Sage Journals: Discover world-class research

Abstract

Background:

Real-world studies of people with type 2 diabetes (T2D) have shown insufficient dose adjustment during basal insulin titration in clinical practice leading to suboptimal treatment. Thus, 60% of people with T2D treated with insulin do not reach glycemic targets. This emphasizes a need for methods supporting efficient and individualized basal insulin titration of people with T2D. However, no systematic review of basal insulin dose guidance for people with T2D has been found.

Objective:

To provide an overview of basal insulin dose guidance methods that support titration of people with T2D and categorize these methods by characteristics, effect, and user experience.

Methods:

The review was conducted according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines. Studies about basal insulin dose guidance, including adults with T2D on basal insulin analogs published before September 7, 2022, were included. Joanna Briggs Institute critical appraisal checklists were applied to assess risk of bias.

Results:

In total, 35 studies were included, and three categories of dose guidance were identified: paper-based titration algorithms, telehealth solutions, and mathematical models. Heterogeneous reporting of glycemic outcomes challenged comparison of effect between the three categories. Few studies assessed user experience.

Conclusions:

Studies mainly used titration algorithms to titrate basal insulin as telehealth or in paper format, except for studies using mathematical models. A numerically larger proportion of participants seemed to reach target using telehealth solutions compared to paper-based titration algorithms. Exploring capabilities of machine learning may provide insights that could pioneer future research while focusing on holistic development.

Keywords

basal insulin dose guidance glycemic control insulin titration type 2 diabetes systematic review

Introduction

Initiation of basal insulin is a complex and time-consuming task associated with clinical inertia.^1
-5 Thus, approximately 60% of people with T2D treated with insulin do not reach glycemic targets.^4,6
-8 Insulin titration is used when determining the optimal dose for an individual.^2,4,9 This is necessary since people with T2D vary in pancreatic insulin production and insulin resistance.^9,10 Hence, the optimal dose of basal insulin differs among people with T2D and may change over time due to, for example, stress levels, lifestyle changes, and sickness.

Suboptimal treatment is partly caused by non-adherence to treatment and failure to initiate or intensify treatment promptly.^9,11 Lack of adjustment to insulin treatment is mainly caused by the complexity of the titration process.⁵ This causes people with T2D to remain on suboptimal insulin doses, leading to less improvement in glycemic control than what could have been accomplished with an optimal dose.^5,12,13 In addition, studies based on real-world data have shown both a delay in the initiation of basal insulin and insufficient dose adjustment during titration.^1,14,15 Suboptimal insulin titration has been shown in the range of 3 to 12 months after initiation of active titration in clinical practice.^3,6,16
-19 This elucidates that people with T2D, in some cases, have not reached glycemic target after 3+ months of active titration. Failure to achieve glycemic targets during the initial 3 months of titration is associated with a higher risk of failure to reach glycemic targets two years after the initiation.¹⁵ This emphasizes the need for dose guidance supporting efficient and individualized basal insulin titration of people with T2D to provide optimal and timely treatment.

In recent years, basal insulin dose guidance has been of rapidly growing interest within international research, emphasized by an increase of publications on the subject. Despite this interest and the fact that it has been a research field for several decades, a preliminary search of the Cochrane Database of Systematic Reviews and Reviews, the International Prospective Register of Systematic Reviews (PROSPERO), and Joanna Briggs Institute (JBI) Evidence Synthesis revealed no systematic review of basal insulin dose guidance for people with T2D. Therefore, this systematic review aims to provide an overview of methods used for basal insulin dose guidance supporting titration of people with T2D and categorize these methods by characteristics, effect, and user experience.

Methods

Study Design

The systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.²⁰ Therefore, a protocol was registered in PROSPERO on December 19, 2021 (CRD42021289364), forming the review’s basis.²¹

Eligibility Criteria

Studies evaluating dose guidance methods supporting titration of people with T2D treated with basal insulin analogs in any setting, including adult participants (≤18 years) diagnosed with T2D, were considered. Studies investigating populations of mixed diabetes types without a transparent subgroup analysis or without a clear statement of diabetes type were excluded. Studies including participants on basal-bolus regimens, human or intermediate insulin, or other injectable antidiabetic treatment were excluded.

Primary studies reporting any glycemic outcome published in English, Danish, Norwegian, or Swedish before September 7, 2022, as peer-reviewed full-text, were included. All study designs except study protocols, animal research, expert opinions, and case studies were considered.

Information Sources and Search Strategy

A comprehensive systematic search was performed in PubMed, Embase, and IEEE by one author (C.H.N.T) with assistance from a research librarian. Citation and reference searches were conducted in Google Scholar. Authors of relevant studies were contacted if additional information was needed.

Unstructured searches in PubMed and Google Scholar were performed to identify relevant search terms. The search was adjusted to each database. Search terms included different synonyms and spellings. Search functions were applied, including thesaurus, Boolean operators, phrase, truncation, free text, and advanced search (Supplementary material).

Selection Process

First, studies identified through the systematic search were uploaded to RefWorks (version 2.1.0.1). Second, duplicates were removed using the functions Exact duplicates and Close duplicates. Third, one reviewer (C.H.N.T.) screened the title and abstract of the remaining studies. Fourth, studies deemed eligible were retrieved in full-text and assessed by one reviewer (C.H.N.T.). Doubt about studies’ eligibility was resolved through discussion with co-authors. Reasons for exclusion of studies were recorded during full-text assessment (Supplementary material). The final sample consisted of studies deemed eligible after full-text assessment.

Data Extraction and Synthesis

One author (C.H.N.T.) extracted data using a sheet in Microsoft Excel (2016). Extracted data included study characteristics (title, authors, publication year, study design, country, sample size, and duration of study), participant characteristics (age, sex, body mass index [BMI], insulin-naïve, and initial HbA1c), characteristics of the dose guidance method (setting, description of the method, and type of insulin used), and glycemic outcomes.

A narrative synthesis of extracted data was conducted, and characteristics of studies and populations were described. The narrative synthesis focused on categorizing dose guidance methods and assessing effect of the interventions and user experience according to the categorization.

Risk of Bias Assessment

Critical appraisal tools from JBI were applied by study design of the studies to assess risk of bias.²² Study design was determined using Andrews and Likis, 2015.²³ One author (C.H.N.T.) assessed included studies with support from co-authors.

Before critical appraisal was performed, authors agreed on a scoring system and cut-off points per the JBI reviewers manual.²⁴ Studies were judged as described in Melo et al.²⁵ A suitable tool for simulation studies was not found from JBI; therefore, the critical appraisal tool from Fone et al²⁶ was used.

Results

Study Selection

A total of 4,363 papers were found. After removing duplicates, 3,327 papers were included in title and abstract screening. Of those, 280 papers were found eligible for full-text screening. Thirty-one papers met the inclusion criteria and were included in the review. Four additional papers were identified through reference and citation searches. Thus, 35 articles were included in this review. The selection process is presented in Figure 1. The supplementary material contains a tabular overview of data extracted from the included studies.

Figure 1.

The selection process is illustrated in a preferred reporting items for systematic reviews and meta-analyses flowchart.²⁰

Some studies seemed eligible but were excluded due to use of human insulin or basal-bolus regime in a subgroup of participants without a transparent subgroup analysis of participants treated only with basal insulin analogs, or the use of bolus insulin as rescue medication.^13,27
-29

Study Characteristics

Seven studies were quasi-experimental design,^30
-36 20 studies were randomized controlled trials (RCT),^37
-56 three studies were mixed method,^57
-59 one study was of qualitative design,⁶⁰ one study was a cohort,⁶¹ and three studies were simulation design.^8,10,62 Mixed method studies were a mix of quasi-experimental and qualitative design. The studies were published from 2006 to 2022 and enrolled 19,432 people with T2D. The length of the studies ranged from 28 days to 12 months.

The studies were conducted in 31 countries across Europa, Asia, North and South America, the Middle East, and Africa. Seven studies did not specify in which country it was conducted.^{8,10,32,48,55,61,62}

Participant Characteristics

Characteristics of participants were similar regarding initial BMI, age, and sex distribution. The most significant difference was whether participants were insulin-naïve at start-of-trial. The study population in 60% of the studies were insulin-naïve.^{8,10,31,34,35,37,39
-41,45,46,48
-51,53,56
-58,61,62} In 14% of studies, the population continued basal insulin treatment initiated before the study,^{30,32,33,36,43} and 26% of studies included a study population of both insulin-naïve and people contuneing previously initiated treatment.^{38,42,44,47,52,54,55,59,60} Initial HbA1c, duration of diabetes, and whether the study population was insulin-naïve are essential factors to consider when comparing the impact on glycemic control from dose guidance interventions.^15,63
-67 All study populations had initial HbA1c above 7%, and diabetes duration ranged from 2.9 to 15.9 years.

Characteristics of the Dose Guidance Methods

Twenty-one of identified dose guidance methods were developed for titration of glargine,^{30,32,34,36
-39,41,42,44,45,47,49,51,53
-56,58,61,62} three for detemir,^40,48,52 five for degludec,^{8,10,31,43,46} one for icodec⁵⁰, and one for glargine and detemir.⁵⁹ Four studies did not specify insulin further than it was basal insulin analogs.^33,35,57,60

Approximately 70% of the studies were in an outpatient clinic. The remaining studies were in primary care^{34,35,42,51,52,61} or did not specify the setting.^{8,10,36,50,62}

Categorization of the dose guidance methods

Identified dose guidance methods were divided into three categories: paper-based titration algorithms, telehealth solutions, and mathematical models (Figure 2).

Figure 2.

Overview of type of dose guidance methods used in the included studies.

Paper-based titration algorithms reflect standard practice at the time of writing. The studies investigated algorithms with varying targets and sizes of dose adjustment carried out during in-person visits. In total, 20 studies investigated paper-based titration algorithms.^{32,34,36
-38,40
-43,46,48
-53,56,58,61,62}

Telehealth solutions covered telemonitoring solutions with titration across a digital platform^{30,45,54,57,59,60} and combined with home visits,³⁵ or self-titration decision support.^{33,39,44,47,55} In contrast to studies addressing paper-based algorithms, the organizational setup was altered in these studies. Interactions between participants and healthcare professionals (HCP) were primarily handled over distance via phone. In total, 12 studies investigated telehealth solutions.^{30,33,35,39,44,45,47,54,55,57,59,60}

Mathematical models were investigated by three studies using compartment modeling and control theory.^8,10,31 Most of these studies did not specify the use case of the method.

Dose guidance methods covered both physician- and patient-led methods. The distribution was similar for paper-based titration algorithms and telehealth solutions, where most approaches based on mathematical models did not specify the intended user (Figure 3).

Figure 3.

Distribution of the intended user of the identified dose guidance methods according to the three main categories: paper-based titration algorithms, telehealth solutions, and mathematical models.

Description of the dose guidance method is presented in Table 1.

Table 1.

Overview of How Basal Insulin Was Titrated in the Included Studies Grouped by the Titration Algorithm Used.

Study	Description of titration algorithm	Category
Yuan et al³⁷	2-0-2 titration algorithm according to three different fasting blood glucose targets; 70 < FBG ≤ 100, 100 < FBG ≤ 110, or 110 < FBG ≤ 126 mg/dL. Titrated based on the lowest of three consecutive fasting SMBG values.	Paper-based titration algorithm
Zhang et al⁵⁸	Comparison of the use of a titration algorithm to reach different glycemic targets (Group 1: 70 < FBG ≤ 100 mg/dL, Group 2: 100 < FBG < 110 mg/dL, and Group 3: 110 < FBG ≤ 126 mg/dL)The titration algorithm used was a modification of the 2-0-2 algorithm.	Paper-based titration algorithm
Misra et al⁴¹	2-0-2-4 titration algorithm as patient-led compared to physician-led. Insulin doses were titrated every three days.	Paper-based titration algorithm
McGloin et al⁵⁷	MyMedic hub. Telemonitoring system where people with T2D were titrated using a 2-0-2 titration algorithm twice weekly for three weeks and once weekly after that.	Telehealth solution
Ngassa Piotie et al³⁵	Nurse-driven and home-based telehealth intervention where participants were titrated using the 2-0-2 titration algorithm to reach the target of 72-126 mg/dL.	Telehealth solution
Seufert et al⁶¹	2-0-2 titration algorithm (adjusted every three days) compared to the 2-0-2-4-6-8 titration algorithm (adjusted every 3-5 days).	Paper-based titration algorithm
Kadowaki et al⁴³	2-0-2 titration algorithm compared to the 2-0-2-4-6-8 titration algorithm at both fixed dosing and flexible dosing. Adjustments to insulin doses were made weekly.	Paper-based titration algorithm
Kennedy et al⁴⁹	Comparison of usual and active insulin titration using the 2-0-2-4-6-8 titration algorithm. If fasting blood glucose was below 70 mg/dL insulin dose was decreased to the previous dose.	Paper-based titration algorithm
Yu et al⁴⁰	3-0-3 titration algorithm compared to the 2-4-6-8 titration algorithm. Titration was performed per three days.	Paper-based titration algorithm
Blonde et al⁴⁸	3-0-3 titration algorithm to the target of 70-90 mg/dL compared to 79-110 mg/dL. Adjustments to insulin doses were made every three days.	Paper-based titration algorithm
Meneghini et al⁵²	3-0-3 titration algorithm, where adjustments were made every three days, compared to standard-of-care, where adjustments were made at the physician’s discretion.	Paper-based titration algorithm
Hsu et al⁴⁵	Diabetes management program. Telemonitoring system where the 3-0-3 titration algorithm was used to reach the target of 79-110 mg/dL.	Telehealth solution
Philis-Tsimikas et al⁴⁶	4-0-4 titration algorithm compared to the 4-2-0-2-4-6-8 titration algorithm. Adjustments of doses were made weekly based on one and the lowest of three consecutive days of fasting SMBG measure, respectively.	Paper-based titration algorithm
Lingvay et al⁵⁰	Comparison of four titration algorithms: three for icodec and one for glargine.Glargine: 4-0-4 titration algorithm to target 79-130 mg/dLIcodec titration A: 21-0-21 titration algorithm to target 79-130 mg/dLIcodec titration B: 28-0-28 titration algorithm to target 79-130 mg/dL (equivalent to the titration algorithm used for glargine)Icodec titration C: 28-0-28 titration algorithm to target 70-108 mg/dL	Paper-based titration algorithm
Garg et al⁵¹	2-0-2-4 titration algorithm as patient-led compared to physician-led. In the physician-led titration, group doses were adjusted at each visit, whereas doses were adjusted twice weekly in the patient-led titration group.	Paper-based titration algorithm
Sethi et al³⁶	Using the 2-0-2-4 titration algorithm to reach HbA1c<7%.The frequency of dose adjustments was made at least weekly and not more than every 3–4 days unless required for safety.	Paper-based titration algorithm
Ji et al⁵³	2-0-2-4-6 titration algorithm at a standard starting dose (0.2 U/kg) or a higher starting dose (0.3 U/kg).	Paper-based titration algorithm
Bajaj et al⁴⁴	LTHome/MyStar WebCoach. Decision support system for self-titration using the 4-2-0-2-4 titration algorithm to the target 90-130 mg/dL.	Telehealth solution
Davies et al⁵⁵	MyStar DoseCoach. Decision support system for self-titration using the 4-2-0-2-4 titration algorithm to reach the 90-130 mg/dL target.	Telehealth solution
Kim et al⁴⁷	Decision support system for self-titration using the 4-2-0-2-4-6 titration algorithm to the target 79-119 mg/dL.	Telehealth solution
Hu et al³⁹	Self-titration decision support program. One in-person visit was followed by five phone calls where insulin dose adjustments were made if needed, along with empowering coaching from a nurse. Otherwise, the participants self-titrated.Titration algorithm used: 6-4-2-0-2-4-6 to target 79-110 mg/dL.	Telehealth solution
Levy et al⁵⁹	MITI. Telemonitoring system where participants were titrated using the 2-1-0-2-3-4-5 titration algorithm through weekly phone calls.	Telehealth solution
Rogers et al⁶⁰	MITI. Telemonitoring system where participants were titrated using the 2-1-0-2-3-4-5 titration algorithm through weekly phone calls to reach the target of 79-130 mg/dL.	Telehealth solution
Levy et al⁵⁴	MITI. Telemonitoring system where participants were titrated using the 2-1-0-2-3-4-5 titration algorithm through weekly phone calls.	Telehealth solution
Bae et al³⁸	Comparison of the INSIGHT and EDITION titration algorithm.INSIGHT: titrate by one unit/day.EDITION: titrate by three units per three days.	Paper-based titration algorithm
Yale et al⁴²	Comparison of the paper-based titration algorithm INSIGHT and EDITION.In the INSIGHT group, insulin was titrated by one unit/day.In the EDITION group, insulin was titration by three units per three days based on median prebreakfast SMBG values of the last three days.	Paper-based titration algorithm
Hasan et al³⁴	ADA/EASD consensus titration algorithm of 2009. Increased with two units every three days until target (70-130 mg/dL) reached. If fasting blood glucose is >180 mg/dL, increase by four units every three days; if fasting blood glucose is <70 mg/dL, reduce by four units or 10% if >60 units.	Paper-based titration algorithm
Larsen et al³³	Electronic diary app to support self-titration by increasing dose by two units every three days if two of the previous three days’ fasting SMBG measures >121 mg/dL and no readings were <72 mg/dL.	Telehealth solution
Sieber et al⁶²	Comparison of three paper-based titration algorithms.Group 1: titrate by two units per three days to target 90-130 mg/dL.Group 2: titrate by four units per three days and by six units if blood glucose if >180 mg/dL to target 90-130 mg/dL.Group 3: titrate by two units per three days to target 110-150 mg/dL	Paper-based titration algorithm
Pfützner et al³²	Comparison of four paper-based titration algorithms.1. Target: 90-130 mg/dL. Increase dose by two units every three days.2. Target: 90-130 mg/dL. Increase the dose by four units every three days if blood glucose is >180 mg/dL, then increase by two units.3. Target: 110-150 mg/dL. Increase dose by two units every three days.4. Target: 70-100 mg/dL. Increase dose two units every three days.	Paper-based titration algorithm
Ishii et al⁵⁶	Comparison of physician and patient-led titration algorithm.Physician-led: 0-1-2-3-4 and decrease according to the physician’s discretion.Patient-led: 1-0-1.The frequency of dose adjustments was not specified.	Paper-based titration algorithm
Tamez-Pérez et al³⁰	MyDoseCoach. A combination of a mobile app and a web portal suggested basal insulin dose adjustments every three days based on a titration algorithm: 10% increase if SMBG>180 mg/dL, 5% increase if 140<SMBG<180 mg/dL, no change if 79<SMBG<140, 5% decrease if 70<SMBG<79 mg/dL, 10% decrease if SMBG<70 mg/dL.	Telehealth solution
Aradóttir et al³¹	Titration was performed using a linear dose-response algorithm.Day 1-4: No insulin.Day 5-9: 10 units insulin.Day 10: Evaluation of whether 10 units is sufficient or if the dose should be adjusted with 0.2 U/kg.Day 15: The dose estimation algorithm used CGM data from day 1-14, and 75% of the estimated dose was given to the participant.Day 20-84: titration using stepwise algorithm until target (72-108 mg/dL) reached.	Mathematical model
Krishnamoorthy et al¹⁰	Model-free titration approach using recursive least square-based extremum seeking control.	Mathematical model
Aradóttir et al⁸	A model predictive control-based dose guidance algorithm.	Mathematical model

Abbreviations: FBG, fasting blood glucose; SMBG, self-measured blood glucose; T2D, type 2 diabetes; MITI, mobile insulin titration intervention; ADA, American Diabetes; EASD, European Association for the Study of Diabetes; CGM, continuous glucose monitoring.

Table 1 elucidates that all identified dose guidance methods, except in Krishnamoorthy et al¹⁰ and Aradóttir et al,⁸ used titration algorithms to titrate basal insulin either in a digital tool or paper-based format. Aradóttir et al³¹ mixed the use of a mathematical model with use of a paper-based titration algorithm. Titration algorithms varied considerably among included studies, as approximately 18 algorithms were used. However, similar titration algorithms were found in studies investigating paper-based titration algorithms and telehealth solutions, for example, the 2-0-2 titration algorithm.

Effect of the dose guidance methods

Studies reported very heterogeneous glycemic outcomes (Supplementary material). The most frequently reported outcome was proportion of participants reaching glycemic target. However, this target differed among studies. Some studies used HbA1c < 7% as target, while others used fasting blood glucose within a specific range. The difference in how target was defined made it challenging to compare effect across studies. To enable a comparison to some degree to elucidate tendencies in effect across different dose guidance methods, an overview of the proportion of participants reaching target is presented in Figure 4. Approximately 23% of studies did not report proportion of participants reaching target at end-of-trial.^{8,10,33,40,45,57,58,62}

Figure 4.

Summary of the proportion of participants that reached a predefined glycemic target. Only studies that reported target as either fasting blood glucose within the target of 79-130 mg/dL, 90-130 mg/dL, or 72-108 mg/dL or HbA1c<7% (marked with *) is included in this figure.

As the only study investigating mathematical models, Aradóttir et al³¹ reported that all participants reached target with a mean time to target of 44 days (n = 8). The mean proportion of participants reaching target in studies investigating telehealth solutions was 61 ± 20% when considering both targets and 46 ± 29% when only considering HbA1c targets. The mean for paper-based titration algorithms was 41 ± 19% in both cases. This may indicate a tendency for a numerically larger proportion of participants titrated using telehealth solutions to reach target compared to paper-based titration algorithms.

Among these studies, few reported time-to-target. None of the studies about paper-based titration algorithms reported time-to-target. Three studies about telehealth solutions reported mean time-to-target, which ranged from 20 to 66 days.^30,54,59 It should be noted that two of these studies investigated the same telehealth solution.^54,59 Since few studies reported time-to-target, it is relevant to consider the mean study duration within the three categories to get an indication of time used to reach target. The mean duration was 22 ± 9 weeks for studies addressing paper-based titration algorithms, 16 ± 6 weeks for studies addressing telehealth solutions, and 11 ± 2 weeks for studies addressing mathematical models, of which most were simulations. On average, study duration of studies investigating paper-based titration algorithms was twice as long as for mathematical models and 6 weeks longer than telehealth studies.

User experience of the dose guidance methods

User experience was investigated by 14 studies, of which 11 studies^{36,38,41,42,44,45,51,54
-57} reported outcomes from standardized questionnaires (eg, Diabetes Treatment Satisfaction Questionnaire [DTSQ]), three studies^35,57,60 reported outcomes from interviews, and three studies^35,58,59 reported outcomes from non-standardized questionnaires. Studies addressing mathematical models did not investigate user experience.

The studies reporting baseline changes in the DTSQ scores showed varying results (Supplementary material). For telehealth solutions, the change ranged from 0.8 to 10.1 and from 0.1 to 11.7 for paper-based titration algorithms. This revealed no apparent difference in the change of DTSQ score between the two methods.

From non-standardized questionnaires and interviews, HCPs and people with T2D found telehealth solutions convenient and appropriate for titration of basal insulin.^35,57,59,60 Two of these studies investigated the same telehealth intervention.^59,60 People with T2D found it convenient to have fewer in-person interactions while maintaining contact with HCP via phone. In Rogers et al,⁶⁰ HCPs found telehealth intervention could reduce the burden of titration. In contrast, McGloin et al⁵⁷ elucidated an increased workload among HCPs caused by a large amount of generated data.

Only the study by Zhang et al⁵⁸ reported qualitative findings on the use of paper-based titration algorithms. The study found a gap between preferences of people with T2D and HCPs when choosing a titration algorithm. People with T2D preferred simple and easy-to-use algorithms. In contrast, HCPs preferred algorithms recommended by guidelines with higher perceived efficacy in lowering blood glucose levels and which were known to the HCP.

Critical Appraisal of the Studies

Table 2 -6 show the results of critical appraisal of the included studies.

Table 2.

Summary of Critical Appraisal Assessed by JBI Critical Appraisal Checklist for Randomized Controlled Trials.

Study	1	2	3	4	5	6	7	8	9	10	11	12	13	Risk of bias
Yuan et al³⁷	+	+	+	−	−	−	+	+	+	+	+	+	+	Low
Bae et al³⁸	U	+	+	−	−	−	+	+	+	+	+	+	+	Moderate
Hu et al³⁹	+	U	+	−	−	−	+	+	−	+	+	+	+	Moderate
Lingvay et al⁵⁰	U	+	+	−	−	−	+	+	+	+	+	+	+	Moderate
Ishii et al⁵⁶	+	U	U	−	−	+	+	+	−	+	+	+	+	Moderate
Yu et al⁴⁰	U	U	+	−	−	−	+	−	−	+	+	−	+	High
Ji et al⁵³	U	U	+	−	−	+	+	+	−	+	+	+	+	Moderate
Misra et al⁴¹	U	+	+	−	−	−	+	+	−	+	+	−	+	Moderate
Davies et al⁵⁵	U	U	+	−	−	−	+	−	+	+	+	−	+	High
Yale et al⁴²	U	U	+	−	−	−	+	+	+	+	+	−	+	Moderate
Kadowaki et al⁴³	U	U	+	−	−	−	+	−	+	+	+	+	+	Moderate
Bajaj et al⁴⁴	U	U	+	−	−	−	+	−	+	+	+	+	+	Moderate
Hsu et al⁴⁵	U	U	+	−	−	−	+	+	+	+	+	+	+	Moderate
Garg et al⁵¹	+	+	+	−	−	−	+	−	+	+	+	+	+	Moderate
Levy et al⁵⁴	+	+	+	−	−	−	+	+	+	+	+	+	+	Low
Philis-Tsimikas et al⁴⁶	U	U	+	−	−	−	+	+	+	+	+	+	+	Moderate
Kim et al⁴⁷	+	U	U	−	−	−	+	+	−	+	+	+	+	Moderate
Blonde et al⁴⁸	U	U	+	−	−	−	+	+	+	+	+	+	+	Moderate
Meneghini et al⁵²	U	U	+	−	−	−	U	−	−	+	+	+	+	High
Kennedy et al⁴⁹	U	U	+	−	−	−	+	+	−	+	+	+	+	Moderate

U = Unclear, + = Yes, and - = No. Question 3: Red marks visual inspection of between-group differences in baseline characteristics of the population to determine if the groups were similar, and green marks studies that performed statistical tests for the difference between groups. Question 9: Red marks intention-to-treat analysis was carried out but did not describe how lost-to-follow-up was handled. Green indicates that intention-to-treat analysis was carried out with an explanation of how lost-to-follow-up was handled.

Table 3.

Summary of Critical Appraisal Assessed by JBI Critical Appraisal Checklist for Quasi-Experimental Studies, Including Assessment of the Qualitative Part of Mixed Method Studies.

Study	1	2	3	4	5	6	7	8	9	Risk of bias
Tamez-Pérez et al³⁰	+	+^a	+	-	+	-	+	+	+	Low
Aradóttir et al³¹	+	+^a	+	-	+	+	+	+	-	Low
McGloin et al^{57 b}	+	+^a	U	U	+	+	U	+	+	Moderate
Zhang et al^{58 b}	+	+	+	+	+	-	U	U	+	Moderate
Levy et al⁵⁹	+	+^a	+	-	+	+	+	+	+	Low
Hasan et al³⁴	+	+^a	+	-	+	-	+	+	+	Low
Pfützner et al³²	+	U	+	U	-	+	+	+	U	Moderate
Larsen et al³³	+	+^a	+	-	+	+	+	+	+	Low
Ngassa Piotie et al³⁵	+	+^a	+	-	+	+	+	+	+	Low
Sethi et al³⁶	+	+^a	+	-	+	+	+	+	-	Low

U = Unclear, + = Yes, and - = No. Question 2: red marks visual inspection of between-group differences in baseline characteristics of the population to determine if the groups were similar.

Single-arm study.

Mixed method study.

Table 4.

Summary of Critical Appraisal Assessed by JBI Critical Appraisal Checklist for Qualitative Studies, Including Assessment of the Qualitative Part of Mixed Method Studies.

Study	1	2	3	4	5	6	7	8	9	10	Risk of bias
McGloin et al^{57 a}	+	+	+	+	+	-	-	+	+	+	Low
Rogers et al⁶⁰	-	U	U	U	U	-	-	+	+	+	High
Zhang et al^58a	U	U	U	U	U	-	-	U	+	U	High

U = Unclear, + = Yes, and - = No.

Mixed method study.

Table 5.

Summary of Critical Appraisal Assessed by JBI Critical Appraisal Checklist for Cohorts.

Study	1	2	3	4	5	6	7	8	9	10	11	Risk of bias
Seufert et al⁶¹	+	+	+	-	-	+	+	+	U	U	U	Moderate

U = Unclear, + = Yes, and - = No.

Table 6.

Summary of Critical Appraisal Assessed by the Checklist in Fone²⁶ for Simulation Studies.

Study	1	2	3	4	5	6	7	8	9	10	Overall
Krishnamoorthy et al¹⁰	2	2	1	2	2	1	1	1	1	1	B
Sieber et al⁶²	2	2	2	1	2	1	1	1	1	2	B
Aradóttir et al⁸	1	2	1	2	2	1	2	1	2	1	B

Scores that can be given to a question; 0, 1, or 2 (poor to good). Overall indicated the overall score; A, B, C, or D (high to low risk of bias).

Discussion

Summary of Evidence

The review aimed to provide an overview of dose guidance methods supporting basal insulin titration of people with T2D and categorize these according to characteristics, effects, and user experience. Overall results showed three categories of methods: paper-based titration algorithms, telehealth solutions, and mathematical models. Most studies investigated implementations of paper-based titration algorithms. Studies investigating digital solutions for basal insulin titration for people with T2D were limited to simple telehealth solutions and, in one case, a mathematical model embedded into a decision support system. In summary, all studies used titration algorithms either in paper form or digital, except for the mathematical models.

Similar findings are seen in Deerochanawong et al,¹⁹ which highlighted use of paper-based titration algorithms and telehealth solutions when investigating titration of insulin glargine 100 U/mL in an Asian population. However, use of mathematical models was not reported. Furthermore, Kerr et al⁶⁸ found indications for improved glycemic control when using digital solutions to manage T2D treatment compared to standard of care. This is further supported by Hangaard et al,⁶⁹ which found a significant improvement in HbA1c when using telemedicine among people with T2D. These studies did not focus on basal insulin titration but overall treatment of people with T2D. However, it is feasible to assume that a similar effect may be seen when using telemedicine for titrating basal, which aligns with the tendency observed in this review.

User experience was not investigated thoroughly by included studies. Yet, common characteristics were the wish of people with T2D for simple and easy-to-use solutions and HCPs’ attention to effect on workload. Concerning telehealth solutions, HCPs, in some cases, uttered concern about increased data being generated compared to standard practice affecting workload.⁵⁷ None of the studies investigating mathematical models looked at user experience. Consideration of user experience when developing methods for basal insulin dose guidance is essential to ensure a holistic solution aimed at the intended end-user and thereby to secure effect in a real-world setting.⁷⁰ Especially considering solutions aimed at people with T2D due to known issues of non-adherence to treatment.^71,72

Strengths and Limitations

The broad scope and comprehensive literature strengthen the present systematic review. However, relevant studies may have been overlooked since the search was limited to use of basal insulin analogs and English, Danish, Norwegian, and Swedish language.

The heterogeneity of reported glycemic outcomes and differences in study design complicated comparison of effect. Validity of the review is weakened since mainly one reviewer screened the search results. To minimize this effect, co-authors were continuously consulted to clarify doubts about inclusion of studies and during critical appraisal. Furthermore, the review was strengthened since the structured search was performed with assistance from a research librarian, ensuring a thorough search.

Implications for Future Research

Mathematical models were limited to three studies which were mainly evaluated through simulation. Expect a study by Aradóttir et al,³¹ where the solution was tested on eight participants showing promising results. The limited use of mathematical models may be due to the complex nature of T2D and heterogeneity of the population caused by varying insulin sensitivity and production. This complicates modeling of insulin’s effect on blood glucose. The modeling task is further complicated by the limited available information about people with T2D. Glucose measures are typically performed using glucometers, and frequency of these measures varies depending on the individual in question.

In contrast, people with type 1 diabetes more often use continuous glucose monitoring to measure blood glucose, enabling more thorough insight into blood glucose levels throughout the day.^73
-75 Similar challenges have been recognized by studies addressing mathematical models.^8,10

New technologies enabling improved data collection might ease some of the challenges in modeling insulin’s effect on blood glucose levels for people with T2D using mathematical models. Kerr et al⁶⁸ highlight that new technology that supports improved data capturing may facilitate better treatment support when combined with dose recommendation software. Furthermore, addition of automated data-driven dose guidance might help rectify the increased workload for HCP that, in some cases, has been reported when introducing new technology.⁷⁶

At the time of writing, machine learning methods, which are frenquently used in other areas of diabetes reasearch, have focused on detection or prediction of hypoglycemic events, blood glucose levels, and optimal bolus insulin dosing in relation to people with T2D.⁷⁷ In the future, exploring the capability of machine learning methods for basal insulin dose guidance for people with T2D may provide insight into the field that could pioneer future research.

Conclusions

Three basal insulin dose guidance categories aimed at people with T2D were identified: paper-based titration algorithms, telehealth solutions, and mathematical models. Compared to paper-based titration algorithms, a numerically larger proportion of participants reached a predefined target using telehealth solutions.

Few studies investigated user experience. Some studies underlined a possible increase in workload when using telehealth solutions due to increased data. Futhermore, it was found that people with T2D preferred simple and easy-to-use solutions and fewer in-person visits.

Future work might benefit from exploring the capabilities of machine learning methods for basal insulin dose guidance for people with T2D, focusing on a simple and easy-to-use solution that does not increase the workload for HCPs.

Supplemental Material

sj-xlsx-1-dst-10.1177_19322968221145964 – Supplemental material for Time for Using Machine Learning for Dose Guidance in Titration of People With Type 2 Diabetes? A Systematic Review of Basal Insulin Dose Guidance

Supplemental material, sj-xlsx-1-dst-10.1177_19322968221145964 for Time for Using Machine Learning for Dose Guidance in Titration of People With Type 2 Diabetes? A Systematic Review of Basal Insulin Dose Guidance by Camilla Heisel Nyholm Thomsen, Stine Hangaard, Thomas Kronborg, Peter Vestergaard, Ole Hejlesen and Morten Hasselstrøm Jensen in Journal of Diabetes Science and Technology

Supplemental Material

sj-xlsx-2-dst-10.1177_19322968221145964 – Supplemental material for Time for Using Machine Learning for Dose Guidance in Titration of People With Type 2 Diabetes? A Systematic Review of Basal Insulin Dose Guidance

Supplemental material, sj-xlsx-2-dst-10.1177_19322968221145964 for Time for Using Machine Learning for Dose Guidance in Titration of People With Type 2 Diabetes? A Systematic Review of Basal Insulin Dose Guidance by Camilla Heisel Nyholm Thomsen, Stine Hangaard, Thomas Kronborg, Peter Vestergaard, Ole Hejlesen and Morten Hasselstrøm Jensen in Journal of Diabetes Science and Technology

Footnotes

Acknowledgements

The authors thank research librarian Connie Skrubbeltrang for competent assistance in the literature search.

Abbreviations

DTSQ, Diabetes Treatment Satisfaction Questionnaire; HCP, healthcare professional; JBI, Joanna Briggs Institute; PRISMA, preferred reporting items for systematic reviews and meta-analyses; PROSPERO, International Prospective Register of Systematic Reviews; RCT, randomized controlled trial; T2D, type 2 diabetes.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Author P.V. is head of research at the Steno Diabetes Center North Denmark, funded by the Novo Nordisk Foundation. Author M.H.J is a former Novo Nordisk employee and holds Novo Nordisk shares.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Camilla Heisel Nyholm Thomsen

Stine Hangaard

Thomas Kronborg

Morten Hasselstrøm Jensen

Supplemental Material

Supplemental material for this article is available online.

References

Reach

Pechtner

Gentilella

Corcos

Ceriello

Clinical inertia and its impact on treatment intensification in people with type 2 diabetes mellitus. Diabetes Metab. 2017;43(6):501-511.

Chun

Strong

Urquhart

Insulin initiation and titration in patients with type 2 diabetes. Diabetes Spectr Publ Am Diabetes Assoc. 2019;32(2):104-111.

Mocarski

Yeaw

Divino

, et al. Slow titration and delayed intensification of basal insulin among patients with type 2 diabetes. J Manag Care Spec Pharm. 2018;24(4):390-400.

Dansk Endokrinologisk Selskab. Date unknown. Type 2 diabetes. https://endocrinology.dk/nbv/diabetes-melitus/behandling-og-kontrol-af-type-2-diabetes/. Accessed December 8, 2022.

Lingvay

Rhee

Raskin

. Type 2 diabetes mellitus: an evidence-based approach to practical management. In: Feinglos

Bethel

, eds. Type 2 Diabetes Mellitus. (Contemporary Endocrinology). Totowa, NJ: Humana Press; 2008:151-167. doi:10.1007/978-1-60327-043-4_10.

Zhang

Zhu

, et al. Observational Registry of Basal Insulin Treatment (ORBIT) in patients with type 2 diabetes uncontrolled with oral antihyperglycaemic drugs: real-life use of basal insulin in China. Diabetes Obes Metab. 2017;19(6):822-830.

Thomsen

Baggesen

Søgaard

, et al. Early glycaemic control in metformin users receiving their first add-on therapy: a population-based study of 4,734 people with type 2 diabetes. Diabetologia. 2015;58(10):2247-2253.

Aradóttir

Boiroux

Bengtsson

, et al. Model predictive control for dose guidance in long acting insulin treatment of type 2 diabetes. IFAC J Syst Control. 2019;9:100067.

Khunti

Gomes

Pocock

, et al. Therapeutic inertia in the treatment of hyperglycaemia in patients with type 2 diabetes: a systematic review. Diabetes Obes Metab. 2018;20(2):427-437.

10.

Krishnamoorthy

Boiroux

Aradóttir

Engell

Jørgensen

JB.

A model-free approach to automatic dose guidance in long acting insulin treatment of type 2 diabetes. IEEE Control Syst Lett. 2021;5(6):2030-2035.

11.

Forst

Choudhary

Schneider

Linetzky

Pozzilli

A practical approach to the clinical challenges in initiation of basal insulin therapy in people with type 2 diabetes. Diabetes Metab Res Rev. 2021;37(6):e3418.

12.

Bode

Clarke

Johnson

Use of decision support software to titrate multiple daily injections yielded sustained A1c reductions after 1 year. J Diabetes Sci Technol. 2018;12(1):124-128.

13.

Bergenstal

Johnson

Passi

, et al. Automated insulin dosing guidance to optimise insulin management in patients with type 2 diabetes: a multicentre, randomised controlled trial. Lancet Lond Engl. 2019;393(10176):1138-1148.

14.

Khunti

Nikolajsen

Thorsted

Andersen

Davies

Paul

SK.

Clinical inertia with regard to intensifying therapy in people with type 2 diabetes treated with basal insulin. Diabetes Obes Metab. 2016;18(4):401-409.

15.

Khunti

Giorgino

Berard

Mauricio

Harris

SB.

The importance of the initial period of basal insulin titration in people with diabetes. Diabetes Obes Metab. 2020;22(5):722-733.

16.

Blonde

Brunton

Chava

, et al. Achievement of target A1C <7.0% (<53 mmol/mol) by U. S. Type 2 diabetes patients treated with basal insulin in both randomized controlled trials and clinical practice. Diabetes Spectr. 2019;32(2):93-103.

17.

Meneghini

Mauricio

Orsi

, et al. The Diabetes Unmet Need with Basal Insulin Evaluation (DUNE) study in type 2 diabetes: achieving HbA1c targets with basal insulin in a real-world setting. Diabetes Obes Metab. 2019;21(6):1429-1436.

18.

Pscherer

Anderten

Pfohl

, et al. Titration of insulin glargine 100 U/mL when added to oral antidiabetic drugs in patients with type 2 diabetes: results of the TOP-1 real-world study. Acta Diabetol. 2020;57(1):89-99.

19.

Deerochanawong

Leelawattana

Kosachunhanun

Tantiwong

Basal insulin dose titration for glycemic control in patients with type 2 diabetes mellitus in Thailand: results of the REWARDS real-world study. Clin Med Insights Endocrinol Diabetes. 2020;13:1179551420935930.

20.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

21.

Thomsen

CHN

Hangaard

Kronborg

Hejlesen

Vestergaard

Jensen

. Dose guidance systems for basal insulin titration of patients with Type 2 Diabetes: a protocol of a systematic review. International Prospective Register of Systematic Reviews 2021 CRD42021289364. https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42021289364. Published 2021. Accessed December 8, 2022.

22.

Joanna Briggs Institute. Critical appraisal tools. Date unknown. https://jbi.global/critical-appraisal-tools. Accessed December 8, 2022.

23.

Andrews

Likis

FE.

Study design algorithm. J Low Genit Tract Dis. 2015;19(4):364-368.

24.

Joanna Briggs Institute. JBI manual for evidence synthesis. Date unknown. https://jbi-global-wiki.refined.site/space/MANUAL. Accessed December 8, 2022.

25.

Melo

Dutra

Rodrigues Filho

, et al. Association between psychotropic medications and presence of sleep bruxism: a systematic review. J Oral Rehabil. 2018;45(7):545-554.

26.

Fone

Hollinghurst

Temple

, et al. Systematic review of the use and value of computer simulation modelling in population health and health care delivery. J Public Health Med. 2003;25(4):325-335.

27.

Simon

ACR

Holleman

Gude

Hoekstra

JBL

Peek

. Safety of a web-based insulin titration system for patients with type 2 diabetes mellitus–pilot study. Qual Life Qual Inf. 2012;180:731-735.

28.

Practical Diabetes. Diabetes Insulin Guidance System: a real-world evaluation of new technology (d-Nav) to achieve glycaemic control in insulin-treated type 2 diabetes. https://wchh.onlinelibrary.wiley.com/doi/full/10.1002/pdi.1966. Published 2015. Accessed December 8, 2022.

29.

Cook

Mann

King

, et al. Management of insulin therapy in urban diabetes patients is facilitated by use of an intelligent dosing system. Diabetes Technol Ther. 2004;6(3):326-335.

30.

Tamez-Pérez

Deceased Author, Cantú -Santos

Gutierrez- González

González-Facio

Romero-Ibarguengoitia

ME.

Effect of digital-tool-supported basal insulin titration algorithm in reaching glycemic control in patients with type 2 diabetes in Mexico. J Diabetes Sci Technol. 2022;16:1513-1520.

31.

Aradóttir

Bengtsson

Jensen

, et al. Feasibility of a new approach to initiate insulin in type 2 diabetes. J Diabetes Sci Technol. 2021;15(2):339-345.

32.

Pfützner

Stratmann

Funke

, et al. Real-world data collection regarding titration algorithms for insulin glargine in patients with type 2 diabetes mellitus. J Diabetes Sci Technol. 2016;10(5):1122-1129.

33.

Larsen

Turner

Farmer

Neil

Tarassenko

Telemedicine-supported insulin optimisation in primary care. J Telemed Telecare. 2010;16(8):433-440.

34.

Hasan

Amer

Junaid

Practical implementation of ADA/EASD consensus algorithm in patients with type 2 diabetes in Pakistan. J Pak Med Assoc. 2018;68(9):1304-1309.

35.

Ngassa Piotie

Wood

Muchiri

Webb

Rheeder

Using a nurse-driven and home-based telehealth intervention to improve insulin therapy for people with type 2 diabetes in primary care: a feasibility study. J Endocrinol Metab Diabetes South Afr. 2022;27(3):108-116.

36.

Sethi

Al- Rubeaan

Unubol

, et al. Efficacy and safety of insulin glargine 300 U/mL in people with type 2 diabetes uncontrolled on basal insulin: the 26-week interventional, sin gle-arm ARTEMIS-DM study. Diabetes Ther Res Treat Educ Diabetes Relat Disord. 2022;13(7):1395-1408.

37.

Yuan

Fengfei

Zhou

, et al. Fasting glucose of 6.1 mmol/L as a possible optimal target for type 2 diabetic patients with insulin glargine: a randomized clinical trial. Date unknown. https://www.hindawi.com/journals/jdr/2021/5524313/. Accessed December 8, 2022.

38.

Bae

Ahn

Yang

, et al. Efficacy and safety of self-titration algorithms of insulin glargine 300 units/mL in individuals with uncontrolled type 2 diabetes mellitus (the Korean TITRATION study): a randomized controlled trial. Diabetes Metab J. 2022;46(1):71-80.

39.

Deng

Zhang

, et al. Efficacy and safety of a decision support intervention for basal insulin self-titration assisted by the nurse in outpatients with T2DM: a randomized controlled trial. Diabetes Metab Syndr Obes. 2021;14:1315-1327.

40.

Park

Hong

, et al. Comparison of the efficacy and safety of insulin detemir administered once daily according to two titration algorithms (3-0-3 and 2-4-6-8) in patients with type 2 diabetes mellitus. Endocrinol Metab (Seoul). 2020;35(1):142-148.

41.

Misra

Patel

Agarwal

, et al. Effectiveness and safety of physician-led versus patient-led titration of insulin glargine in Indian patients with type 2 diabetes mellitus: a subanalysis of the Asian Treat to Target Lantus Study (ATLAS). Diabetes Technol Ther. 2019;21(11):656-664.

42.

Yale

Berard

Groleau

Javadi

Stewart

Harris

SB.

TITRATION: a randomized study to assess 2 treatment algorithms with new insulin glargine 300 units/mL. Can J Diabetes. 2017;41(5):478-484.

43.

Kadowaki

Jinnouchi

Kaku

Hersløv

Hyllested-Winge

Nakamura

Insulin degludec in a simple or stepwise titration algorithm in a Japanese population of patients with type 2 diabetes: a randomized, 26-week, treat-to-target trial. Diabetol Int. 2017;8(1):87-94.

44.

Bajaj

Venn

Aronson

Randomized trial of long-acting insulin glargine titration web tool (LTHome) versus enhanced usual therapy of glargine titration (INNOVATE trial). Diabetes Technol Ther. 2016;18(10):610-615.

45.

Hsu

Lau

Huang

, et al. Utilization of a cloud-based diabetes management program for insulin initiation and titration enables collaborative decision making between healthcare providers and patients. Diabetes Technol Ther. 2016;18(2):59-67.

46.

Philis-Tsimikas

Brod

Niemeyer

Ocampo Francisco

Rothman

Insulin degludec once-daily in type 2 diabetes: simple or stepwise titration (BEGIN: once simple use). Adv Ther. 2013;30(6):607-622.

47.

Kim

Park

Kang

, et al. Insulin dose titration system in diabetes patients using a short messaging service automatically produced by a knowledge matrix. Diabetes Technol Ther. 2010;12(8):663-669.

48.

Blonde

Merilainen

Karwe

Raskin

; TITRATE Study Group. Patient—directed titration for achieving glycaemic goals using a once—daily basal insulin analogue: an assessment of two different fasting plasma glucose targets—the TITRATE study. Diabetes Obes Metab. 2009;11(6):623-631.

49.

Kennedy

Herman

Strange

Harris

; GOAL AIC Team. Impact of active versus usual algorithmic titration of basal insulin and point-of-care versus laboratory measurement of HbA1c on glycemic control in patients with type 2 diabetes: the Glycemic Optimization with Algorithms and Labs at Point of Care (GOAL A1C) trial. Diabetes Care. 2006;29(1):1-8.

50.

Lingvay

Buse

Franek

, et al. A Randomized, open-label comparison of once-weekly insulin icodec titration strategies versus once-daily insulin glargine U100. Diabetes Care. 2021;44(7):1595-1603.

51.

Garg

Admane

Freemantle

, et al. Patient-led versus physician-led titration of insulin glargine in patients with uncontrolled type 2 diabetes: a randomized multinational ATLAS study. Endocr Pract. 2015;21(2):143-157.

52.

Meneghini

Koenen

Weng

Selam

JL.

The usage of a simplified self-titration dosing guideline (303 algorithm) for insulin detemir in patients with type 2 diabetes–results of the randomized, controlled PREDICTIVE 303 study. Diabetes Obes Metab. 2007;9(6):902-913.

53.

Wan

Wen

, et al. Higher versus standard starting dose of insulin glargine 100 U/mL in overweight or obese Chinese patients with type 2 diabetes: results of a multicentre, open-label, randomized controlled trial (BEYOND VII). Diabetes Obes Metab. 2020;22(5):838-846.

54.

Levy

Moynihan

Nilo

, et al. The mobile insulin titration intervention (MITI) for insulin adjustment in an urban, low-income population: randomized controlled trial. J Med Internet Res. 2015;17(7):e180.

55.

Davies

Bain

Charpentier

, et al. A Randomized controlled, treat-to-target study evaluating the efficacy and safety of insulin glargine 300 U/mL (Gla-300) administered using either device-supported or routine titration in people with type 2 diabetes. J Diabetes Sci Technol. 2019;13(5):881-889.

56.

Ishii

Nakajima

Kamei

, et al. Comparison of patient-led and physician-led insulin titration in Japanese type 2 diabetes mellitus patients based on treatment distress, satisfaction, and self-efficacy: the COMMIT-patient study. Diabetes Ther Res Treat Educ Diabetes Relat Disord. 2021;12(2):595-611.

57.

McGloin

O’Connell

Glacken

, et al. Patient empowerment using electronic telemonitoring with telephone support in the transition to insulin therapy in adults with type 2 diabetes: observational, pre-post, mixed methods study. J Med Internet Res. 2020;22(5):e16161.

58.

Zhang

Zhao

, et al. Lack of coordination between partners: investigation of Physician-Preferred and Patient-Preferred (4P) basal insulin titration algorithms in the real world. Patient Prefer Adherence. 2018;12:1253-1259.

59.

Levy

Orzeck-Byrnes

Aidasani

, et al. Transition of a text-based insulin titration program from a randomized controlled trial into real-world settings: implementation study. J Med Internet Res. 2018;20(3):e93.

60.

Rogers

Aidasani

Friedes

, et al. Barriers and facilitators to the implementation of a mobile insulin titration intervention for patients with uncontrolled diabetes: a qualitative analysis. JMIR Mhealth Uhealth. 2019;7(7):e13906.

61.

Seufert

Fritsche

Pscherer

, et al. Titration and optimization trial for the initiation of insulin glargine 100 U/mL in patients with inadequately controlled type 2 diabetes on oral antidiabetic drugs. Diabetes Obes Metab. 2019;21(2):439-443.

62.

Sieber

Weinheimer

Kongable

Riddle

Chang

Flacke

In silico examination of initiation of long-acting insulin analogs toujeo compared to lantus under 3 dosing titration rules in virtual type 2 diabetes subjects. J Diabetes Sci Technol. 2020;14(5):898-907.

63.

Jones

Lonergan

Henley

Pearson

Hattersley

Shields

BM.

Should studies of diabetes treatment stratification correct for baseline HbA1c?

PLoS ONE. 2016;11(4):e0152428.

64.

Park

Cho

, et al. Influence of the duration of diabetes on the outcome of a diabetes self-management education program. Diabetes Metab J. 2012;36(3):222-229.

65.

Cummings

Cao

Hadjiyianni

Ilag

Tan

MH.

Characteristics of insulin-naïve people with type 2 diabetes who successfully respond to insulin glargine U100 after 24 weeks of treatment: a meta-analysis of individual participant data from 3 randomized clinical trials. Clin Diabetes Endocrinol. 2018;4:10.

66.

Lovre

Fonseca

Benefits of timely basal insulin control in patients with type 2 diabetes. J Diabetes Complications. 2015;29(2):295-301.

67.

Aso

Suzuki

Chiba

, et al. Effect of insulin degludec versus insulin glargine on glycemic control and daily fasting blood glucose variability in insulin-naïve Japanese patients with type 2 diabetes: I’D GOT trial. Diabetes Res Clin Pract. 2017;130:237-243.

68.

Kerr

Edelman

Vespasiani

Khunti

New digital health technologies for insulin initiation and optimization for people with type 2 diabetes. Endocr Pract. 2022;28(8):811-821.

69.

Hangaard

Laursen

Andersen

, et al. The effectiveness of telemedicine solutions for the management of type 2 diabetes: a systematic review, meta-analysis, and meta-regression. J Diabetes Sci Technol. 2021.

70.

Fico

Hernanzez

Cancela

, et al. What do healthcare professionals need to turn risk models for type 2 diabetes into usable computerized clinical decision support systems? lessons learned from the MOSAIC project. BMC Med Inform Decis Mak. 2019;19(1):163.

71.

Krass

Schieback

Dhippayom

Adherence to diabetes medication: a systematic review. Diabet Med J Br Diabet Assoc. 2015;32(6):725-737.

72.

Blackburn

Swidrovich

Lemstra

Non-adherence in type 2 diabetes: practical considerations for interpreting the literature. Patient Prefer Adherence. 2013;7:183-189.

73.

Garg

SK.

Emerging landscape of continuous glucose monitoring. Diabetes Technol Ther. 2021;23(S3):S1-S4.

74.

Kompala

Neinstein

A new era: increasing continuous glucose monitoring use in type 2 diabetes. Am J Manag Care. 2019;25(4):SP123-SP126. https://www.ajmc.com/view/a-new-era-increasing-continuous-glucose-monitoring-use-in-type-2-diabetes-. Accessed December 8, 2022.

75.

Slattery

Choudhary

Clinical use of continuous glucose monitoring in adults with type 1 diabetes. Diabetes Technol Ther. 2017;19(S2):S55-S61.

76.

Simon

Holleman

Gude

, et al. Safety and usability evaluation of a web-based insulin self-titration system for patients with type 2 diabetes mellitus. Artif Intell Med. 2013;59(1):23-31.

77.

Contreras

Vehi

Artificial intelligence for diabetes management and decision support: literature review. J Med Internet Res. 2018;20(5):e10775.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.61 MB

0.01 MB