Innovations in Diabetes Device Training: A Scoping Review

Abstract

Objective

The coronavirus disease 2019 pandemic highlighted a pre-existing need for alternatives to traditional in-person diabetes device trainings. Barriers to care, which include the heavy burden of training, pose a threat to optimal adoption and utilization of these devices. We searched the literature for alternative methods of training, evaluated user satisfaction, and compared short-term clinical outcomes with guideline-based glucometric targets and historical training results.

Methods

A scoping review of Embase articles from 2019 to 2021 was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines using key words relevant to diabetes technologies. Original full-text articles investigating training of new users on devices were included. Titles and abstracts were screened for eligibility by 2 independent reviewers, and results were summarized.

Results

Of 25 articles retrieved from the database, 11 met the criteria. Alternative training strategies included video conferencing, phone calls, mobile applications, and hybrids with traditional trainings. Overall, there was a high degree of user satisfaction with virtual visits, with a preference for hybrid approaches (6 articles). Although glucometrics varied between articles, short-term glucometrics were satisfactory overall (8 articles), including improved glycated hemoglobin measurements and time in range. Two articles compared time in range over various time points after traditional and remote training. One found equivalency, and the other identified a 5% improvement with remote training.

Conclusion

Alternative training approaches are a viable option to reduce the barriers to care and to alleviate training burden. Intentional implementation of alternatives should be considered a solution to address current barriers.

Highlights

• •Digital device trainings can reduce barriers to optimal utilization of technology
• •Time in range and glycated hemoglobin levels were comparable between virtual and traditional trainings
• •Alternative trainings yielded high user satisfaction
• •Future training models must include digital solutions

Clinical Relevance

Patient training can present significant barriers to clinical workflow. This review identified alternatives to traditional, face-to-face training methods that showed equivocal short-term clinical and user satisfaction outcomes. Providing hybrid options for device training may decrease perceived barriers to technology adoption and improve utilization for a more diverse population.

Introduction

The coronavirus disease 2019 (COVID-19) pandemic raised awareness about existing challenges with in-person care and training approaches for people with diabetes (PwD) and health care providers (HCPs). The pandemic required HCPs to deliver care, monitor glucometrics, and treat PwD remotely to avoid the risk of infection during face-to-face visits.^{1 , 2} Some pre-existing barriers to treatment and training for PwD, such as potential loss of income and/or employment to attend face-to-face clinic visits and concerns about obtaining diabetes supplies, needed to be addressed.3, 4, 5 The pandemic also motivated more members of the health care community to explore, implement, and analyze alternatives to the traditional face-to-face training method for the use of insulin delivery and glucose monitoring devices for PwD. These alternative methods included virtual visits or self-led training modules.

Alternative methods of training may address some barriers to device adoption and utilization, such as geographical limitations to access to care, the need for time off work to attend face-to-face training, and the need to arrange for transportation.⁶ Initial training on pumps alone may take approximately 3 hours, with the expectation that ongoing support and follow-up are needed to ensure optimal utilization.⁷ Additional trainings for continuous glucose monitors (CGMs) or upgrading to automated insulin delivery systems may cause even more strain for PwD and HCPs.⁸ These barriers and other factors have created a health disparity in device technology prescriptions and utilization.⁹

Even with the potential benefits of alternative training methods, questions regarding the effectiveness of these training methods with regard to clinical outcomes, patient satisfaction with virtual training, and confidence in alternative methods of device training among HCPs and PwD may delay the adoption of alternative methods.¹⁰ This scoping review presents an overview of publications on alternative methods of device training and reports on short-term clinical outcomes, user satisfaction, and confidence of HCPs or PwD with regard to those alternative methods. We designed this review to evaluate the following research questions: (1) What alternative methods are currently used in place of or alongside traditional face-to-face training?, (2) Do alternative methods of device training yield equivalent outcomes (both clinical and user satisfaction outcomes) to those seen with in-person training?, and (3) What are the key issues to consider for the implementation of alternative methods of device training?

Materials and Methods

A scoping review of the literature¹¹ was conducted on alternative methods of training on diabetes devices, including pumps, CGMs, and automated insulin delivery systems, per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines.¹² This was a review of previously published articles; therefore, no ethics board approval was required.

Data Sources and Search Terms

Technology & Patent Research International, Inc, was contracted to conduct the search using OvidSP (Embase). Free-text key words were combined with controlled vocabularies (Medical Subject Headings and Emtree) to identify references dealing with the various training methods used to train/educate PwD on the use and maintenance of diabetes devices. The search was limited to English-language articles published from 2019 through 2021. Search terms included “diabetes” or “insulin,” “care”/“education”/“training,” “technology,” “pump,” and “infusion system.” The code used to search for articles and the number of returns are provided in Figure 1 .

Fig. 1.

Search terms and number of articles returned in the selection of records for inclusion in this scoping review. Search was conducted by Technology and Patent Research International.

Eligibility Criteria

Both authors independently screened the articles identified in the search to determine their relevance. To be considered eligible for inclusion, only original, full-text articles documenting in-person, virtual, or self-led trainings on diabetes devices (eg, CGM and insulin pumps) were considered; review articles were excluded. The trainings were restricted to new users or onboarding and could be conducted in any country. Articles considered to be outside the study focus, including those describing general nutrition and diabetes education, were excluded.

Screening of Included Articles

The authors first screened the title and abstract of the identified articles to ensure that they met the inclusion criteria. Full-text articles of all those deemed eligible were then retrieved, if possible, for further review by the authors. Articles were excluded if they did not meet the eligibility criteria (Fig. 2 ).

Fig. 2.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram showing the selection process of articles included in this review.

Of the 25 articles returned from the database search, 5 were excluded during the preliminary title and abstract screening for not meeting the inclusion criteria. Full text could not be found for 1 article, and it was excluded from this review. Of the 19 remaining articles, 8 were excluded because they were abstracts, which led to 11 articles being included in this scoping review (Fig. 2).

Results and Discussion

Article Characteristics

Table 1 summarizes the characteristics of the 11 articles and provides information on study location, participant characteristics, and objectives relevant to this review. Six of the articles provided information collected in the United States,13, 14, 15, 16, 17, 18 and the other 5 articles presented information from participants from the United Kingdom,¹⁹ Colombia,²⁰ Qatar,²¹ Italy,²² and Israel.²³ Seven articles specified participants as having type 1 diabetes,^{13 , 16 ,} 19, 20, 21, 22, 23 1 included people with either type 1 diabetes or type 2 diabetes,¹⁵ 1 included HCPs,¹⁹ and 2 did not provide this type of participant information.^{17 , 18} Two articles reported only on measures of satisfaction or perceptions related to alternative methods of device training,^{19 , 23} 5 reported only on clinical (eg, glucometric) measures and utilization of alternative methods of device training,^{13 , 14 ,} 20, 21, 22 and 4 reported on both clinical measures as well as satisfaction with alternative training method(s) or device utilization.15, 16, 17, 18

Table 1.

Overview of Articles Investigating Alternative Methods of Device Training

Author, year	Country	Participant characteristics	Relevant objective
Perception of virtual visits
Kirzhner et al,23 2022	Israel	71 adults with T1D	Assess perceptions of telemedicine and preference vs traditional in-person visits
Forde et al,19 2022	UK	143 HCPs: 48% diabetes physicians, 52% diabetes educators	Understand HCP experiences with remote training and identify barriers/issues to remote training
Clinical measures and utilization of alternative training methods
Cherubini et al,22 2021	Italy	43 children with T1D using the Basal-IQ system for >3 mo	Analyze the impact of a virtual education camp on glucose control after in-person training on the use of a CLC system
Gómez et al,20 2021	Colombia	91 adults with T1D	Describe the efficacy and safety of a virtual training program for PwD upgrading to the HCL system from MDI or SAP therapy
Petrovski et al,21 2020	Qatar	30 children with T1D	Assess glycemic control after a 10-d HCL training protocol for children on MDI, with or without CGM, and with no prior pump experience
Faulds et al,13 2019	US	34 adults with T1D	Examine real-world utilization and glycemic control after the HCL initiation process
Lyons et al,14 2021	US	5 diabetes centers in the T1DX-QI collaboration	Develop strategies, including telehealth, to increase insulin pump use among youth (aged 12-26 y) with T1D
Both satisfaction and clinical measures
Gal et al,15 2020	US	34 adults (n = 27 with T1D, n = 7 with T2D) using insulin and interested in initiating CGM	Evaluate the feasibility of telehealth to expand access to initiation of CGM use and participant self-reported satisfaction with CGM
Pinsker et al,17 2021	US	8984 virtual trainings, 14,284 in-person trainings	Compare user experience and glucometric outcomes of remote and in-person training sessions
Berget at al16 2020	US	72 participants (age at BL,14 ± 4.3 y) trained to use the HCL system through a combination of face-to-face and remote training; 51% completed the feedback survey	Assess glucose control (TIR and TBR) and describe the lessons learned from the training and follow-up of the new HCL training/education program
Vigersky et al,18 2021	US	Pediatric and adult PwD new to MiniMed 670G receiving face-to-face or remote pump training	Evaluate glucometric outcomes and patient satisfaction with the virtual training program

Abbreviations: BL = baseline; CGM = continuous glucose monitor; CLC = closed-loop control; HCL = hybrid closed loop; HCP = health care professional; MDI = multiple daily injections; PwD = people with diabetes; SAP = sensor-augmented pump; T1D = type 1 diabetes; T1DX-QI = Type 1 Diabetes Exchange Quality Improvement; T2D = type 2 diabetes; TBR = time below range; TIR = time in range; UK = United Kingdom; US = United States.

Alternative Methods of Device Training Identified

Table 2 summarizes the alternative training methods identified, including video conferencing, phone calls, and virtual education camps (vECs).13, 14, 15, 16, 17, 18 ^, 20, 21, 22 Some articles highlighted the development of an intentional virtual-based curriculum in direct response to the pandemic.^{17 , 18 , 20} Gómez et al²⁰ demonstrated a live, virtual, video conference–based structured curriculum with 3 to 5 steps based on user experience. For those participants using multiple daily injections (MDI) or sensor-augmented pumps, for example, there was a 5-step training program that included disease state education. Conversely, for those participants who upgraded from using a predictive low glucose management pump, training consisted of the final 3 steps of using the hybrid closed-loop (HCL) system in manual and automated mode. Pinsker et al¹⁷ also reported a 3-phase, structured virtual curriculum with pretraining work including task completion using a simulator, video conferencing sessions on the actual pump with feedback using the teach-back method, and follow-up including data review through a mobile application.

Table 2.

Methods of Device Training

Author, year	Training method
Perception of virtual visits
Kirzhner et al,23 2022	This was a survey-based questionnaire for people with T1D attending a hospital-affiliated diabetes clinic that assessed preference for in-person vs telemedicine (calls, texts, and video conferencing) visits.
Forde et al,19 2022	This was a survey-based questionnaire for HCPs who provide device training to people with T1D to assess their experience and preference/satisfaction with in-person vs virtual visits.
Clinical measures and utilization of alternative training methods
Cherubini et al,22 2021	In-person training to help PwD upgrade from using the Basal-IQ system to the Tandem Control-IQ system, followed by a 3-d virtual education camp (Zoom meetings) to strengthen educational support.
Gómez et al,20 2021	Structured Zoom video conference curriculum and follow-up for MiniMed 670G across all levels of experience with technology.
Petrovski et al,21 2020	Structured in-person 10-d protocol for initiation on the MiniMed 670G HCL system for PwD using MDI, with phone calls and in-person follow-up in automated mode.
Faulds et al,13 2019	Initial in-person device training on the MiniMed 670G HCL system across all levels of experience with technology, with phone and electronic messaging support on entering automated mode.
Lyons et al,14 2021	Comprehensive handouts or telehealth education on pump benefits and considerations, in-person initiation, and ongoing support via mobile app, virtual visits, and booklets; each participating center independently developed and implemented strategies to increase insulin pump use.
Both satisfaction and clinical measures
Gal et al,15 2020	CGM initiation via video conferencing or phone calls with phone, text, and email follow-up.
Pinsker et al,17 2021	Structured video conference curriculum for initial training on all Tandem t:slim X2 insulin pump starts across all levels of experience with technology, with a mobile application for data downloading for participants seen during the COVID-19 pandemic (March 19, 2020, to July 28, 2020); in-person training for participants prior to COVID-19 (January 1, 2020, to March 18, 2020).
Berget at al,16 2020	A 3-phase program consisting of an initial in-person introduction class to educate current open-loop pump users on the HCL system, a live web-based video conference group class to initiate HCL mode, and 3 follow-up phone calls for ongoing support during the first 4 wk of HCL use.
Vigersky et al,18 2021	In-person training (before the COVID-19 pandemic, January 20, 2020, to February 22, 2020) vs Zoom video conferencing (during the COVID-19 pandemic, March 20, 2020, to April 22, 2020) on the MiniMed 670G HCL system across all levels of experience with technology.

Abbreviations: CGM = continuous glucose monitor; COVID-19 = coronavirus disease 2019; HCL = hybrid closed loop; HCP = health care professional; MDI = multiple daily injections; PwD = people with diabetes; T1D = type 1 diabetes.

In addition to reporting on structured curricula tailored toward a virtual training environment, some articles reported the need for additional consideration of the complexity of the device. In 4 of the 11 articles, participants received initial in-person training followed by training through virtual modes (phone calls, electronic messaging, mobile application, and video conferencing) that allowed for continued learning and optimal use of more advanced devices, such as the HCL system.^{13 , 16 , 21 , 22} In another article, Gal et al¹⁵ reported initiating CGM use for PwD solely through live, virtual video conferencing or phone calls.

Alternative methods were also reported in articles focused on pediatric PwD when additional support for both PwD and their families was needed. Lyons et al¹⁴ reported on 5 centers that used a hybrid of in-person training and alternative ongoing support tools, such as comprehensive handouts, more frequent touchpoints through virtual visits, as well as mobile applications and booklets. Berget et al¹⁶ further stressed the importance of ongoing educational support and data review for pediatric PwD and caregivers using the HCL system through a 3-phase training program that included live, virtual video conference education sessions and follow-up phone calls.

Equivalent or Improved Clinical and User Satisfaction Findings with Alternative Methods of Device Training

Six of the 11 articles provided information regarding user satisfaction with alternative training methods or with the use of CGMs (Table 3 ).15, 16, 17, 18, 19 ^{, 23} Questionnaires were completed by phone, via internet, or on paper and ranged from a few questions about preference (in-person vs remote) to more formal templates, depending on the study. There was a high degree of user satisfaction with virtual visits overall, with a preference for a combination of in-person and virtual visits, although user satisfaction evaluations may be subject to selection bias due to the requirement to return questionnaires. Some articles suggested that not only device users but also caregivers and HCPs may prefer alternative training methods over traditional face-to-face methods.^{19 , 23} Vigersky et al¹⁸ reported a 98% overall satisfaction rating with video conference training for the MiniMed 670G system (Medtronic) across all users, including those receiving baseline therapies of the following: (1) MDI and sensor; (2) MDI and self-monitored blood glucose; (3) a pump and sensor, with an upgrade of the pump; and (4) a pump and sensor, with an upgrade of both. In addition, they reported an increase in participants’ satisfaction as measured by an increase in the Net Promoter Score, a single-question management tool. For question 1, “based solely upon your recent training experience, how likely might you be to recommend Medtronic to another person who themselves are insulin-dependent (if they were considering pump therapy),” the pre-COVID Net Promoter Score was 78 compared with an intra-COVID score of 84. The percentage of calls to the 24-hour technical support line for educational support decreased by 0% to 19% after the transition to video conferencing. However, this was accompanied by a 187% increase in software support inquiries. Similarly, in the article by Berget et al,¹⁶ participants gave the virtual training program a 9.3 out of 10 satisfaction rating and suggested an increase in follow-up time from 1 to 3 months. Gal et al¹⁵ reported that all 34 participants who underwent remote training via video conferencing or phone calls for CGM initiation self-reported a high degree of satisfaction with its use. Participants also reported significant positive changes in diabetes technology attitudes, hypoglycemia confidence, and glucose monitoring satisfaction. In addition, they reported decreases in diabetes management distress, emotional burden, and behavioral burden.

Table 3.

Perceptions of and Satisfaction with Alternative Methods of Device Training: Key Results and Conclusions

Author, year	Study results	Study conclusions
Kirzhner et al,23 2022	•62% of participants preferred a combination of in-person and virtual visits •14% preferred only virtual visits •24% preferred only conventional visits •Participants aged ≥61 y preferred in-person visits •Sex, origin, education, diabetes duration, mode of insulin treatment, and distance from the clinic were not associated with method preference •66% felt confident to download data from their personal medical devices	Although participants from a wide range of treatment methods are willing to use telemedicine, virtual visits do not fully replace in-person visits, especially for older people with T1D.
Forde et al,19 2022	•Percentage of and time for training (median, IQR), by training method: 20% (5%-30%), 30 min (20-36 min) face-to-face; 50% (30%-80%), 20 min (15-30 min) telephone; and 10% (0%-35%), 30 min (20-30 min) video •63% of HCPs felt that remote consultations offered an effective strategy for care delivery •The most common barriers to virtual training were patient familiarity with technology (72%) and access to patient device data (67%) •Having a smartphone-based device with automated cloud uploads made consultations easier (73%) and more effective (85%) and facilitated hospital discharge (61%) •HCP perceived effectiveness of remote training for new CSII starts and CSII renewals rated high on a 7-point scale, median (IQR): 5 (4-6) and 5.5 (5-6), respectively. The most common barriers to remote pump trainings were patient digital literacy (61%), limited HCP experience (46%), and time required per patient (44%) •73% noted a reduction in new CSII starts •61% noted a reduction in CSII renewals •Variation in the proportion of new CSII starts done virtually (53.9% of HCPs reported <20%, and 32.9% reported >60%) •Most HCPs felt that remote consultations and trainings will continue after COVID-19	In general, HCPs have embraced remote care delivery, but digital literacy remains a barrier for some PwD.
Gal et al,15 2020	•Participants reported reduced diabetes management distress, emotional burden, and behavioral burden and increased satisfaction with glucose monitoring, hypoglycemic confidence, trust, and attitudes on diabetes technology	Remote CGM initiation and training were associated with self-reported improvements in quality of life.
Pinsker et al,17 2021	•Virtual training was associated with higher user satisfaction (4.78 ± 0.52 vs 4.64 ± 0.68; P < .01) and higher user confidence (4.61 ± 0.75 vs 4.47 ± 0.85; P < .01) than in-person training	The new virtual training curriculum received higher participant approval than traditional in-person training.
Berget at al,16 2020	•The “Plan-Do-Study-Act” HCL training program emphasizes education on HCL exits, CGM use, and optimizing insulin-to-carbohydrate ratio settings •Mean sensor wear time of 83% over 4 wk after HCL mode initiation, with 72% of time spent in the HCL mode	The strategies used in this study could potentially inform best practices for implementing new diabetes technologies at other diabetes clinics.
Vigersky et al,18 2021	•The Zoom video conferencing application had a 98% satisfaction •NPS for question 1a increased from 78 in the pre–COVID-19 cohort to 84 in the intra–COVID-19 cohort •Calls for educational assistance to the technical support team decreased by 0%-19%, whereas requests for login and software installation support increased by 187%	Virtual training on the MiniMed 670G system resulted in high satisfaction.

Abbreviations: CGM = continuous glucose monitor; COVID-19 = coronavirus disease 2019; CSII = continuous subcutaneous insulin infusion; HCL= hybrid closed loop; HCP = health care professional; NPS = Net Promotor Score; PwD = people with diabetes; T1D = type 1 diabetes

^aQuestion 1 of the Net Promotor Score was “based solely upon your recent training experience, how likely might you be to recommend Medtronic to another person who themselves are insulin-dependent (if they were considering pump therapy).”

Eight of the 11 articles reported on a variety of short-term clinical outcomes, including glucose variability, glycated hemoglobin (HbA1c), time in range/target (TIR), and time below range (TBR) (Table 4 ).^{13 ,} 15, 16, 17, 18 ^, 20, 21, 22 Overall, these articles reported that clinical outcomes of alternative methods of device training were satisfactory and comparable with those typically seen with traditional in-person training.^{13 ,} 15, 16, 17, 18 ^, 20, 21, 22 Cherubini et al²² reported glucometrics 12 weeks after a 3-day vEC attended by children and adolescents who had upgraded to the Tandem Control-IQ system (Tandem Diabetes Care). The median percentage TIR increased from 64% on the Basal-IQ system to 76% after 12 weeks of using the Control-IQ pump and attending the vEC. Furthermore, >75% of the participants achieved a TIR of >70%, and HbA1c improved from 7.0% to 6.6% (53.0 to 48.6 mmol/mol) after the pump upgrade and the vEC. There were no episodes of diabetic ketoacidosis or severe hypoglycemia during the follow-up period. In addition, Gal et al¹⁵ reported a decrease in mean HbA1c from 8.3% to 7.2% (67 to 55 mmol/mol) in PwD remotely initiated on CGM. There was also a mean TIR increase from 48% at baseline to 59%, an increase that was sustained through the 12-week follow-up period. Although this study did not directly compare training methods, the authors concluded that remote CGM initiation was successful in achieving sustained use and improving glycemic control after 12 weeks and that this alternative training method could substantially increase the adoption of CGM.

Table 4.

Short-Term Clinical Measures Reported With Alternative Methods of Device Training: Key Results and Conclusions

Author, year	Study results	Study conclusions
Cherubini et al,22 2021	•Median (IQR) percentage of TIR (70-180 mg/dL) increased from 64% (56%-73%) with Basal-IQ to 76% (69%-82%) with Control-IQ (P < .001) •Percentage of time with glucose at 180-250 mg/dL and ≥251 mg/dL decreased from 24% to 18% (P < .001) and 9% to 4% (P < .001), respectively •Time with glucose at 54-70 mg/dL and <54 mg/dL remained low and did not change (1% and 0%, respectively, at both time points) •HbA1c decreased from 7.0% to 6.6% (53.0 to 48.6 mmol/mol) (P < .001) •More than 75% of participants achieved a TIR of >70% after vEC •No episodes of DKA or severe hypoglycemia were reported.	Over 75% of children attending a vEC after start-up of a CLC system obtained and maintained a TIR of >70%. The significant and persistent improvement in TIR supports the feasibility of a vEC in children and adolescents with T1D.
Gómez et al,20 2021	•Overall, mean TIR (70-180 mg/dL) increased from 77.3% ± 11.3% in manual mode to 81.6% ± 7.7% (P < .0001) after 2 wk in automated mode •Participants using MDI previously had a mean TIR of 81.5% in automated mode •Previous PLGM users’ mean TIR was 82.5% in automated mode •Previous LGS users’ mean TIR was 81.6% in automated mode •Overall, mean TBR (54-70 mg/dL and <54 mg/dL) decreased from 2.7% and 0.5% in manual mode to 1.8% and 0.3%, respectively, after 2 wk in automated mode (P = .001 and P = .01, respectively) •Overall, GV decreased (% coefficient of variation 32.4 in manual mode vs 29.7 after 2 wk in automated mode; P < .001), independent of BL therapy •Sensor compliance was >90% during training and follow-up; automated mode was used 98.5% of the time and remained at >95% after training	Virtual training on HCL systems was safe and effective, resulting in improvements in TIR, TBR, and GV independent of previous treatment.
Petrovski et al,21 2020	•Median percentage time of sensor use was 92% •Median time in automated mode was 89% •Mean HbA1c decreased from 8.2% ± 1.4% (66 ± 15.3 mmol/mol) at BL to 6.7% ± 0.5% (50 ± 5.5 mmol/mol) over 84 d in automated mode (P = .017) •Time in TGR (71-180 mg/dL) increased from 46.9% ± 18.5% at BL to 74.8% ± 7.1% over 84 d in automated mode (P < .001) •Time BTR (51-70 mg/dL and ≤50 mg/dL) was 2.7% and 0.4% at BL, respectively, and 2.5% and 0.3% over 84 d in automated mode, respectively, with no reports of severe hypoglycemia or DKA •TAR (181-250 mg/dL and ≥251 mg/dL) decreased from 25.8% and 24.2% at BL, respectively, to 16.9% and 5.5% over 84 d in automated mode, respectively	A concise, structured 10-d protocol allows children and adolescents with T1D on MDI to successfully initiate HCL use.
Faulds et al,13 2019	•Mean SMBG/d increased from 5.2 at BL to 6.2 at weeks 6-12 (final follow-up; P < .05) •3.3 sensor calibrations/d •Time in automated mode decreased from 79.3% at 2 wk to 72.3% at the final follow-up •82% of participants spent >50% of time in automated mode •Automated mode exits over the final 14-d download was 8.2 •TIR was 67.3% in manual mode, 73.4% at 2 wk (P = .09), and 71.1% by week 6-12 (P = .08) •HbA1c decreased by 0.51% (P = .001), with no change in total daily insulin dose or % basal insulin dose •Those with BL HbA1c <7.0% (<53 mmol/mol) spent more TIR than those with BL HbA1c ≥7.0% (≥53 mmol/mol; 78.0% vs 68.3%) even though they spent less time in automated mode (66.5% vs 74.8%)	These results indicate that implementing HCL technology using a structured education program is feasible, with overall pump benefit potentially varying depending on BL characteristics, eg, HbA1c.
Lyons et al,14 2021	Over 22 mo (May 2018 to February 2020), the study average of insulin pump uptake increased by 13% from 45% to 58%	The rate of insulin pump use increased over 22 mo. The project also promoted a cooperative culture among diabetes centers in sharing best practices and population data.
Gal et al,15 2020	•All participants were using CGMs at 12 wk; median (IQR) CGM use in the final 4 wk = 7.0 d/wk (6.7-7.0) •Mean HbA1c decreased from 8.3% (67 mmol/mol) at BL to 7.2% (55 mmol/mol) at 12 wk (P < .001) •Mean TIR was 59% over 12 wk compared with 48% at BL as estimated from HbA1c, corresponding to an improvement of approximately 2.7 h/d Over 12 wk, median TBR of <70 mg/dL was 1.4%, and TBR of <54 mg/dL was 0.2%	Remote CGM initiation and training allowed for success in achieving sustained CGM use and improved glycemic control after 12 wk. These types of trainings have the potential to increase CGM use among PwD who use insulin.
Pinsker et al,17 2021	TIR was 72% (60%-81%) up to 6 mo after virtual training vs 67% (54%-78%) after in-person training	The new virtual training curriculum proved safe and effective.
Berget at al,16 2020	•Mean sensor wear time was 83% over 4 wk after initiating the HCL mode, with 72% of time spent in HCL mode •Significant increase in TIR from 49% at BL to 62.2% at week 4 after HCL class (P < .001) •Nonsignificant decrease in TBR from 1.8% at BL to 1.6 at week 4 after HCL class	The strategies used in this study could potentially inform best practices for implementing new diabetes technologies at other diabetes clinics.
Vigersky et al,18 2021	Comparing pre–COVID-19 and intra–COVID-19 cohorts: •TIR was 70.1% and 68.0%, respectively •TBR 54-69 mg/dL was 1.7% and 1.8%, respectively •TBR <54 mg/dL was 0.4% and 0.5%, respectively •TAR 181-250 mg/dL was 28.2% and 30.3%, respectively •TAR >250 mg/dL was 6.6% and 7.8%, respectively	Virtual training on the MiniMed 670G system resulted in short-term glycemic results comparable with those of in-person training.

Abbreviations: BL = baseline; BTR = below target range (51-70 mg/dL and ≤50 mg/dL); CGM = continuous glucose monitor; CLC = closed-loop control; COVID-19 = coronavirus disease 2019; DKA = diabetic ketoacidosis; GV = glucose variability; HbA1c = glycated hemoglobin (%); HCL= hybrid closed loop; LGS = low glucose suspend; MDI = multiple daily injections; PLGM = predictive low glucose management; PwD = people with diabetes; SMBG = self-monitored blood glucose; T1D = type 1 diabetes; TAR = time above range; TBR = time below range; TGR = target glucose range (71-180 mg/dL); TIR = time in range; vEC = virtual education camp.

Two of the articles compared data from in-person trainings and alternative methods of training.^{17 , 18} Pinsker et al¹⁷ reported a 6-month posttraining sensor TIR of 72% for those receiving remote training via a structured, virtual training curriculum compared with 67% for those receiving in-person training. Vigersky et al¹⁸ retrospectively compared data from pre–COVID-19 in-person training with data from intra–COVID-19 virtual training. They reported no significant difference in posttraining sensor TIR between the 2 training methods (70.1% in-person training and 68.0% remote training); other glycemic outcome measures were also not significantly different between the training methods.

Planning Alternative Methods of Device Training to Enhance Learning, Retention, and Competence

Some articles included in this review reported on a variety of techniques recommended for improved learning and retention.^{18 , 20 , 22} Principles of adult learning theory have shown that novice learners organize knowledge with more superficial structures, and thus, new material may be more difficult to assimilate.²⁴ When considering alternative methods of training, use case-based examples to increase relevancy and build a deeper understanding for everyday problem solving. Methods such as simulations and problem-based assessments can reduce the time spent acquiring superficial knowledge and may decrease the need for frequent face-to-face encounters. Another important lesson learned from these articles is that irrespective of the method used for onboarding, resources for reinforced learnings and follow-up/ongoing support are needed.^{15 , 16 ,} 20, 21, 22 These resources include self-guided problem-based microlearnings, such as short videos, quick reference guides, or tools,^{14 , 17} and staff to conduct follow-up phone calls or send texts/emails to device users to provide encouragement, answer questions, and troubleshoot device issues.^{15 , 16 , 20} Both effective training for onboarding and ongoing support are vital for successful device utilization and optimal clinical outcomes in the long term. Finally, 2 of the articles showed that factors such as race, sex, baseline education status, and distance from clinics were not predictive of training method preference,^{17 , 23} suggesting that alternatives to in-person training can address some disparities in optimal adoption and use of diabetes devices. However, alternative training methods also have limitations, including digital literacy for both PwD and HCPs and access to adequate broadband. Being that this solution is not appropriate for everyone, hybrid options based on patient preferences are needed. HCPs must be prepared to provide choices in the type of training that fits the individual patient’s need, learning style, and socioeconomic circumstance.

Strengths and Limitations

This review used a rigorous process per the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews checklist to synthesize recommendations for alternative methods of device trainings from a body of literature with diverse methods and outcomes. The cited articles provided information on participants from both within and outside the United States, making these recommendations relevant to a variety of countries. This review lends support to the efficacy of alternative methods of device training using both subjective (eg, PwD and caregiver-reported satisfaction) and objective (eg, TIR and HbA1c) measurements. There are also limitations to this review. Only articles in the English language were included; therefore, some relevant articles may have been overlooked. There is potential for reporting bias because articles covering both traditional and alternative methods of training with null or inconclusive short-term clinical outcomes or user satisfaction data are unlikely to be published or reported in the literature. Finally, heterogeneity was difficult to assess because there are currently no clearly identified outcome measures or definite expectation of metrics to evaluate diabetes device training.

Conclusions

The articles included in this scoping review address our research questions. First, they demonstrate that a variety of alternative methods of device training are being offered and utilized by different device manufacturers. This review supports more widespread use, reimbursement, and resources allocated for these alternative methods. Second, short-term clinical outcomes, user satisfaction, and confidence were found to be similar across the different training methods. Further work is needed to evaluate these interventions for longer-term outcomes. Finally, important considerations for providers include utilizing adult learning best practices, such as problem- and case-based learning and assessments; providing strong multimodal ongoing support; and giving PwD options when considering successful device training. Even with in-depth training and experience, PwD face challenges when using technology. Setting expectations and creating patient-centered goals in collaboration with the user (and their caregivers, if appropriate) are essential in diabetes management systems.

Disclosure

Both authors are employees of Eli Lilly and Company. L.N.T. holds Medtronic stock. N.A.X. has received payment from Boehringer Ingelheim for speaking engagements.