The Review of Insulin Pens—Past, Present, and Look to the Future

Currently, there are about 150–200 million diabetic patients treated with insulin globally. The year 2021 is special because the 100th anniversary of the insulin discovery is being celebrated. It is a good occasion to sum up the insulin pen technology invention and improvement which are nowadays the leading mode of an insulin delivery. Even though so many years have passed, insulin is still administered subcutaneously, that is why devices to deliver it are of great importance. Insulin pens have evolved only through the last decades (the reusable, durable pens, and the disposable, prefilled pens) and modern smart insulin pens have been developed in the last few years, and both types of the devices compared to traditional syringes and vials are more convenient, discrete in use, have better dosing accuracy, and improve adherence. In this review, we will focus on the history of insulin pens and their improvement over the previous decades.


INTRODUCTION
The International Diabetes Federation (IDF) estimates that over 537 million people all over the world are currently struggling with diabetes mellitus (DM) (1) and there are about 150-200 million of them treated with insulin (2). The history of insulin dates back to the last century, when in 1921 Frederick Banting and Charles Best with the support of John Macleod and James Collip discovered insulin and thereby revolutionized the treatment of DM (3)(4)(5). The first injection of insulin on January 11, 1922, to a 14-yearold boy with the use of reusable glass-bodied syringes (6) started an entirely new era of diabetes management (4,5) and led to the improvement of insulin delivery methods (3,5). Even though insulin has been used for 100 years already, its administration remains subcutaneous where insulin pens which evolved only through the last four decades are the leading method of its delivery (about 60% of patients treated with insulin use insulin pens all over the world) (7)(8)(9). Insulin pen utility is not the same in different regions of the world. According to a report from the year 2008, insulin pens were used by only 15% of patients in the US, compared with 80%-90% in Europe, and it was suspected that it could be due to limited education regarding the benefits of insulin pens but also their higher price (10). The situation has changed in the next years where data from the year 2011 indicate that the number of patients initiating vial/syringe in the US decreased from 2005 to 2011 to approximately 30% while patients initiating pens increased to approximately 60% (11). According to a IQVIA ® report for the period from June 2020 till June 2021 prepared for the purpose of this manuscript [data not published (12)], the usage of pens in US rose to 59% where in Europe it is comparably high and assessed to be 93.6%. Insulin pens have numerous advantages over traditional vial and syringe injections, among others easy use especially for patients with vision problems or manual dexterity, accuracy of delivering small doses of insulin, and discretion of use (13). It is worth noting that aspects of insulin administration may also contribute to the treatment outcomes even though the type of insulin and its efficacy and safety are the primary factors to consider. It is important to underline that each insulin-producing company has its own insulin pen dedicated to use with the produced insulin. It was proved in some studies that patients who use insulin pens are more adherent to the treatment regimen and have less hypoglycemic events compared to insulin vial users (14)(15)(16)(17)(18). Also, numerous studies report that patients' preference for insulin pens exceeds that for vials or syringes (19)(20)(21) and portability of insulin pens improves patients' convenience (22). However, it is important to note that the superiority of insulin pens in achieving and maintaining glycemic control has been questioned, and this question has not been resolved up to day (23). American and European guidelines underline the necessity of undertaking patient preference when selecting diabetes treatment especially when treatment is accoutered with pain due to injection (24). That is why recently a study assessing the patient perspective of injectable treatment among patients with type 2 diabetes (T2DM) has been performed and showed that there are some features of the injection device that patients choose more often which may help in future improvement of insulin pens (25). Development of insulin pens is parallel to the development of newer insulin formulation where insulin pen must adapt to changes to dosing and timing requirements like it is in case of modern ultra-long-acting insulin analogue icodec, administered once weekly, which is under development (26). This year, the discovery of insulin turns 100 years, and this provides an opportunity to reflect on its administration methods over the past years.
In this review, we will focus on the history of insulin pens and their improvement over the previous decades, starting from the first-generation insulin pens throughout modern smart insulin pens ( Figure 1). It must be noticed that clinical trials in relation to the newest smart insulin pens and insulin pen caps are very limited to date, that is why information related to this new technology comes also from manufacturer websites and commercial data resources.

FIRST-GENERATION INSULIN PENS
The first insulin injections were made with large and heavy glass or metal syringes and reusable needles (4,7,(24)(25)(26). Syringe was the only possible way of delivering insulin in clinical practice for the next several decades (4,7,27). This method of administration had several and serious disadvantages including poor dose accuracy, lack of social acceptance, and fear of injections (7,27,28). These inconveniences of the vial and syringe led to the manufacture of insulin pens. Majority of insulin pens are proprietary devices and are developed to work with specific insulin from the same manufacturer (29). Insulin pens are classified into two categories: being reusable (durable) or prefilled (disposable). The reusable insulin pen is loaded by the patient with replaceable insulin cartridges, and the prefilled insulin pen has the insulin reservoir cartridge already installed and the pen is discarded when the cartridge is empty. Both types of insulin pens can contain a maximum of 3 ml of insulin (30) and can deliver insulin in 0.5-, 1-, or 2-unit (U) increments up to 160 U with the use of a needle which has to be attached to the insulin pen.

REUSABLE INSULIN PENS
In 1985, Novo Nordisk has launched the first reusable insulin pen injector called NovoPen ® to overcome barriers of the vial and syringe (31) and started a series of NovoPen ® insulin injectors. The new device was a combination of the syringe and insulin vial in one mechanism, resembling a fountain pen (31). NovoPen ® contained a disposable, replaceable 1.5-ml insulin cartridge connected with a single-use needle and one-unit incremental dosing (29,30) which was ready to use whenever needed. This allowed patients to administer multiple, preprandial injections discreetly, and their daily schedule became more flexible (32)(33)(34). First studies related to insulin pen comprised only several patients in 1995 (31), but as the development of the devices has grown up, also the number of patients studied increased to several hundreds per study in 2002 (35) and up to several thousands in 2020 (27,36). Initially, insulin cartridges dedicated to insulin pen contained short-acting insulin for numerous injections before meals and basal insulin was injected with conventional syringes (37). Soon after, in 1988 a new insulin pen NovoPen ® 2 was presented to administer NPH and premixed insulins (38)(39)(40). Analogically as with short-acting insulins, majority of patients using the device to administer basal or mixed insulin preferred to continue the therapy with pens (38)(39)(40). In 1992, NovoPen ® 3 was launched which had a maximum dose that could be administered at one time which increased to 70 U (from 36 U with NovoPen ® 2) and the dialed doses could be reset without insulin waste. Soon after, in 1996 NovoPen ® 1.5 was released which had a smaller insulin cartridge and was shorter in length, followed by NovoPen ® 3 Demi to administer 0.5 U dose increments in 1999 and NovoPen ® Junior in 2003 which was designed with vibrant colors and developed specifically for children with diabetes. In 2005, NovoPen ® 4 was introduced which required reduced force to perform an injection, which had dose increments of 1.0 U and a maximum dose of 60 U (41). Moreover, NovoPen ® 4 was reported as simpler to learn and easier to use for both insulin-naïve and currently using NovoPen ® 3 patients (42). Following the release of NovoPen ® s, other manufacturers have also introduced reusable insulin pens, including the HumaPen ® range (Eli Lilly and Company, Indianapolis, IN, USA) and the OptiPen ® Pro, OptiClik ® , and ClikSTAR ® pens (Sanofi, Bridgewater, NJ, USA) The inconvenience of the first insulin pens was no possibility of dialing backward without wasting insulin, but the thing changed with the introduction of NovoPen ® 3 and HumaPen Ergo ® (35,41). This option translated to device acceptability in comparison with previous generations of insulin injectors and syringes (43). With time, the option of insulin-free dialing forward and backward became a prevailing way of setting the insulin doses. All mentioned insulin pens had the trigger placed at on the opposite site of the needle attach end, but there are also insulin pens with a side-mounted release button used for half-automatic insulin delivery, first developed in AutoPen (44), and this mechanism was also present later on in 2010 in GensuPen ® and in 2017 in GensuPen ® 2 insulin pens (27,45). Such a mechanism ensured patients about proper insulin administration, simplified the way of injection, and was convenient for elderly patients (27). Moreover, it was proven that the GensuPen ® 2 injector in comparison to NovoPen ® 4 (Novo Nordisk, Bagsvaerd, Denmark) and HumaPen Ergo ® (Eli Lilly, Indianapolis, IN) requires reduced force for insulin administration, especially at high doses of the drug (46). In recent years, further improvement in insulin pen function has been made and there are several ones which possess the memory function of the last dose taken. In 2007, Eli Lilly released the world's first digital insulin pen with memory function, namely, HumaPen Memoir (47). Soon after, in 2010, Novo Nordisk launched NovoPen ® Echo (48), the first insulin pen with memory function and half-unit dosing feature. Most of the insulin pens available in the market have the feature to deliver insulin in 1-unit increments, and only a few deliver in half-units. 0.5-increment insulin pens are designed for patients who need small insulin doses, and the available ones are HumaPen Luxura HD, Humalog ® Junior KwikPen ® , NovoPen ® Demi, Junior, Echo, JuniorSTAR ® , and InPen ™ . Based on the trials' outcomes, children, adolescents, and their parents appreciated both the memory function and simplicity of junior devices (49,50).
Cited studies related to reusable insulin pens are summarized in Table 1, and the technical characteristics of reusable insulin pens are presented in Table 2.   6-month randomized, control, crossover trial. All participants were using MIR of soluble insulin (Velosulin Human) from pen injector (Insuject) with a single injection of NPH insulin (Insulatard Human) from the conventional syringe before the study. The first group was continuing multiple injections with the pen injector and NPH insulin from the conventional syringe in the first 3 months of the trial. In the next study period, the group started to administer NPH insulin (Ultratard Human) in the pen injector Insuject-X. The second study group started the trial in the reverse order.
No differences in the metabolic control were found between both study groups. In the final questionnaires, 86% of the patients found the NPH pen injector less complicated to use than usual syringes. All but 2 patients wished to continue using Insuject-X in the future.   Patients administering insulin at least 3 months prior to study entry were asked to answer the questionnaire to assess the level of satisfaction with their current delivery device. Next, participants were instructed how to use HumaPen ® Ergo and started injecting insulin in their previous regimen with the new injector. After 5-7 weeks, in the end of the study, patients were asked to answer the questionnaire regarding the acceptability of HumaPen ® Ergo, compared with their previous devices.
>70% of both syringe and NovoPen ® 3 users rated HumaPen ® Ergo as easy to use in all aspects. The main advantages of the new device were ease of holding during injection, possibility of correcting the doses and the procedure of cartridge changing. At the end of the study, 74% of syringe users and 72% of previous injector users decided to continue administering insulin with HumaPen ® Ergo. Ristic  Participants who were using another injector before the study started the insulin administration with HumaPen ® Ergo. The visits took place in the beginning of the study, after the next 3 weeks, and again in the 7th week of the study. The acceptability of the HumaPen ® Ergo was evaluated with a questionnaire in the end of the trial. The HCPs assessed the pen injector with the same criteria as the patients.
Participants considered HumaPen ® Ergo as easy/very easy in learning to use (97%), reading the dose (95%), correcting the dose (97%), and holding during injection (62%). Most of patients (Study 1/2: 89%/93%) found the pen easier/much easier to correct the dose than the previously used injector. 60%/ 69% of the study group would continue using HumaPen ® Ergo and recommend the model to the others HCPs and would recommend the injector because of the ease in dialing back with no insulin waste (80%) and reading the dose (74% Overall preference for the IIPD was higher than that for vial and syringes among both groups (insulin users and nonusers), mainly because of social acceptability. However, current insulin users claimed that social acceptability and ease of use were the most significant predictors of preference vial and syringes. For insulin non-users, these preference predictors were activity interference and also ease of use.  6-month observational study. Patients were treated with NPH insulin for at least 6 months before the study. In the beginning of the trial, they were transferred to glargine insulin administered with OptiPen ® Pro-1. After 6 months of observations, the patients were asked to complete an inquiry form and rate the OptiPen ® Pro-1 on a scale (0 = worst, 5 = best).
Participants were asked to assess the usability of the device they were using before and the NovoPen ® Echo. Firstly, they completed specially designed tasks (setting up the pen, adjusting and injecting a dose, operating the memory function and subjective assessment). Afterward, participants filled up rating scales (1 = most favorable; 6 = least favorable) to rank each pen. Multicenter, open-label, crossover study.
In the first step of the study, the group of patients currently using NovoPen ® 3 were asked to handle NovoPen ® 4 and complete a sequence of tasks within 5 min. The second, crossover part of the trial concerned both groups of patients (NovoPen ® 3 users and insulinnaïve patients). The first half of every group received a time-recorded training about using NovoPen ® 3 before completing a series of tasks. In the end of the tasks, the patients were asked to evaluate handling the device in a questionnaire. In the second step, the participants completed the same sequence of tasks with another device-NovoPen ® 4. The other half of the study groups assessed the injectors in the reverse order.  Participants were asked to assess the JuniorSTAR ® pen injector on 3 five-point scales: -when rating the product: 1 = very poor; 5 = very good or 1 = very difficult; 5 = very easy, -when asked to agree/disagree: 1 = completely disagree; 5 = completely agree. Positive response means a percentage of either a 4 or a 5 score.
practicality, ease of carrying (84%), ease of reading the dose (96%), ease of dialing back (87%), and a suitable injection force (87%). When the respondents were asked to describe the pen in one word, the most common replies were as follows: practical, easy, and simple. Randomized, open-label, crossover study. Patients received basal insulin (glargine) in one of two treatment sequences (2 weeks of using pen followed by 2 weeks of using vial and syringe or vice versa). Patient device preference was evaluated by the Insulin Injection Preference Questionnaire in the first end point (at week 4-the end of the crossover period). Then, patient preference and HCP recommendation were assessed with one global item and 3 others (blood glucose control, reluctance to use insulin, long-term insulin use) using a 5-point scale (1 = not preferred, 5 = preferred/recommended). Next, patients were re-randomized to pen or vial and syringe group for further observation (6, 10, and 30 weeks) to evaluate clinical end-points (HbA1c, fasting blood glucose levels) and safety outcomes (hypoglycemia, adverse events).  (87), and Junior KwikPen ® launched in 2017, a half-unit insulin pen (88). Similarly to reusable insulin pens, prefilled ones when compared to vials and syringes were rated as much easier to handle, discreet in public use, confident in proper dose delivery, and preferred by majority of patients (with T1DM and T2DM), healthcare professionals (89)(90)(91), and patients' caregivers (parents, relatives) (92). Moreover, both non-experienced healthcare practitioners and needle-naïve patients found the prefilled insulin pens much easier to teach and learn (93,94).  For years, insulin pens were used with insulin 100 U/ml, but since the development of higher-concentration insulins, also new insulin pens for 200 and 300 U/ml have been manufactured and used since 2017, namely, Humalog ® 200 U/ml KwikPen ® (Eli Lilly) (95), Tresiba ® 200 U/ml prefilled FlexTouch ® (Novo Nordisk) (96), and Glargine U300 SoloSTAR ® insulin pen (Sanofi-Aventis) (97). However, we must consider that disposable pens are less environment friendly and this is a globally growing importance nowadays (98). One can just imagine that if a patient is using approximately 40 units of insulin a day there is about 50 prefilled plastic pens thrown away every year and accounting for thousands of patients using insulin pens the number of insulin pens being thrown away per year is accounted in millions. Based just on a small study form Bosnia and Herzegovina published in 2020, it was predicted that only in this small country there were 3.2 million pens used and dispensed annually (99).
Cited studies related to prefilled insulin pens are summarized in Table 3, and the technical characteristics of prefilled pens are presented in Table 4. Multicenter, observational study The trial consisted of 1-hour face-to-face interviews aimed at evaluating the usability of the devices and patients' preferences. Firstly, participants were asked to prepare the device and deliver a 40-unit dose relying on their intuition and/or relevant manuals. Any training and maintenance was not provided. Next, respondents evaluated abovementioned procedures for each pen in a five-point scale (1 = poor, 5 = excellent).  Observational study Non-experienced participants were educated how to use the injectors. All the respondents made 2 injections with 5 examined devices. The first one was a self-injection in the prosthetic skin attached in the respondents' flank. The second injection was made to the prosthetic skin placed in the upper arm of the mock patient (other injection). Every injection contained 10 units of saline. In the end, both self-and other injections with every device were evaluated in the questionnaire and rated from 1 to 5.
FlexPen ® was rated as the best device for selfinjections. However, FlexPen ® was also selected the worst one for the other-injections because it was too long, was less stable, and had inadequate visibility of the dial. OptiClik ® was evaluated as the best device for other injection but the second worst one to selfinjection. 1-day, randomized, crossover study. Respondents were randomly assigned to one of the groups (starting the study with FlexTouch ® or KwikPen ® ) and then crossover to test the second pen. Participants were trained how to use the devices before the test injections. Next, both patients and HCPs made multiple injections (with randomly altered doses including 20, 40, and 60 U) into a foam cushion and answered questions concerning ease of use, confidence, and preferences.
FlexTouch ® (compared to KwikPen ® ) was rated as most preferred device (86% vs. 7%; p < 0.001), easier to use (85% vs. 4%; p < 0.001), and recommended to others (88% vs. 6%; p < 0.001). Additionally, FlexTouch ® was characterized as the better device in the injections for ease of depressing the push button and ease of injecting the doses (p < 0.001 for all). FlexTouch ® was found as the most confident in correcting and completing insulin delivery (73% vs. 6%; p < 0.001). Hancu   1-day multicenter, open-label, randomized, crossover study. Respondents were randomly assigned into the study groups (starting test with FlexTouch ® or SoloSTAR ® ). Participants were instructed how to use the pen and performed test injections into a foam cushion, dosing 20, 40, and 80 U. In the following step, both study groups were crossed over to test another pen device. Each pen device was assessed separately (in a form evaluating handling and operation of the pen). Moreover, in the final questionnaire respondents completed regarding their preferences.
Patients' persistence was defined as the lack of gaps in insulin therapy during the follow-up period. 1-day randomized, multicenter, open-label, crossover study. Participants were trained how to use the devices. Next, test injections into foam cushion (dosing 20, 55, and 80 U) were made with both vial and syringe and FlexTouch ® . Then, respondents separately rated the devices in respect of ease and confidence of use.
FlexTouch ® (compared to vial and syringe) was found a preferred device (88% vs. 5%; p < 0.001), easier to use (91% vs. 6%; p < 0.001), and willingly recommended (91% vs. 3%; p < 0.001). Moreover, participants considered FlexTouch ® easier to use, more stable during injection, and better in depressing the push-button and reading the dose scale (all p < 0.001). Patients and HCPs using FlexTouch ® were also more confident in properly insulin delivery and metabolic control than the ones using vial and syringe (p < 0.001). Lajara  1-day randomized, multicenter, open-label, crossover study. All participants received an instruction on how to use the injection device. Then they were asked to make a test injection into a foam cushion (dosing 20, 55, and 80 U) with FlexTouch ® or vial and syringe (in a random order) and answer questions on confidence and ease of use (1 = very difficult/not at all confident; 5 = very easy/ very confident). In the next step, respondents followed the abovementioned procedures with another device. Finally, all participants completed a preference questionnaire to evaluate both methods.
Both HCPs (nurses: 100% vs. 0%; physicians 87% vs. 7%), needle-naïve (83% vs. 7%), and vial-and syringe-experienced (73% vs. 7%) patients preferred FlexTouch ® over vial and syringe for ease of teaching. Moreover, the insulin pen was rated as very/fairly easy for depressing the push-button (physicians: 93% vs. 80%; nurses: 97% vs. 80%; vial and syringe-experienced patients: 93% vs. 90% and needle-naïve ones: 100% vs. 77%).  with T1DM or  T2DM and 61  HCPs) 1-day, randomized, crossover study. The study consisted of 2 comparison groups: FT to KP and FT to SS. Participants were randomized to start the trial with FT or another pen device and then were crossed over to test the second injector. All respondents were educated about using the devices. Both patients and HCPs were asked to make multiple injections of different doses with each pen (FT vs. KP study: 20, 40, 60 U; FT vs. SS study: 20, 40, 80 U). In the end, participants answered the questions regarding ease of use, learning and teaching, confidence in use, and preference.
FlexTouch ® was rated as very/fairly easy to inject, particularly in the maximum dose (compared to KP or SS: ≥80% vs. ≤38% and ≤23%) and very/rather confident in the ability to manage daily injections. FT was also considered as easier to teach and learn to use than KP and SS (all p < 0.001) and preferred for learning and teaching (≥39% vs. ≤4% for KP and ≤6% for SS). Most of the patients and HCPs would recommend FT (≥95%) than KP (≤72%) and SS (≤71%).

(Continued)
Masierek et al.  Patients became stratified into 4 study groups: A-visually impaired with T1DM and impaired dexterity; B-visually impaired with T2DM and impaired dexterity; C-visually impaired with T1DM or T2DM; D-patients without any impairment with T1DM or T2DM.
Participants were asked to perform some test injections (dosing 10, 30, and 50 U) and complete a standardized questionnaire assessing the handling of the pen device. The procedure was repeated with a second insulin injector. In the end, patients evaluated the study by completing a comparative questionnaire.
FlexTouch ® was preferred in all study groups including 100% of group D (unimpaired patients).
Only a few patients with visual/dexterity impairment preferred InnoLet ® (group A-13%, group B-3%, group C-14%). 1-day single-center, randomized, crossover study. Participants (in random order) were asked to perform testing injections into laboratory tubes (doses of 5, 25, 43, and 79 U) with the devices. Dosing accuracy was measured, and patients completed final questionnaires (device assessment questionnaire, patient perception questionnaire). Next, respondents were crossed over to test another device. At the end of the trial, all participants answered the questions in the device preference questionnaire.
FlexTouch ® presented significantly better dosing accuracy when used by all cohorts and at all doses (p < 0.005 for all doses). The pen injector was rated significantly higher than vial and syringe in both device preference questionnaire (93% vs. 2% for vial and syringe; p < 0.001)) and patient perception questionnaire (in all aspects). The study consisted of one visit. Participants were asked to make 4-6 injections into a foam cushion (dosing 2, 20, 40, 80, 120, and 160 units). Next, they were asked to complete a questionnaire to evaluate the device. These procedures were repeated in each of the three analyzed injectors. After the tests, participants answered the final, overall questions.
Significant majority of participants preferred to use FlexTouch ® U100 (93.0%) and U200 (91.4%), even dexterity-impaired and pen-naïve patients in comparison with SoloSTAR ® (p < 0.001), respectively. Patients were asked to dial up from zero to maximum and next, dial down from maximum to zero with each pen. Dialing up and down was recorded with a video, and the torque of the devices was analyzed. Next, 16 pen-experienced people with T2DM rated the subjective comfort for each insulin injector after dialing up and down again.
SS was rated as most comfortable in dialing up by 8 and dialing down by 6 of the 16 respondents; analogically, FP was ranked by 5 and 8, respectively; FT1: 2 and 1; KP: 1 and 1. FT2 was evaluated as least comfortable by 12 and 10 patients. Comfort of up-and down-dialing was considered "very comfortable" for SS by 15 patients each and next, FP (12 and 14), KP (10 each), and FT1 (9 and 7). FT2 was ranked "less/not comfortable" by 10

NEXT-GENERATION INSULIN PENS (SMART INSULIN PENS) AND INSULIN PEN CAPS
Nowadays, the way of health delivery is becoming more digital than ever before where face-to-face visits are often replaced by telephone or video contacts and continuous glucose monitoring or glucometer data can be revived through cloud-based data sharing technology which was very pronounced in the COVID-19 era. One of the key problems for patients with T1DM and T2DM treated with multiple daily insulin (MDI) is omitting or late insulin doses which has been found in the study which analyzed data from a continuous glucose monitoring system (CGM) (113). It was also described lately in the study with a Bluetooth ® -enabled insulin pen cap that all of the patients taking part in the study missed the insulin doses and it could be intentionally missed because of inconvenience or eating pattern or just forgotten (113). It is important to note that it was also calculated already a years ago that omitting only two meal-related insulin doses per week is associated with a 0.4% increase in HbA1c value (114). Another problem with MDI is that patients rely on numeracy skills while deciding about the meal insulin dose, and it has been proven that these skills are many times not good enough which leads to errors in insulin dosing and to poor glycemic control (115)(116)(117). Because patients treated with MDI have to make their insulin dosing decisions without access to the amount and timing of previous insulin doses or residual active insulin, this can, on the other hand, cause overlapping of insulin boluses and put a patient at risk of hypoglycemia (118). That is why smart insulin pens and pen caps were and are being developed to overcome these barriers. Information coming from business research indicates that the smart insulin pen market value will significantly increase by the year 2027 in Latin America, the Middle East, and Africa (119) with the greatest market growth in Europe with a trend toward increased use of smart insulin pen market seen also in North America (120).

SMART INSULIN PENS
Smart insulin pens are digital, connected insulin pens which go beyond memory function and automatically transmit information about time and amount of insulin administered to the user's mobile device and can remind about the insulin dose and help to calculate the bolus (7). The clinical data from the smart insulin pen are transferred wirelessly via Bluetooth ® technology to an application (app) available for smartphones (7,121,122). Therefore, smart insulin pens require the use of an app to collect the data sent from the pen but eliminate the need for manual self-report logbooks (121). Thus, smart insulin pens can help to overcome the challenges that users of pen injectors have to deal with on a daily basis. Smart insulin pens are a relatively new invention, so it should come as no surprise that a few studies have been conducted in this field to date (121). In 2017, the world's first US Food and Drug Administration (FDA)-approved insulin smart pen which uses Bluetooth ® technology, namely, InPen ™ (Companion Medical, San Diego, Ca, USA), was launched, and in November 2020 its new version was launched by Medtronic (123). This pen combines the insulin pen with a smartphone app which has the ability to record and store data of insulin injections and recommend doses, as well as display glycemia and related data on the paired smartphone app (124)(125)(126). InPen ™ is designed for use with rapid-acting insulin U-100 Lilly Humalog ® and Novo Nordisk NovoLog ® (127). InPen ™ is the first of its kind of smart insulin pen that allows to prepare reports for healthcare professionals, reminds about missed doses, and tracks insulin on board, but also alerts the user about an exposure of the device to abnormal (very high or very low) temperatures that may inactivate insulin (124)(125)(126)128). What is likewise important, in InPen ™ the dose can be increased or decreased in half-unit steps, and therefore the dose administered is very precise (128, 129). Later on, several new smart insulin pens emerged on the market, namely, ESYSTA ® pens (Emperra), Pendiq 2.0 pens (Pendiq), and NovoPen ® 6 (Novo Nordisk). It cannot escape the attention that insulin pen injectors may help not only patients but also diabetes care teams. They provide accurate information about missed doses as well as injection times in relation to meals and dose sizes, which is useful in making correct therapeutic decisions and giving personalized treatment plans (121,(130)(131)(132). The first study of clinical outcomes using a smart insulin pen was reported in 2020 (133). This investigation was conducted in Sweden and indicated that among patients with T1DM using smart insulin pens, clinical outcomes improved at lower costs compared to standard care. What is even more important, this research suggested that smart insulin pens have the potential to improve glycemic control and decrease glucose variability (133,134).

INSULIN PEN CAPS
Insulin pen caps are another device which does not have a clear definition but displays the quantity of insulin in the pen and integrate the insulin-related information with a mobile app. Insulin pen caps are usually attached to the side or fit in the end of the pen.
A first-of-its-kind smart pen cap for insulin pens (Bigfoot Unity ™ Diabetes Management System) launched by Bigfoot Biomedical received FDA clearance in May 2021. This insulin pen cap is integrated with Abbott's FreeStyle Libre 2 system and translated continuous monitored glucose data into on-demand insulin dose recommendations displayed on the pen cap screen. It is the first and only device which integrates a continuous glucose monitoring system (CGMS) to insulin dose recommendation (135).
Another smart cap integrated with a dedicated mobile app is GoCap (Common Sensing company) (136). The integration with the application helps calculate the meal or correct boluses, preventing overdosing by active insulin display (125,136). Moreover, individual reminders allow to keep the schedule of basal insulin (136). Similarly, Insulclock ® is an electronic device attached into the insulin pen and connected with a smartphone app and has an insulin reminder system to reduce insulin omissions (137); this device helps to improve glycemic control and reduce glycemic variability with improved adherence in a recent pilot, randomized study among T1DM (138) and among T2DM patients (139). Another two devices do not connect with any mobile app but present an interactive display (Timesulin ® ) or flash diode (Dukada ® Trio), which define the time of last insulin injection (140, 141). The GoCap device received FDA approval (125). Clinical trials which compare different insulin pen caps are not available yet.
Cited studies related to smart insulin pens and their technical characteristics are summarized in Tables 5 and 6. As for the studies related to insulin pen caps and thier technical details the summery is provided in Tables 7 and 8, accordingly.

CONCLUSIONS
Insulin remains the primary medication in the treatment of T1DM and is often used therapy in T2DM. The methods and

FUNDING
There is no external funding for this review. Bioton S.A. company funded the publication fee only.