What are insulin pumps

The insulin pump is a small computerized medical device, worn continuously by the patient. It delivers insulin subcutaneously almost without interruption and is one of the main treatment options in type 1 diabetes, alongside the multiple daily injection regimen [1]. It can replace multiple daily injections in patients with type 1 diabetes and in some patients with type 2 diabetes. It is especially useful when glycemic control remains difficult to achieve with conventional insulin regimens [2]. The insulin reaches the body through a cannula, a short, soft tube placed under the skin and usually changed every two to three days.

The pump works permanently in two complementary ways, explicitly defined by the current diabetes technology guidelines [2]. On the one hand, it provides a small, constant flow of insulin throughout the day and night, called the basal rate. This basal rate can be adjusted separately for each hour. On the other hand, larger doses are added to the basal rate, given at meals or to correct hyperglycemia, called boluses [1]. Thus, a single cannula practically replaces all the pricks required by classic treatment with multiple daily injections, offering an important gain in treatment flexibility.

Insulin pumps used today come in two main types, with quite different designs; there is also an implantable version, which is rarely used. The tubed pump is a small computerized box, roughly the size of a deck of playing cards. It generally has a screen that helps with programming. It connects through a long, thin tube (catheter) to the cannula inserted under the skin and secured with an adhesive. The patch pump has no external tube and no screen of its own. It is a small, rounded device that sticks directly onto the skin. It contains the insulin reservoir, a short internal (hidden) tube, and the cannula [3].

Regardless of the type, modern pumps are small, light, and designed to be discreet under clothing. The tubed pump can be clipped to a belt, kept in a pocket, or fixed to clothing. The patch pump is controlled remotely, through a dedicated handset or a phone app, with no visible wires [3]. Many patch models also have a water-resistant housing, tested to specific sealing standards, which makes them convenient for showering or swimming without disconnection [3].

The pump is worn permanently, day and night, because it contains only rapid-acting insulin and must ensure a continuous flow. If the flow is interrupted, blood glucose rises quickly and ketone bodies (acidic substances produced when insulin is lacking) can appear, with a risk of ketoacidosis. The body of the tubed pump is worn away from the skin — clipped to a belt, in a pocket, or fixed to clothing. The cannula through which the insulin enters is usually placed on the abdomen or buttock, but also in other areas. The connection between the pump and the cannula is provided by the catheter. The patch pump sticks directly onto the skin, on the abdomen, arms, thigh, hip, or buttock [4].

It is recommended to change the cannula site every two to three days and to rotate the areas used systematically. This keeps the skin healthy and reduces the risk of lipohypertrophy (a local thickening of the fatty tissue under the skin that appears when the same site is used repeatedly) or of uneven insulin absorption [4]. Most patch pumps are water-resistant, being tested to specific sealing standards, so they can be worn while showering or swimming [3]. Tubed pumps can be temporarily disconnected for such activities. After a disconnection, it is important to reconnect the pump as soon as possible, in order to limit the period without basal insulin supply. The pause should not exceed one hour. If it does, measure your blood glucose, check for ketones, and give yourself the missing insulin according to the protocol agreed with your medical team.

An ordinary, stand-alone insulin pump does not measure your blood glucose. Its role is to deliver insulin, not to determine the blood glucose value. That is why you separately need a glucose meter or a continuous glucose monitoring sensor (CGM). The sensor actually measures the glucose in the interstitial fluid (the fluid between the cells) and uses that to estimate the glucose in your blood. The value displayed can differ slightly from the glucose meter reading, especially when blood glucose is changing rapidly.

There are also automated systems, in which the pump is wirelessly connected to a sensor that measures the glucose concentration in the interstitial fluid, under the skin. In this case, the pump uses the values sent by the sensor to adjust its insulin delivery. The measurement itself, however, remains the sole task of the sensor, not of the pump. Therefore, whether you use a simple pump or an integrated automated system, blood glucose monitoring remains essential for the safety of the treatment.

No, an insulin pump does not replace the pancreas. It takes over only one of its tasks, namely the delivery of insulin. The pancreas, however, performs many other functions, such as producing digestive enzymes and other hormones that regulate blood glucose (glucagon), and no pump can substitute for them [5].

Even the most advanced automated systems, popularly called the "artificial pancreas," reproduce a single function of the pancreas, insulin delivery, though they perform it better and better. They do not produce the other hormones and still require you to manually announce meals, being unable to fully replace natural physiological regulation [5]. The pump therefore reproduces, imperfectly, the role of pancreatic insulin secretion, but not the entire pancreas, as the figure below shows.

A simple pump, without automation, does not decide the insulin doses on its own. You program the basal rate, and at meals you enter the amount of carbohydrates and the blood glucose value. On this basis, the pump calculates the recommended bolus (or you calculate it yourself). In practice, the pump faithfully follows the settings established together with your medical team [2]. Hybrid closed-loop systems go a step further and can automatically adjust part of the insulin.

They increase, reduce, or temporarily stop the basal rate according to the data received from the glucose sensor. Most can also add small automatic corrections. Studies have been carried out in children who switched from simple pumps or multiple injections to such systems. They show an improvement in glycated hemoglobin (HbA1c, the average blood glucose over the past few months) and in the time spent in the glycemic target, with a reduction in glycemic variability [6]. Time in target is the percentage of the day when blood glucose stays within the recommended range. Even so, you still have to announce meals manually, so the pump helps you through partial automation, without being completely independent.

Insulin pumps have been used for almost fifty years. The first models that could actually be worn by patients appeared in the 1970s. An early prototype, the size of a backpack, existed as far back as the 1960s and was initially used only for clinical research [7]. These early devices were large, cumbersome, and often associated with complications, being very different from today's discreet devices.

Over the decades, pumps have become ever smaller, safer, and easier to use. Bolus calculators appeared gradually, along with the possibility of multiple basal rates for different times of day. More recently came the link with glucose sensors and the automatic delivery of insulin through hybrid closed-loop systems [7]. Thus, a technology that was initially bulky and reserved for research has become a discreet, safe, and intelligent device, integrated into current clinical practice.