Time to Insulin Initiation in Type 2 Diabetes Patients in 2010/2011 and 2016/2017 in Germany

Abstract

Background:

The aim of the current study was to determine whether the time to insulin therapy initiation in patients with type 2 diabetes in primary care in Germany has changed in recent years.

Methods:

Longitudinal data from general practices in Germany (Disease Analyzer) were analyzed. These data included information of 7128 patients (age: 68.5 [SD: 11.5] years; 54.4% male) receiving incident insulin therapy in 2010/2011 and 8216 patients (age: 69.1 [SD: 11.9] years; 54.9% male) receiving incident insulin therapy in 2016/2017. Changes in time to insulin initiation in the practices and the last HbA1c value before the start of insulin therapy were analyzed, stratified by index date. To analyze the impact of covariables on the time to insulin initiation, a multivariate regression analysis was performed, adjusted for age, sex, diabetologist care, and HbA1c as independent variables.

Results:

Median time from T2D diagnosis to insulin therapy in the Disease Analyzer database increased from 1717 days in 2010/2011 to 1917 days in 2016/2017 (P < .001). The proportion of patients with a HbA1c value ≥9% before insulin initiation was high in both groups (2010/2011: 33.0%, 2016/2017: 34.2%, P = .347). The time to insulin initiation in DPP-4i patients was 112 days longer, and in SGLT2 patients 346 days longer than in patients treated with sulfonylurea.

Conclusions:

The present analysis confirms an increasing delay of the insulin therapy initiation as a consequence of the more frequent use of newer oral antidiabetics. However, the rather moderate increase of time to insulin might display insufficient long-term glycemic control using these agents. Still, more than one-third of patients receive insulin only when HbA1c levels exceed 9%.

Keywords

insulin therapy primary care glycemic control DPP-4 SGLT2

Treatment of type 2 diabetes with lifestyle interventions and antidiabetic drugs aims at preventing or delaying complications while maintaining quality of life. Cornerstones of the treatment are both control of glycemia and cardiovascular risk factor management. Treatment decision should always follow a patient-centered approach.¹ The initiation of insulin therapy in type 2 diabetes is often delayed for a considerable time period—not least for fear of weight gain or hypoglycemia.²

Avoidance or considerable delay of insulin treatment is reported to be associated with a progression of hyperglycemia and insulin resistance which finally may increase the risk of macro- and microvascular complications.³ Many general practitioners perceive insulin therapy as too complex to manage in a busy primary care practice, resulting in additional delays of required therapy intensification.⁴

Throughout the last decade, novel antidiabetic agents have been launched on the market, including dipeptidyl peptidase 4 (DPP-4) inhibitors,⁵ glucagon-like peptide-1 (GLP-1) receptor agonists (GLP-1-RA),⁶ and sodium glucose cotransporter 2 (SGLT-2) inhibitors (SGLT-2i ).⁷ Use of SGLT-2i therapy has been shown to be associated with a reduction in heart failure risk in patients with T2D.⁸ GLP-1-RA are associated with a weight reduction in T2D patients.⁹ Several Cardiovascular Outcome Studies in patients with type 2 diabetes had a profound impact on recent treatment guidelines.^10-12

In 2013, Kostev and Rathmann showed that the time to insulin therapy initiation in primary care T2D patients significantly increased from 2005 to 2010 in Germany. Using data from the Disease Analyzer database, median time to insulin therapy in the practices increased from 943 days in 2005 to 1549 days in 2010 in Germany. In the United Kingdom, time to onset of insulin treatment increased from 1700 days in 2005 to 2061 days in 2010. This could be due to a preference given to prescriptions of novel antidiabetic agents instead of or prior to insulin. Moreover, the last HbA1c values prior to insulin initiation were high and even increased during the study period (Germany: 2005: 8.2%, 2010: 8.4%; UK: 2005: 9.5%, 2010: 9.8%; both P < .001), indicating insufficient blood glucose control.^13,14

There are no studies available to investigate whether the time to insulin therapy initiation in general practices has further increased during the time period from 2010 to 2017. Thus, the aim of the current study was to determine whether the time to insulin therapy initiation in type 2 diabetes patients in primary care in Germany has further increased in recent years.

Methods

Database

This study is based on data from the Disease Analyzer database (IQVIA), which compiles drug prescriptions, diagnoses, basic medical, and demographic data obtained directly and in anonymous format from computer systems used in the practices of general practitioners and specialists.¹⁵ Diagnoses (International Classification of Diseases, 10th revision [ICD-10]), prescriptions (Anatomical Therapeutic Chemical [ATC] Classification system), and the quality of reported data are being monitored by IQVIA based on a number of criteria (eg, completeness of documentation and linkage between diagnoses and prescriptions).

In Germany, the sampling methods used for the selection of physicians’ practices were sufficient to produce a representative database of general and specialized practices.¹⁵ Finally, this database has already been used in several studies focusing on diabetes.^16,17

Inclusion Criteria

The analyzed database period was January 1995 to December 2017. The first type 2 diabetes diagnosis (ICD: E11) in the practice records was defined as the index date. All subjects with a first-time insulin prescription (ATC: A10C) were selected. Further inclusion criteria were (1) continuous treatment in the same practice (≥1 visit during the twelve months prior to the index date and ≥1 visit each year for at least 1 year after the index date), (2) age of over 40 years at index date, and (3) HbA1c value documented within 6 months prior to the first insulin prescription. Patients with prescriptions of insulin at the same point as their first diagnosis of diabetes or within one year after the index date were excluded.

Outcomes

The main outcome measures were (1) the time to initiation of insulin therapy and (2) last HbA1c value before start of insulin therapy. Patients who started their insulin treatment in the years 2010/2011 and 2016/2017 were analyzed to determine the time from the first diagnosis of diabetes in the practices (index date) and the first insulin prescription. Patients were retrospectively observed for a maximum of 13 years to ensure comparability in terms of time to insulin therapy between patients who initiated insulin treatment in 2010/2011 versus those who began it in 2016/2017. We carried out a sensitivity analysis by estimating the increase in HbA1c prior to insulin initiation. In order to do this, the difference between the last value prior to insulin therapy (0-183 days) and the first value in the time period 184-365 days prior to insulin therapy were calculated.

In addition, the reduction in HbA1c levels in the 9 months following insulin initiation was estimated. To analyze the change in HbA1c levels, the mean value of the difference between the last HbA1c value prior to insulin initiation and the last HbA1c value in the period 91-270 days following insulin initiation was calculated.

Covariables

Therapy with noninsulin antidiabetic agents (metformin, sulfonylurea, DPP-4 inhibitors, GLP-1 receptor agonists, SGLT2 inhibitors and other antidiabetics, including alpha-glucosidase-inhibitors, glinides, and glitazones), age, sex, physician specialty: diabetologist care, and last HbA1c value documented within the 6 months prior to insulin initiation were included as independent variables.

Statistical Methods

Descriptive analyses were obtained for all variables. Means (SD) were calculated for normally distributed continuous variables (time to insulin) and the medians (interquartile range) were determined for continuous variables that were not distributed normally. A multivariate regression analysis was performed with time to insulin as dependent and antihyperglycemic therapy, age, sex, and HbA1c as independent variables to analyze the impact of antihyperglycemic therapy, age, sex, and HbA1c on the time to insulin. Metformin was used as the reference group when determining the associations of the various antidiabetic agents with time to insulin initiation. In the subgroup of patients treated with at least two antidiabetic agents prior to insulin, sulfonylurea was used as the reference group. Two-tailed tests were used and a P value of <.05 was considered statistically significant. All analyses were carried out using SAS 9.4 (SAS Institute, Cary, NC, USA).

Results

Patient Characteristics

The clinical characteristics of patients with type 2 diabetes receiving incident insulin therapy (2010/2011 versus 2016/2017) in primary care practices in Germany are shown in Table 1.

Table 1.

Clinical Characteristics of Primary Care Patients With Type 2 Diabetes Receiving Incident Insulin Therapy (2010 Versus 2016): Disease Analyzer Germany.

	2010-2011	2016-2017	P value
N	7128	8216
Age (mean, SD)	68.5 (11.5)	69.1 (11.9)	.005
Sex, % male	54.4	54.9	.640
Last antihyperglycemic therapy (prior to insulin) (%)
Metformin (mono)	17.9	18.7	<.001
Metformin + DPP-4	11.7	41.4	<.001
Metformin + sulfonylurea	36.3	19.6	<.001
Metformin + TZD	8.4	2.8	<.001
Metformin + GLP-1	1.8	3.2	<.001
Metformin + SGLT2	0.0	3.8	<.001
Sulfonylurea (mono)	9.5	3.5	<.001
Sulfonylurea + DPP-4	1.5	2.8	<.001
All other combinations or mono therapies	14.0	5.7	<.001

Data are listed as means (SD) or prevalence (%). P values: chi-square tests or Wilcoxon tests (for age).

There was no significant difference in age and sex between patients who had initiated insulin in 2010/2011 and those who had initiated it in 2016/2017. By contrast, there has been substantial change in the last 6 years with regard to the last antihyperglycemic therapy prior to insulin initiation.

The proportion of patients receiving a combination of metformin + DPP-4i increased from 11.7% to 41.4% (P < .001), that of patients receiving metformin + SGLT-2 from 0% to 3.8% (P < .001), and that of patients receiving metformin + GLP-1 from 1.8% to 3.2% (P < .001). The proportion of sulfonylurea monotherapy users decreased from 9.5% to 3.5% (P < .001), while patients receiving a combination of metformin + sulfonylurea decreased from 36.3% to 19.6% (P < .001), and those receiving metformin + TZD from 8.4% to 2.8% (P < .001)

Time to Insulin

There was a significant difference in the time from first diabetes diagnosis to start of insulin therapy between 2010/2011 (median duration = 1717 days) and 2016/2017 (median duration = 1917) days (Table 2).

Table 2.

Changes in Time to Insulin Initiation in Type 2 Diabetes Patients in Primary Care Stratified by Index Date (2010/2011 Versus 2016/2017): Disease Analyzer (Germany).

	2010-2011	2016-2017	P value
Total patients (N)	7128	8216
Mean HbA1c (%)*	8.5	8.6	.264
HbA1c >9%*	33.0	34.2	.347
Mean increase of HbA1c (%)**	0.8	0.8	.146
Increase ≥ 1%**	36.6	38.6	.127
Time to insulin (days): median (IQR)	1717 (928-2632)	1917 (1122-2950)	<.001

Data are listed as means (SD) or prevalence (%).

Last value prior to insulin therapy (0-183 days).

Difference between last value prior to insulin therapy (0-183 days) and first value in the time 184-365 days prior to insulin therapy.

IQR: interquartile range.

In total, 33.0% of the patients from 2010/2011 and 34.2% from 2016/2017 had poor glycemic control (defined as recorded HbA1c values of >9%) before starting insulin treatment. There was a slight, but insignificant, increase in the average last HbA1c values before start of insulin therapy between 2010/2011 and 2016/2017. Moreover, 12.6% and 10.0% of the patients from 2010/2011 and 13.5% and 10.3% from 2016/2017 had HbA1c value of 8-8.5% and 8.6-9%, respectively.

In an average period of six months prior to the last HbA1c value before insulin initiation, HbA1c increased by 0.8% during both time periods. A total of 36.6% of patients in 2010/2011 and 38.6% in 2016/2017 had an HbA1c increase of at least 1%.

Impact of Predefined Variables on the Time to Insulin Initiation

The association of predefined variables with the time to insulin initiation in patients with type 2 diabetes is shown in Table 3. Compared to individuals whose only therapy prior to insulin had been metformin monotherapy (reference group), patients with second- or third-line therapy were significantly more likely to start insulin therapy. The time to insulin increased by 276 days in sulfonylurea patients, 398 days in DPP-4 patients, and 637 days in SGLT2 patients (all P < .001). Among those patients who had received at least two different antihyperglycemic drugs prior to insulin, the time to insulin was 112 days longer in DPP-4i patients and 346 days longer in SGLT2 patients than in those given sulfonylurea. No significant increase in the time to insulin initiation was observed in patients who had received GLP-1-RA. Further significant variables included age, diabetologist care, and HbA1c value. An increase in HbA1c levels had no significant impact on the time to insulin initiation. Diabetologist treatment was associated with a decrease of 469 days in the time to insulin initiation. Older patients and patients with HbA1c values of 8.5% or higher received insulin therapy significantly earlier.

Table 3.

Association Between Oral Antidiabetic Agents, Age, Diabetologist Care, and HbA1c Values and the Time to Insulin Initiation in Type 2 Diabetes Patients in Primary Care (Germany, France, UK; Disease Analyzer Database).

Variable	Adjusted time difference in days	P value
Model 1: Number of additional days without insulin compared to metformin as the only therapy
Metformin	Reference
Sulfonylureas	+276	<.001
Acarbose, glinides, glitazones	+141	<.001
DPP-4 inhibitors	+398	<.001
GLP-1 analogs	+230	<.001
SGLT2	+637	<.001
Model 2: Number of additional days without insulin compared to sulfonylurea as second line therapy
Sulfonylurea	Reference
Acarbose, glinides, glitazones	−136	<.001
DPP-4 inhibitors	+112	<.001
GLP-1 analogs	−64	.183
SGLT2	+346	<.001
Further significant variables
Age (per year)	−12	<.001
Diabetologist care	−469	<.001
HbA1c ≥8.5	−29	.050
HbA1c ≥9.5	−94	<.001

Reference group: metformin. Linear regression models of time to insulin initiation (dependent variable), adjusted for age, sex, physician specialty: diabetologist care.

Change in HbA1c Levels

The HbA1c values in the period 91-270 days after insulin initiation was documented for 12 690 patients. In all patients on insulin, the mean HbA1c reduction was 1.1% (SD 1.8), from an average of 8.6% to an average of 7.5%. However, in patients with HbA1c values ≥8.5 prior to insulin initiation, the reduction was 2.1% (SD: 2.0), from an average of 10.2% to an average of 8.1%.

Discussion

This real-world data show that the time to initiation of insulin therapy in patients with type 2 diabetes continues to increase between 2010/2011 and 2016/2017 in primary care practices in Germany. The share of patients in which the last HbA1c value recorded before insulin initiation was ≥9% was high in both periods showing a slight increase in the 2016/2917 cohort. In more than one-third of the patients insulin is initiated at only such high HbA1c values.

One likely explanation is the use of novel antidiabetic agents during this period, including SGLT-2 inhibitors, which are associated with a strongest increase in the time to insulin initiation. The median time to insulin initiation increases when patients continue novel therapies for longer periods but also if a higher number of patients receive these therapies. For a large portion of T2D patients, drugs like DPP-4i and SGLT-2i lead to a delay of insulin therapy. However, the delay is rather moderate and suggests time limited efficacy of these agents.

The present study confirms clearly the trend seen also in other studies showing later initiation of insulin treatment at high HbA1c levels.¹³ It is unclear whether this trend might lead to an increased risk of developing diabetes-related complications.

The results are consistent with those of previous investigations, for example, in a large cross-sectional survey performed in Europe, only one-quarter of patients with type 2 diabetes reached the target value for adequate control (HbA1c <6.5%), despite various combinations of oral antidiabetic agents.^18,19 Although the proportion of patients who did not achieve the optimum HbA1c values has probably decreased in the last decade, there is still a clinically relevant number of T2D patients who might require insulin therapy at an earlier stage.

Laxy et al reported that the proportion of people with type 2 diabetes receiving oral antidiabetic/cardio-protective medication and those reaching treatment goals for glycemic control increased significantly between 2000 and 2014, but that the improvements observed were generally less significant between 2007 and 2014 than between 2000 and 2007.²⁰ Hosomura et al showed that the number of antidiabetic agents prescribed per patient increased over time, but that the HbA1c levels at which physicians stated they would introduce insulin also increased.²¹

In terms of the association between the use of novel antidiabetic agents and increased time to insulin initiation, our results are in line with a study carried out by Montvida et al.²² The authors estimated that the restricted mean time to insulin initiation was marginally longer in patients receiving second-line DPP-4i (7.1 years) than in those given sulfonylurea (6.3 years).²²

In our study the median time to insulin was between 4.7 and 5.3 years. However, it should be noted that patients with very long time to insulin (longer than 13 years) could not be included in the study (see methods). The share of patients who stay for longer than 13 years without insulin may be relatively high in the clinical practice. In the EDITION 3 study the time to insulin was 9.8 years, whereby patients with long time to insulin were included in the cohorts.²³

Our results also show that the time to insulin initiation in patients treated in diabetologist practices is shorter than in general practices. Diabetologists may be more likely to prescribe insulin than GPs due to their greater experience with diabetes treatment. On the other hand, diabetologists may tend to treat more severe cases, which require earlier insulin initiation

In order to help patients who need insulin, a greater effort must be made to remove the barriers currently preventing earlier initiation of insulin therapy in patients with type 2 diabetes. Moreover, it is important to identify these patients at an early stage and to distinguish them from T2D patients who can and should be treated with novel antidiabetic drugs, either as standalone therapy or in combination with insulin. Our study found an increase in HbA1c values of slightly over 1% in only 6 months. On the one hand, this increase was not associated with a shorter time to insulin initiation, but on the other, even in such patients the initiation of insulin six months earlier could have an impact on the drastic HbA1c reduction. As our results have shown, the mean HbA1c level increased by 1.1%, and by as much as 2.1% in patients with poor control.

Ahlqvist et al reported data from a large Scandinavian database, which may explain differences in the time to insulin therapy.²⁴

Retrospective primary care database analyses are generally limited by the validity and completeness of the data they contain. In particular, HbA1c values were only available for about 50% of patients shortly before insulin initiation, which is why only patients with known HbA1c values were included in this study. In addition, variables such as BMI, smoking status, or physical activity were not or were only rarely available. Finally, therapy-related details, for example, mono- or combination therapies but also codiagnoses or other cotherapies, were not analyzed.

In conclusion, the time to start of insulin therapy in primary care type 2 diabetes patients increased significantly in Germany between 2010/2011 and 2016/2017. The last HbA1c values recorded prior to insulin initiation were high in some cases. Further research is necessary to determine the underlying reasons for this trend.

Footnotes

Abbreviations

ATC, anatomical therapeutic chemical; BMI, body mass index; DPP-4i, dipeptidyl peptidase-4 inhibitor; GLP-1, glucagon-like peptide-1; ICD-10, International Classification of Diseases, 10th revision; SAS, Statistical Analysis Software; SD, standard deviation; SGLT2, sodium-glucose-co-transporter-2; T2D, type 2 diabetes; TZD, thiazolidinediones; UK, United Kingdom.

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: KK is an employee of IQVIA without conflict of interest. SG served as a consultant, advisor, and/or speaker for Abbott, Amgen, Animas, Berlin-Chemie, MSD, NovoNordisk, Novartis, Roche, Sanofi-Aventis. BMS served as a consultant, advisor, and/or speaker for AstraZeneca, Bayer, Berlin-Chemie Menarini, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Roche, and Sanofi. MK served as a consultant, advisor, and/or speaker for AstraZeneca, Boehringer Ingelheim, Lilly, MSD, Novo Nordisk. SP served as a consultant, advisor, and/or speaker for Berlin-Chemie, Boehringer Ingelheim, Eli Lilly, Medical Tribune, MSD, NovoNordisk, Novartis, OmniaMed, Sanofi-Aventis.

Funding

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

ORCID iD

Karel Kostev

References

Davies

D’Alessio

Fradkin

, et al. Management of hyperglycemia in type 2 diabetes, 2018. A consensus report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2018;41:2669-2701.

Pilla

Yeh

Juraschek

Clark

Maruthur

NM.

Predictors of insulin initiation in patients with type 2 diabetes: an analysis of the look AHEAD randomized trial. J Gen Intern Med. 2018;33:839-846. doi:10.1007/s11606-017-4282-9.

Paul

Klein

Thorsted

Wolden

Khunti

Delay in treatment intensification increases the risks of cardiovascular events in patients with type 2 diabetes. Cardiovasc Diabetol. 2015;14:100. doi:10.1186/s12933-015-0260-x.

Hirsch

Bergenstal

Parkin

Wright

Buse

JB.

A real-world approach to insulin therapy in primary care practice. Clin Diabetes. 2005;23:78-86.

Karagiannis

Paschos

Paletas

Matthews

Tsapas

Dipeptidyl peptidase-4 inhibitors for treatment of type 2 diabetes mellitus in the clinical setting: systematic review and meta-analysis. BMJ. 2012;344:e1369.

Owens

Monnier

Hanefeld

A review of glucagon-like peptide-1 receptor agonists and their effects on lowering postprandial plasma glucose and cardiovascular outcomes in the treatment of type 2 diabetes mellitus. Diabetes Obes Metab. 2017;19:1645-1654.

Santos

Lima

FJC

Sousa-Rodrigues

Barbosa

FT.

Use of SGLT-2 inhibitors in the treatment of type 2 diabetes mellitus. Rev Assoc Med Bras (1992). 2017;63:636-641. doi:10.1590/1806-9282.63.07.636.

Zinman

Wanner

Lachin

, et al. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med. 2015;373:2117-2128. doi:10.1056/NEJMoa1504720.

Sánchez-Garrido

Brandt

Clemmensen

Müller

DiMarchi

Tschöp

MH.

GLP-1/glucagon receptor co-agonism for treatment of obesity. Diabetologia. 2017;60:1851-1861. doi:10.1007/s00125-017-4354-8.

10.

Cefalu

Kaul

Gerstein

, et al. Cardiovascular outcomes trials in type 2 diabetes: where do we go from here? Reflections from a Diabetes Care Editors’ Expert Forum. Diabetes Care. 2018;41:14-31.

11.

Marso

Daniels

Brown-Frandsen

, et al.; LEADER Steering Committee; LEADER Trial Investigators. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2016;375:311-322.

12.

Marso

Bain

Consoli

, et al.; SUSTAIN- 6 Investigators. Semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2016;375:1834-1844.

13.

Jacob

Waehlert

Kostev

Changes in type 2 diabetes mellitus patients in German primary care prior to (2006) and after (2010, 2014) launch of new drugs. J Diabetes Sci Technol. 2016;10:414-420. doi:10.1177/1932296815607860.

14.

Kostev

Rathmann

Changes in time to insulin initiation in type 2 diabetes patients: a retrospective database analysis in Germany and UK (2005-2010). Prim Care Diabetes. 2013;7:229-233. doi:10.1016/j.pcd.2013.03.003.

15.

Rathmann

Bongaerts

Carius

Kruppert

Kostev

Basic characteristics and representativeness of the German Disease Analyzer database. Int J Clin Pharmacol Ther. 2018;56:459-466. doi:10.5414/CP203320.

16.

Rathmann

Bongaerts

Kostev

Association of characteristics of people with type 2 diabetes mellitus with discordant values of fasting glucose and HbA1c. J Diabetes. 2018;10:934-941. doi:10.1111/1753-0407.12823.

17.

Kostev

Jacob

Lucas

Rathmann

Low annual frequency of HbA1c testing in people with Type 2 diabetes in primary care practices in Germany. Diabet Med. 2018;35:249-254. doi:10.1111/dme.13556.

18.

Alvarez

Guisasola F

Mavros

Nocea

Alemao

Alexander

Yin

Glycaemic control among patients with type 2 diabetes mellitus in seven European countries: findings from the Real-Life Effectiveness and Care Patterns of Diabetes Management (RECAP-DM) study. Diabetes Obes Metab. 2008;10(suppl 1):8-15.

19.

Blak

Smith

Hards

Maguire

Gimeno

A retrospective database study of insulin initiation in patients with type 2 diabetes in UK primary care. Diabet Med. 2012; 29:e191-e198.

20.

Laxy

Knoll

Schunk

Meisinger

Huth

Holle

Quality of diabetes care in Germany improved from 2000 to 2007 to 2014, but improvements diminished since 2007. Evidence from the population-based KORA studies. PLOS ONE. 2016;11:e0164704. doi:10.1371/journal.pone.0164704.

21.

Hosomura

Malmasi

Timerman

, et al. Decline of insulin therapy and delays in insulin initiation in people with uncontrolled diabetes mellitus. Diabet Med. 2017;34:1599-1602. doi:10.1111/dme.13454.

22.

Montvida

Shaw

Atherton

Stringer

Paul

SK.

Long-term trends in antidiabetes drug usage in the U.S.: Real-world evidence in patients newly diagnosed with type 2 diabetes. Diabetes Care. 2018;41:69-78. doi:10.2337/dc17-1414.

23.

Bolli

Riddle

Bergenstal

, et al.; on behalf of the EDITION 3 study investigators. New insulin glargine 300 U/ml compared with glargine 100 U/ml in insulin-naïve people with type 2 diabetes on oral glucose-lowering drugs: a randomized controlled trial (EDITION 3). Diabetes Obes Metab. 2015;17:386-394. doi:10.1111/dom.12438.

24.

Ahlqvist

Storm

Käräjämäki

, et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol. 2018;6:361-369. doi:10.1016/S2213-8587(18)30051-2.