills and knowledge, assessment of available support systems, and education to ensure that the patient and care-giver have realistic expectations about what can be achieved and what will be necessary to reach glycemic goals.27,31
How Do Patients Begin Insulin Pump Therapy?
In our clinics, we do not start pump therapy until the patient has undergone diabetes education and medical nutrition therapy. The following discussion highlights the critical role played by a CDE. The patient must know how to recognize and treat hypoglycemia and manage ‘sick days.’ Prerequisite skills include carbohydrate counting and the use of correction doses of insulin to manage hyperglycemic occurrences. After the patient has mastered these topics, he or she needs to be able to manage his or her diabetes using a basal–prandial regimen with insulin pens or syringes. This is necessary for two reasons. First, it allows the patient to learn insulin management based on a basal–prandial system that will be utilized in pump therapy. Second, the patient must know how to use pens or syringes because a back-up system must be in place if pump failure occurs. After facility with the basal–prandial approach is achieved, the pump can then be introduced and attention given to the psychomotor components of operating the pump.
Which Pump and Which Insulin?
Is one insulin pump superior to another? Probably not. Different pumps have different features and a CDE can provide much guidance on pump selection. Recommendations will be different for a child and an adult. Lifestyle requirements can enter into the decision. Some patients may opt for a disposable pump without connecting tubing, while others will select the durable units. Reservoir size, basal and bolus dosing increments, and safety features are important. Other factors include personal preferences such as the programming protocols, screen features, physical size, food menus, and communication ports for peripheral equipment such as sensors.
What type of insulin is used in pump therapy? Buffered regular insulin and the fast insulin analogs aspart, glulisine, and lispro are currently approved for use in external pumps for adults. Glulisine was recently approved by the FDA for use in children above six years of age. Most patients now use the fast analogs in their pumps because there are data showing improved post-prandial glycemic control compared with regular insulin.32–36 Many patients have fewer episodes of hypoglycemia using aspart, glulisine, or lispro compared with regular insulin.37–43 It is not clear if there are fewer occlusions of the tubing and needle when comparing the various insulin preparations. Some data suggest that glulisine use results in the least number of blockages,44 but other data do not show any significant differences.45
Achieving Glycemic Control
Glucose monitoring is essential for patients using insulin pumps or basal–prandial regimens. Without this information, there is no good way to make insulin changes. Patients with pumps often monitor their glucose six to eight times a day. Frequent interactions between the patient and diabetes team will be needed to optimize pump or basal–prandial insulin therapy. At these visits the CDE will review glycemic control, the average total daily insulin dose, number of boluses per day, and the basal–bolus insulin ratio. As the insulin requirements are adjusted, the insulin-to-carbohydrate ratio and insulin sensitivity will be recalculated.46 The CDE will focus on the fasting glucose to make decisions on the initial basal insulin infusion rate. Glucose levels monitored one to two hours post-prandially are used to make changes in the bolus doses of insulin and provide clues about the accuracy of carbohydrate counting. Monitoring at other time-points such as before meals, at bedtime, and in the early morning gives further information on basal requirements. Sometimes only one basal infusion rate is needed, but usually two or three rates are used. More than four basal rates should signal that diabetes and nutrition education need to be re-addressed. Remember, basal insulin is needed to manage basal hepatic glucose production, which is fairly constant, and adding additional basal rates compensates for changes in insulin resistance. For example, if early-morning hypoglycemia is detected, the basal rate might decrease by 20% during the first half of the night. In patients exhibiting a dawn phenomenon, the basal rate might increase by 20% during the second half of the night.46
There are some general concepts that should be well-known by patients. An important one is the ratio of basal to bolus insulin usage. The ratio should be somewhere around 1:1. Some patients may require 60% of the insulin in basal form. In pregnancy, the basal needs may be 40%, while the bolus requirements may increase to 60%. However, the point is a patient should not be using twice as much bolus as basal insulin or the reverse. Becoming ‘upsidedown’ on the insulin ratios can easily occur if one narrowly focuses on pre- or post-prandial blood glucose measures and neglects the global overview. A red flag is the deterioration of glycemic control with increasing doses of insulin. Other important management strategies include achieving fasting glycemic control because the fasting glucose level will set the tone for the entire day, checking urine ketones if blood glucose is >300mg/dl in persons with type 1 diabetes, and using a correction dose of insulin to control a random hyperglycemic episode, or adding a correction dose to the prandial insulin if there is pre-meal hyperglycemia. Again, these are strategies that a CDE will review periodically with the patient.
Another very useful tool that people with diabetes need to know and must be able to use is the calculation of an insulin correction factor (insulin sensitivity factor). This calculation applies to people using pumps or multidose basal–prandial regimens and is obtained by dividing 1,800 by the total daily insulin dose. The number 1,800 is based on common experience of managing diabetes with fast analogs. It is not an absolute value because some physicians use 1,700 and others 1,900 (note that 1,500 is commonly used if the patient remains on regular insulin). An example will clarify the utility of the correction factor. Assume that a patient is using 100 units of insulin per day, in which case dividing 1,800 by 100 equals 18. The number ‘18’ is the correction factor and means that one unit of insulin will reduce the glucose by 18mg/dl. The patient can use the correction factor in two ways. If blood glucose is high before a meal, a correction factor dose of insulin is given in addition to the prandial insulin. If the blood glucose goal was <140mg/dl before a meal but the measured blood glucose was 195mg/dl, the patient would conclude that the glucose was 55mg/dl over the goal. If one extra unit of insulin reduces the glucose by 18mg/dl, three units will decrease the glucose by almost 55mg/dl. Thus, the patient will inject the usual prandial insulin dose to cover the meal and three extra units to correct the hyperglycemia. Alternatively, if the patient is not eating but checks the blood glucose and finds it to be 195mg/dl, just three units of insulin would be used to correct the hyperglycemic event.
Most people with diabetes do not achieve tight glycemic control despite the availability of new insulin preparations that can be used to approximate normal insulin physiology. The reasons for this include lack of emphasis on patient education, timely initiation of insulin therapy, and defining the patient’s responsibilities in the management of their medical problems. Healthcare providers can play a very positive role if they believe in tight glycemic control. The study by Belfiglio et al.47 emphasized this point. He demonstrated that patients seen by physicians who believed in achieving fasting glucose goals had HbA1c values near 7%. Patients who were seen by physicians who believed that a fasting glucose of 140mg/dl was acceptable had HbA1c levels of around 8%. Healthcare providers can also play a positive role by providing realistic information on insulin therapy. The role of insulin management should be discussed at the first encounter for people with type 1 or type 2 diabetes.
Obviously, no-one would view insulin injections as a prize to be sought after, but a simple demonstration of insulin administration with a pen device almost always dispels the fears associated with the device. People with type 1 diabetes are relieved and people with type 2 diabetes stop worrying that one day they may need insulin. Too often, insulin therapy is discussed as a punishment and impressions are conveyed that it is the patient’s fault for having diabetes.
So, what is the best way to use insulin therapy? Simple once- or twice-daily regimens may work for some people for a while, but eventually many patients will need physiological management. This is currently achieved by using insulin pumps or a basal–prandial injection regimen. The available evidence seems to point toward pump therapy as the better method for those with type 1 diabetes in terms of HbA1c reduction and decreased episodes of hypoglycemia. Pump therapy or multidose approaches appear to be equally effective in type 2 diabetes, but pumps may afford the same degree of control with fewer glycemic excursions. Thus, two good therapeutic pathways are available and the decision of management will involve the patient’s preferences and clinical judgments on candidate selection. In the end analysis, we need to remember that insulin is ‘the good stuff’ and will be needed to reach glycemic goals.■