Basal insulin initiation use and experience among people with type 2 diabetes mellitus with different patterns of persistence: results from a multi-national survey

Abstract

Background and objective: People with type 2 diabetes mellitus (T2DM) often interrupt basal insulin treatment soon after initiation. This study aimed to describe the experiences during and after basal insulin initiation among people with T2DM with different persistence patterns.

Methods: Adults with T2DM from France, Germany, Spain, UK, US, Brazil, and Japan were identified from consumer panels for an online survey. Respondents who initiated basal insulin 3–24 months prior to survey date were categorized as continuers (no gaps of ≥7 days in insulin treatment); interrupters (first gap ≥7 days within 6 months of initiation and restarted insulin); and discontinuers (stopped insulin for ≥7 days within 6 months of initiation without restarting).

Results: Among 942 participants, continuers were older than interrupters and discontinuers (46, 37, and 38 years, respectively, p < .01). Continuers reported having fewer concerns before and after insulin initiation than interrupters and discontinuers, while interrupters had the most concerns. Continuers also reported fewer challenges during the first week of insulin use. Continuers were more likely to respond that insulin use had a positive impact on specific aspects of life than interrupters and discontinuers, for example on glycemic control (73.0%, 63.0%, and 61.8%, respectively; p < .01 vs. continuers).

Conclusion: Among people with T2DM with different persistence patterns after basal insulin initiation there were significant differences in patient characteristics and experience during and after insulin initiation. Interrupters and discontinuers more frequently reported having concerns and challenges during the initiation process, negative impacts after initiation, and less improvement in glycemic control than continuers.

Keywords:

• Type 2 diabetes
• long-acting insulin
• basal insulin
• initiation

Introduction

More than 285 million people are currently suffering from diabetes1, a chronic disease resulting from insulin insufficiency and/or insulin resistance2. Due to projected increases in worldwide obesity trends, by 2030, it is estimated that this number will increase to 438 million1. Currently, diabetes is a relatively common ailment in developed countries and an emerging one in the developing world: in North America and the Caribbean region, 11.4% of the population is affected by diabetes, and the prevalence of diabetes ranges from 7.9% to 8.3% in most other parts of the world3. The most common form of diabetes (85% to 95% of all cases) is type 2 diabetes mellitus2.

Diabetes can progress to increased complications both microvascular and macrovascular (e.g. heart attacks, strokes, retinopathy, nephropathy, nerve damage, infections, and/or amputations) as well as increased mortality, especially when patients are not achieving glycemic targets4^,5. Treatment, however, can restore blood sugar control6. Early intervention can prevent future diabetes-associated complications7 and even reverse early disease8^,9. Current treatment guidelines recommend a stepwise approach to the management of glycemic control. This includes the use of metformin in most initially diagnosed cases, insulin for symptomatic diabetes and/or patients with elevated blood glucose levels, and sulfonylurea, dipeptidyl peptidase 4 (DPP-4) inhibitors, glucagon-like peptide-1 receptor [GLP-1R] agonists, or basal insulin in patients who fail to achieve/maintain their glycemic target with non-insulin monotherapy10. Given the multitude of options for patients who initially fail to achieve or maintain their glycemic target, treatment choice is made based on patient preferences and disease, drug and patient characteristics10.

Insulin has demonstrably the highest efficacy among all antihyperglycemic medication options in lowering blood glucose levels10^,11 which have been shown to prevent or delay the development of micro- and macro-vascular complications10^,12. Yet, despite strong guideline recommendations, nearly one in three patients are reluctant to start insulin13^,14 and a large proportion of patients (62–82%) interrupt or discontinue treatment shortly after initiation15–17. Even patients who continue insulin treatment often do so on an intermittent basis, frequently forgetting to take the medication at the appointed times18. However, nonadherence19–22 or non-persistence15^,16^,23 with insulin may be associated with poor glycemic control, increased frequency of emergency department visits and hospitalizations, and often higher medical costs (relative to patients who persist with treatment and adhere to treatment guidelines). It is, therefore, essential for clinicians who treat T2DM patients with insulin to better understand the patient experience with this treatment and what differentiates those who continue treatment from those who interrupt or discontinue treatment – so that adequate measures can be taken at the start of prescribed therapy to better support patients throughout their treatment and to help them achieve the targeted outcomes.

Basal insulin options, currently including both intermediate-acting and long-acting formulations of insulin, are simpler and more convenient than mealtime or pre-mixed insulin24^,25, ideally improving treatment persistence and/or adherence. However, while several studies have examined persistence with basal insulin treatment15–18^,26–30, few studies have explored the patients’ reasons for continuing treatment29 or distinguished among patients who may temporarily interrupt therapy and those who discontinue17^,29. Studies also have not described the patient experience during or after insulin initiation for patients with different persistence patterns. Therefore, the objective of this multi-national study was to assess real-world experience and resource use during and immediately after basal insulin initiation among people with T2DM and to separately examine the experiences of patients with different persistent patterns in the developed world. Specific topics included experience immediately before starting basal insulin, such as respondent’s motivation for starting insulin, degree to which respondent felt their views were considered, feelings when considering insulin, concerns before starting insulin, degree of confidence before first self-injection, training and support resources before insulin initiation; and experience after basal insulin initiation such as concerns after first week of insulin use, challenges during first week of insulin use, adverse events experienced while using insulin, and impact of insulin on specific aspects of life, and training and support after insulin initiation.

Methods

Data source

An online survey was administered to participants from the US, the UK, Germany, Spain, France, Brazil and Japan between July and September of 2015. Participants for this study were identified from the Harris Panel – and included members of its chronic illness sub-panel – and from other third-party panels (Branded, SSI, and Toluna). No patient-identifying information was collected, and exemption from review was granted for this study by the Western Institutional Review Board. Participants were compensated for their time according to market research panels’ compensation policies.

Patient sample

Patients were considered eligible to participate in the study if they fulfilled the following selection criteria: (1) were insulin-naïve adults (aged ≥18 years) with T2DM; and (2) had initiated a basal insulin analog (i.e. insulin glargine, insulin detemir, or insulin degludec) 3–24 months before survey administration. Patients were excluded from the study if: (1) they were pregnant or breastfeeding at or after insulin initiation; or (2) the first time they had interrupted for at least 7 days or stopped using basal insulin was after the first 6 months following insulin initiation.

Eligible patients were grouped into three cohorts based on their patterns of treatment persistence: continuers, interrupters, and discontinuers. Continuers had no gap of at least 7 days in their intake of basal insulin (i.e. if patients did have a gap in treatment, but it was of 6 days or less, they were categorized as continuers). Switching of basal insulin analogs was considered a continuation of treatment (unless it involved a gap of at least 7 days between the discontinuation of the first analog and the initiation of the second). Interrupters did not take a basal insulin for at least 7 days within the first 6 months after basal insulin analog initiation and subsequently restarted at least once on any basal insulin within the 24 month period before the survey administration date. Discontinuers stopped using a basal insulin for at least 7 days within the first 6 months of its initiation and subsequently did not restart on any type of basal insulin before the survey administration date. The study focused on patients with early interruption and discontinuation (i.e. within 6 months from initiation) because studies have shown that the majority of patients who interrupt or discontinue therapy do so soon after initiation15^,16 and because patient experiences may differ for patients who stop therapy early versus late in the treatment process, in particular with respect to the patient experience during insulin initiation.

Patient enrollment was ceased soon after either the target quota of 50 respondents per persistence category for each country was reached or the enrollment plateaued.

Data collected

The survey questionnaire was developed based on previous findings from semi-structured qualitative interviews of T2DM patients31^,32. The previous interviews used an open-ended interview guide with questions on the same topics and were administered to T2DM patients recruited from market research panels from the same countries31^,32. Topics in the current questionnaire included: (1) patient demographic and socioeconomic characteristics; (2) disease and treatment history; (3) basal insulin initiation experience (the decision-making process for starting insulin, concerns before initiating insulin, available resources – such as training, support/assistance); (4) patient experience while taking basal insulin (challenges during use, available resources, the impact of insulin use on different aspects of disease control and life; and (5) self-reported reasons for different patterns of persistence. Data on self-reported reasons for different patterns of persistence will be disclosed in a separate manuscript focusing on factors correlated with persistence. The study was determined exempt from review by the Western Institutional Review Board, Puyallup, WA, USA.

Measures and outcomes

Patient baseline characteristics and experience during and after insulin initiation were evaluated separately for continuers, interrupters, and discontinuers, but patients from all counties were pooled together within those treatment groups. In general, results are described by group category if there are statistically significant differences between groups and described for the overall population where there are no differences between groups.

Patient baseline characteristics included demographics and disease characteristics. The following demographic data were collected for all survey respondents: country, age, gender, education level, employment status, and whether living with a spouse/partner. The following disease characteristics were collected for all survey respondents: disease duration (the number of years since the first T2DM diagnosis), type of basal insulin initiated before survey administration, the number of months since the initiation of basal insulin, the mode of delivery of basal insulin initiated, and whether the patient had any prior use of other injected or oral antihyperglycemic medications.

Insulin initiation experience questions included questions regarding the patient’s decision to start insulin: motivations for starting insulin, sources of recommendations to start insulin, provider’s reasons for recommendation to start insulin, degree to which the respondent felt their views were considered, patient’s feelings when considering insulin, patient’s concerns before starting insulin, patient’s degree of confidence before first self-injection, and patient challenges during first week of insulin use. The patients were also asked regarding the availability of the following resources during insulin treatment initiation: initial training and information, support and assistance at initiation, preferred formats for training and information, and types of medical personnel who provided the initial training.

Insulin use experience questions collected data regarding the impact of insulin use on the patient’s life: impact on specific aspects of life, adverse events experienced while taking insulin, and concerns after one week of using insulin, as well as the availability of training and support resources after insulin initiation.

Statistical analyses

Study results were pooled across all countries and responses were weighted to give equal representation to each of the seven countries within each persistence category. Results were compared between continuers, interrupters, and discontinuers: means and standard deviations were used to summarize continuous measures and proportions were reported for categorical variables. Pairwise comparisons between persistence groups were made using t-tests for continuous variables and chi-squared tests for categorical variables. Because there were no corrections for multiple comparisons, a two-sided p-value lower than .01 was used to determine statistical significance. Analyses were performed using SAS v.9.3 (Cary, NC, USA) software.

Results

Patient characteristics

A total of 942 patients participated in the survey; this included 357 continuers, 330 interrupters, and 255 discontinuers (Table 1). Among participants, 154 were from the US, 131 from the UK, 137 from France, 131 from Germany, 150 from Spain, 156 from Brazil, and 83 from Japan.

Table 1. Demographic and diabetes-related characteristics.

	Overall^a	By persistence category^a^,^b
	(n = 942)	Continuers	Interrupters	Discontinuers	p-value^c
		(n = 357)	(n = 330)	(n = 255)	[A vs. B]	[A vs. C]	[B vs. C]
		[A]	[B]	[C]
Country, unweighted N (%)						*	*
US	154 (16.3)	52 (14.6)	52 (15.8)	50 (19.6)
UK	131 (13.9)	50 (14.0)	50 (15.2)	31 (12.2)
France	137 (14.5)	50 (14.0)	50 (15.2)	37 (14.5)
Germany	131 (13.9)	55 (15.4)	50 (15.2)	26 (10.2)
Spain	150 (15.9)	50 (14.0)	50 (15.2)	50 (19.6)
Brazil	156 (16.6)	50 (14.0)	51 (15.5)	55 (21.6)
Japan	83 (8.8)	50 (14.0)	27 (8.2)	6 (2.4)
Age, mean (SD)	40.6 (14.1)	46.1 (15.5)	36.6 (11.2)	37.9 (13.0)	*	*
Male, %	66.6	62.2	78.0	57.8	*		*
College degree, %	52.0	50.2	53.3	52.8
Employment status, %					*	*
Employed	78.7	69.9	86.0	81.9
Not employed	12.9	13.4	11.0	14.6
Retired	8.4	16.7	3.1	3.5
Living with a spouse/partner, %	74.7	76.7	73.6	73.4
Years since first T2DM diagnosis, mean (SD)	6.8 (7.0)	6.7 (6.3)	7.5 (8.0)	6.1 (6.6)
Months since initiation of basal insulin, %					*	*
3 to 6	29.1	25.1	28.0	36.3
7 to 12	37.7	31.1	42.1	41.1
13 to 24	33.2	43.8	29.9	22.6
Mode of delivery of basal insulin initiated, %
Pen (prefilled/disposable)	59.4	63.2	58.8	54.7
Pen (reusable) & cartridge	29.2	26.4	30.3	31.7
Vial and syringe	11.4	10.3	10.9	13.6
Prior use of antihyperglycemic medications to treat T2DM, %
Any prior use of antihyperglycemic medication	65.5	58.1	78.2	59.6	*		*
Oral antihyperglycemics	57.7	54.7	67.5	49.4	*		*
Injectables other than insulin	24.1	12.6	34.8	26.3	*	*

^a Overall results and results by persistence category were weighted to give equal representation to each of the seven countries in the data.

^b Continuers had no gaps of ≥7 days in basal insulin treatment. Interrupters interrupted basal insulin for ≥7 days within the first 6 months after initiation and since restarted basal insulin. Discontinuers stopped using basal insulin for ≥7 days within the first 6 months after initiation and had not restarted basal insulin by the time of the survey.

^c p-values were calculated using t-tests for continuous variables and chi-square tests for categorical variables without adjustment for multiple comparisons. p < .01 was considered statistically significant and denoted with an asterisk (*).

There were significant differences between persistence categories in terms of patient age, gender and employment status (Table 1). Continuers were, on average, significantly older than interrupters and discontinuers (46.1, 36.6, and 37.9 years, respectively). A lower proportion of continuers were employed and a higher proportion of continuers were retired compared with interrupters and discontinuers. More men were found in the interrupter cohort than in the continuers and discontinuers cohorts (78.0% vs. 62.2% and 57.8%, respectively). There were no significant differences across categories regarding educational attainment.

Disease duration was similar across all categories. Overall, for 34.5% of patients basal insulin was the first antihyperglycemic treatment. Prior antihyperglycemic medication (and in particular orally administered medication) use was more frequent in the interrupter group relative to all other groups, while a lower proportion of continuers had prior non-insulin injectable antihyperglycemic medications compared with interrupters and discontinuers (12.6%, 34.8%, and 26.3%). Significantly more continuers had initiated basal insulin treatment 13–24 months ago than interrupters and discontinuers (43.8%, 29.9%, and 22.6% respectively); conversely, significantly more discontinuers had initiated basal insulin treatment 3–6 months ago than continuers and interrupters (36.3%, 25.1%, and 28.0%, respectively).

Insulin initiation experience: motivations and concerns

Across all groups, the most common motivations for starting basal insulin were encouragement from a healthcare provider (HCP, 66.3%), the patient’s expectation of improved glycemic control (48.8%), and concerns about developing other complications of diabetes (32.4%) (Table 2). Most respondents were recommended insulin therapy by their primary care physician (67%) or endocrinologist (18.1%) and improved glycemic control was most common reason for a provider’s recommending insulin in all groups, though more discontinuers had providers recommending insulin after intolerance to other antihyperglycemic medications. Significantly higher proportions of continuers and interrupters than discontinuers reported that their views were somewhat or very/fully considered in the decision to start insulin (73.7% and 66.8% vs. 62.6%, respectively).

Table 2. Insulin initiation experience.

	Overall^a	By persistence category^a^,^b
	(n = 942)	Continuers	Interrupters	Discontinuers	p-value^c
		(n = 357)	(n = 330)	(n = 255)	[A vs. B]	[A vs. C]	[B vs. C]
		[A]	[B]	[C]
Respondent’s motivations for starting insulin, %
Encouragement from physician/healthcare provider	66.3	68.7	64.9	64.9
Improved glycemic control	48.8	44.6	54.3	47.6
Concern about developing other complications of diabetes	32.4	33.8	34.8	27.5
Preference for injections over pills	9.1	6.1	11.5	10.0
Inability to tolerate other antihyperglycemic medications	3.4	3.6	3.9	2.6
Sources of recommendations to start insulin, %
Primary care physician	67.0	67.1	70.2	62.8
Other individual in physician’s office	5.4	3.7	6.1	6.8
Endocrinologist	18.1	19.5	16.1	19.0
Diabetes educator	7.7	8.4	6.2	8.5
Respondent	0.8	0.3	1.1	1.1
Other	1.0	1.1	0.3	1.8
Provider’s reasons for recommendation to start insulin, %
Improved glycemic control	73.4	76.6	71.8	71.1
Prevention of other complications of diabetes	51.2	50.8	53.7	48.5
Inability to tolerate other antihyperglycemic medications	8.1	6.9	6.4	11.8
Not applicable^d	0.8	1.1	0.8	0.3
Degree to which respondent felt views were considered, %						*	*
Not at all/slightly	31.7	26.3	33.2	37.4
Somewhat	27.7	27.8	21.9	35.0
Very/fully	40.6	45.9	45.0	27.6
Feelings when considering insulin, % agree/strongly agree
Sense of failure	34.8	32.0	41.7	29.7	*		*
Reassurance that insulin would help manage diabetes	75.6	79.9	78.2	66.5		*	*
Belief that insulin was not necessary	33.1	27.6	40.3	31.6	*
Feeling that insulin indicated that diabetes was worsening	61.3	62.4	67.2	51.9		*	*
Fear of developing other complications of diabetes	66.7	69.5	69.8	58.6		*	*
Degree of confidence before first self-injection, %						*
Not at all confident	17.7	15.5	15.7	23.2
Somewhat confident	40.4	36.7	41.5	44.1
Confident/very confident	41.9	47.7	42.8	32.6
Concerns before starting insulin, % agree/strongly agree
Worry that insulin would not provide glycemic control	43.4	34.9	51.4	45.0	*
Fear of self-injection	55.5	47.5	61.7	58.7	*	*
Fear of making mistakes during self-injection	57.5	48.5	64.5	61.1	*	*
Worry that scarring or bruising would result from injections	48.4	40.7	58.3	46.2	*		*
Worry about proper insulin storage	51.8	44.7	55.6	56.9	*	*
Concern about carrying insulin around	52.8	47.7	59.7	51.1	*
Worry that regular insulin use would interfere with daily activities	54.3	43.9	63.4	57.1	*	*
Concern that he/she would need to visit physician/nurse more often	50.0	41.5	59.7	49.2	*
Concern that he/she would need to monitor blood glucose more often	55.6	49.4	63.9	53.6	*
Concern about potentially frequent hypoglycemia	54.7	47.7	60.9	56.3	*
Concern about becoming insulin dependent	60.7	58.0	62.3	62.5
Concern about ability to pay for insulin therapy	47.9	38.3	55.3	51.8	*	*
Worry about potential weight gain	51.7	47.4	59.2	47.8	*		*
Worry about injecting insulin in front of other people	50.0	41.4	60.5	48.5	*		*
Number of concern before starting insulin, mean (SD)	7.3 (4.6)	6.3 (4.5)	8.4 (4.7)	7.5 (4.3)	*	*
Number of concerns after one week of using insulin, mean (SD)	6.1 (4.7)	5.0 (4.4)	6.8 (4.9)	6.6 (4.4)	*	*

^a Overall results and results by persistence category were weighted to give equal representation to each of the seven countries in the data.

^b Continuers had no gaps of ≥7 days in basal insulin treatment. Interrupters interrupted basal insulin for ≥7 days within the first 6 months after initiation and since restarted basal insulin. Discontinuers stopped using basal insulin for ≥7 days within the first 6 months after initiation and had not restarted basal insulin by the time of the survey.

^c p-values were calculated using t-tests for continuous variables and chi-square tests for categorical variables without adjustment for multiple comparisons. p < .01 was considered statistically significant and denoted with an asterisk (*).

^d Some patients did not receive a recommendation to start insulin from their providers. For these patients, there was not a reason the provider recommended they start insulin.

Patient attitudes relative to the need for treatment when insulin was first discussed differed among cohorts: higher proportions of interrupters than all others reported a sense of failure and a belief that insulin was not necessary, while lower proportions of discontinuers than all others felt reassured that insulin would manage diabetes, that their diabetes was worsening, or feared developing other complications (Table 2).

Among all respondents, the most common concerns before starting insulin were fear of dependence on insulin (60.7%), fear of making mistakes during self-injection (57.5%), more frequent blood glucose monitoring (55.6%), and fear of self-injection (55.5%) (Table 2). Overall, continuers had fewer concerns regarding insulin initiation than the other two groups (means of 6.3 among continuers, 8.4 among interrupters, and 7.5 among discontinuers; p < .01 for comparisons vs. continuers). In particular, continuers were less likely to express a strong concern about the ability of insulin to provide glycemic control, or to worry about self-injections, treatment inconvenience, side effects, or cost than interrupters and discontinuers. For example, only 48.5% of continuers agreed or strongly agreed that they feared making mistakes during self-injection compared with 64.5% of interrupters and 61.1% of discontinuers. Discontinuers reported more often that they felt not at all confident before the first injection of insulin and continuers felt the most confident (only the comparison of continuers vs. discontinuers was significant at p < .01). After one week of using insulin, patient concerns were similar to those before initiation, but numerically lower proportions of participants agreed or strongly agreed that they had specific concerns. Similar to before initiation, continuers had on average the lowest number of concerns among all groups after one week of using insulin (mean of 5.0 among continuers, 6.8 among interrupters, and 6.6 among discontinuers; p < .01 for comparisons vs. continuers). In addition, similar to the trends noted among persistence groups before insulin initiation, lower proportions of continuers had specific concerns related to the ability of insulin to provide glycemic control, or to worry about self-injections, inconvenience, side effects (not significant at p < .01 for continuers vs. discontinuers), or the cost of treatment (not significant at p < .01 for continuers vs. discontinuers) than interrupters and discontinuers (data not shown).

During the first week of insulin use, the most common challenges encountered by patients were injecting insulin (31.4%), titration (30.6%), dealing with emotions about needing insulin (29.4%), remembering to inject insulin regularly (28.7%), and needing more frequent blood glucose monitoring (27.5%) (Table 3). Fewer continuers reported difficulty in dealing with challenges during the first week of insulin initiation than interrupters and discontinuers: for example, 19.6% of continuers reported experiencing challenges with more frequent blood glucose monitoring compared with 33.2% of interrupters and 31.3% of discontinuers. Continuers also reported significantly fewer challenges than interrupters and discontinuers (mean of 1.9 among continuers, 2.7 among interrupters, and 2.9 among discontinuers).

Table 3. Insulin use experience.

	Overall^a	By persistence category^a^,^b
	(n = 942)	Continuers	Interrupters	Discontinuers	p-value^c
		(n = 357)	(n = 330)	(n = 255)	[A vs. B]	[A vs. C]	[B vs. C]
		[A]	[B]	[C]
Challenges during first week of insulin use, % difficult/very difficult
Injecting insulin	31.4	25.0	32.0	39.5		*
More frequent blood glucose monitoring	27.5	19.6	33.2	31.3	*	*
Titration	30.6	26.0	29.2	38.9		*
Proper insulin storage	21.9	15.0	25.7	26.4	*	*
Remembering to inject insulin regularly	28.7	21.2	34.5	31.8	*	*
Making time during the day to inject insulin	25.0	17.3	29.7	29.8	*	*
Feeling confident about treating hypoglycemia	25.6	18.7	27.2	33.0	*	*
Dealing with emotions about needing insulin	29.4	23.5	31.5	34.8		*
Worry about reaction to insulin use from friends/family	23.8	19.3	27.8	25.2	*
Number of challenges, mean (SD)	2.4 (2.8)	1.9 (2.7)	2.7 (3.0)	2.9 (2.7)	*	*
Adverse events experienced while taking insulin, %
Uncontrolled high blood glucose	22.3	15.4	29.4	23.0	*
Symptoms of hypoglycemia	36.2	31.9	41.8	35.1	*
Weight gain	39.4	34.8	44.8	38.8	*
Injection site reaction	15.8	16.6	15.7	14.9

^a Overall results and results by persistence category were weighted to give equal representation to each of the seven countries in the data.

^b Continuers had no gaps of ≥7 days in basal insulin treatment. Interrupters interrupted basal insulin for ≥7 days within the first 6 months after initiation and since restarted basal insulin. Discontinuers stopped using basal insulin for ≥7 days within the first 6 months after initiation and had not restarted basal insulin by the time of the survey.

^c p-values were calculated using t-tests for continuous variables and chi-square tests for categorical variables without adjustment for multiple comparisons. p < .01 was considered statistically significant and denoted with an asterisk (*).

Insulin initiation experience: resources, training and support

Before insulin initiation, the majority of respondents across all three persistence groups had received training on self-injection (97.8%), titration (97.3%), insulin therapy in general (97.5%), diet/exercise (98.3%), and diabetes in general (97.6%). Most participants either found treatment-related training helpful (80.7% to 87.7%) or would have found it helpful if it had been available (59% to 74%). The distributions of training availability and use and helpfulness of different types of training were generally similar across the persistence groups.

Similar to patient-reported experiences with training at insulin initiation, after insulin initiation the majority of patients had available training about diet/exercise, diabetes management, and symptoms of hypoglycemia. The distributions of training availability and use and helpfulness of different types of training were generally similar across the persistence groups.

Across the entire sample, training was most often provided by a physician (68.9%), a nurse/physician’s assistant (44.5%), and a diabetes educator (32.7%). Significantly lower proportions of continuers received training by a physician or a pharmacist compared with interrupters and discontinuers (training by physician: 62.3% among continuers, 73.3% among interrupters, and 72.5% among discontinuers; training by pharmacist: 14.9%, 23.1%, and 28.8%).

Most participants (65.2%) preferred to receive in-person training, 46.3% preferred written materials, 42.1% preferred a website, and 37.7% preferred videos. There was no clear pattern of preference for a specific training material format among the three persistence groups, though continuers exhibited significantly less interest in written materials than interrupters, and interrupters had a stronger preference for telephone hotlines than continuers and discontinuers (all p < .01).

Most participants had access to support services during insulin treatment initiation, such as patient support groups (88.4% of continuers, 96.4% of interrupters, and 92.8% of discontinuers), support from medical personnel (97.7%, 99.7% and 97.8%, respectively), friends/family support (94.5%, 98.9%, 97.7%, respectively), and therapist availability (92.2%, 97.5%, 92.7%, respectively). More than 40% of patients who did not have access to support services reported that support services would have been helpful. Higher proportions of continuers reported not using different types of support compared with interrupters and discontinuers: 27.2% of continuers did not use patient support groups (vs. 14.8% for interrupters and 19.0% for discontinuers; p < .01 vs. continuers), 8.4% of continuers did not use medical personnel support (vs. 5.6% and 4.2%, respectively), 14.1% of continuers did not use social (friend/family) support (vs. 4.7% and 10.5%, respectively; p < .01 for comparison between interrupters vs. continuers), and 22.5% of continuers did not require therapist sessions (vs. 13.7% and 19.7%, respectively; p < .01 for comparison between interrupters vs. continuers). Use of support services after initiation was similar to that before/during initiation, with continuers being most likely to not use any services, and interrupters being most likely to use most available services.

Insulin use experience: positive and negative factors of insulin use and side effects

There were significant differences across persistence groups in terms of the impact of insulin treatment on various aspects of life and disease control. Continuers were more likely than interrupters and discontinuers to respond that insulin use had a positive impact on specific aspects of life such as glycemic control (73.0% vs. 63.0% and 61.8%, respectively), physical well-being (67.5% vs. 64.7% and 58.7%, respectively), and emotional well-being (60.5% vs. 57.8% and 46.1%, respectively) (Figure 1). Continuers were also less likely to report a negative impact of insulin use on these and other specific aspects of life.

Figure 1. Impact of insulin use on specific aspects of patients’ lives. p-values based on whether insulin had a positive, neutral or negative impact on specific aspects of life were calculated using chi-square tests. The proportions of patients who reported a neutral/no impact for specific aspects of life are not shown. p < .01 for continuers compared to interrupters and compared to discontinuers are marked with an asterisk (*) and a cross (†) respectively. p-values less than .01 for interrupters compared to discontinuers are marked with a double cross (‡).

Side effects related to insulin use were reported as follows: overall, 39.4% of participants reported experiencing weight gain while on basal insulin therapy, 36.2% experienced hypoglycemia, 22.3% experienced uncontrolled high blood glucose, and 15.8% experienced injection site reactions (Table 3). Continuers were significantly less likely to have experienced uncontrolled high blood glucose, hypoglycemia, or weight gain after starting insulin compared with interrupters. There were no significant differences in the rate of other adverse events for continuers vs. discontinuers.

Discussion

According to the position statement of the American Diabetes Association and European Association for the Study of Diabetes, insulin has “nearly universal response and theoretically unlimited efficacy”11. However, many patients interrupt or discontinue their insulin treatment soon after initiation15–18^,26–30^,33. While the barriers to starting insulin and factors associated with persistence have been investigated before13^,14^,29^,34, the experience during and after basal insulin initiation of patients with different persistence patterns is unknown. The current study uses data on patient experiences during insulin initiation and treatment from an online survey to understand the differences among different patterns of persistence.

In this study, different persistence cohorts exhibited differences in terms of age and gender: continuers were older than interrupters and discontinuers and interrupters were more likely to be male than other persistence cohorts. A significant association between older age and increased likelihood of persistence has been noted in several prior studies evaluating the factors associated with persistence after basal or pre-mixed insulin initiation among T2DM patients15^,16^,18^,27^,35. In the current study, 65.5% of patients had any prior use of oral or injectable antihyperglycemic medications (57.7% had used oral antihyperglycemics; 24.1% had used a non-insulin injectable). Compared to previous studies15–17, prior oral antihyperglycemic use is lower here than in other studies and prior use of non-insulin injectable treatments is higher than in other studies. Continuers in the current study had lower rates of prior antihyperglycemic medication use (both oral and injectable) than interrupters and had similar rates to discontinuers. This is contrary to findings from prior studies on the role of prior exposure to antihyperglycemic medication and persistence, which generally concluded that prior use of multiple classes of antihyperglycemic medications is associated with higher likelihood of persistence15^,16^,36. This may be due to several possibilities. The need for insulin may play a role in the patient’s attitude toward treatment. For example, the likelihood of continuing insulin is different for patients who are prescribed insulin as a necessary regimen versus those who are presented insulin treatment as an option among other treatments for glucose control. The first type of patients tend to be older and/or have gone through several stepwise regimens (diet/exercise, then oral medications, then non-insulin injectable therapies) compared to the second category of patients. It is possible that patients categorized as continuers were taking medication because it was prescribed as necessary for a group of patients that was older and/or had a more advanced disease state. Conversely, given the younger age of discontinuers/interrupters, it is also possible that at least a subset of these patients were given insulin as an optional treatment choice rather than a necessity. Known benefits of treatment alternatives to basal insulin such as GLP-1R (for example weight loss and low rate of hypoglycemia) may have influenced the patient decision for insulin treatment.

The patient’s involvement in the decision to start insulin, attitude, and anticipation of potential treatment challenges is of great importance in initiating and continuing with a course of treatment as prescribed13^,14^,18^,29. Significantly higher proportions of continuers and interrupters than discontinuers reported that their views were somewhat or very/fully considered in the decision to start insulin. The concerns before and after insulin initiation identified in this study are consistent with common concerns reported in previous studies investigating patient attitudes regarding insulin therapy in insulin-naïve diabetes respondents13^,14. Those concerns typically centered on the risk of hypoglycemia associated with insulin therapy, dependence on insulin therapy, and less flexibility in their lifestyle13^,14. In the current study, continuers were less likely to have concerns before and after initiating insulin (e.g. permanence of insulin therapy, fear of self-injection or making mistakes during self-injection, concerns about inconvenience related to more frequent monitoring and insulin storage, and concerns about hypoglycemia and weight gain side effects) and were less likely to experience specific challenges following insulin initiation than discontinuers or interrupters.

In this study continuers were less likely than interrupters to report side effects with treatment (such as uncontrolled high blood glucose, hypoglycemia, and weight gain). Higher proportions of continuers reported positive impact of insulin use on glycemic control compared with interrupters and discontinuers which suggests the importance of patients actually seeing the benefits of therapy so that they continue taking it. Given equal availability of training and other resources to the other persistence groups, continuers were less likely to be somewhat independent both before and after insulin initiation (e.g. support from medical personnel, patient support groups, and support from friends/family) with respect to insulin treatment. A potential explanation is that continuers had less need to seek out support. This is consistent with the finding that lower proportions of continuers had concerns before or after insulin initiation or experienced specific challenges during the first week of insulin use. This limited use of resources and reported challenges by continuers may reflect greater confidence in treatment efficacy, but may also be due to greater personal comfort with using a new medication.

Discontinuers were more likely than all other groups to have had insulin treatment recommended due to an inability to tolerate other antihyperglycemic medications and to report that they did not feel that their views were considered in the decision to start insulin. Finally, despite having the highest use of support resources, interrupters also reported having the highest number of concerns and challenges with insulin treatment (before and after initiation) of all persistence groups. This finding may indicate that existing resources for insulin treatment support could be improved in order to address the higher emotional/conceptual burden of discontinuers and interrupters.

Nonetheless, given a relatively high number of concerns before and after insulin initiation and challenges during the first week of insulin use across all treatment groups, the results of this study indicate that there is a clear opportunity to improve patients’ expectations and experiences with insulin treatment overall, regardless of persistence behaviors.

Strengths and limitations

The current study’s main strength is its large sample size and multi-national design with a diverse representation of countries. In addition, the survey collected data across the global spectrum of patients’ experiences and concerns during the insulin initiation and maintenance processes.

However, the study is also subject to several limitations. First, the findings of this study may not be generalizable to all insulin-treated patients, as sample bias may be present37 – participants were recruited from online panels consisting of volunteers who have agreed to participate in surveys. This type of sample bias may be especially problematic in countries with limited internet use, where respondents may be younger, more likely to be in urban areas, and have different socio-economic characteristics from the overall population of interest. In the current study, the average patient age was 41 years. Furthermore, respondents had been diagnosed an average of 7 years before initiating insulin and 35% of them used insulin as the first line therapy. By contrast, according to the Centers for Disease Control, the average age at first diabetes diagnosis for adults 18–79 years in the US is 54 years38 (an older population than in the current study) and patients in other persistence studies were more likely, overall, to have had other therapies before starting insulin treatment (though this differs widely between various persistence groups)15–17. Other limitations of using a web-based data collection methodology may also apply39^,40.

Second, because of quotas for different persistence groups, the results for the overall sample are not representative of findings for the overall population of people with T2DM initiating basal insulin. Actual continuation, interruption, and discontinuation rates are expected to be closer to those estimated from administrative claims analyses, such as approximately 20% of basal insulin initiators with T2DM continuing therapy, 62% interrupting, and 18% discontinuing therapy in the year after initiation in the US16.

Third, approximately 30% of participants this study started insulin treatment within 3–6 months of survey administration. These participants could only be evaluated for 3–6 months, and there is a possibility of misclassification of their persistence pattern. Some of the participants categorized as continuers may stop therapy within 6 months of insulin initiation and may be more similar to those in the interrupter or discontinuer group. Similarly, participants categorized as discontinuers may restart insulin therapy and may be more similar to those in the interrupter group.

Fourth, survey responses in this study are subject to recall bias, as respondents could have initiated insulin 3–24 months prior to the date of study administration.

Finally, while this study describes the experience of people with T2DM with different persistence patterns after basal insulin initiation, the study cannot firmly draw any conclusions regarding an association or causational relationship between reported patient characteristics and specific experiences and the likelihood of persistence on treatment because the descriptive analysis does not adjust for plausible confounders (demographics, disease and treatment characteristics). Findings should be used as a hypothesis-generating tool for further research that controls for possible confounding variables. The topics of self-reported reasons for different patterns of persistence and correlates of insulin persistence are discussed in a separate manuscript.

Conclusion

This study described the real-world experiences of people with T2DM who have different patterns of persistence after initiating basal insulin. This study found that about half of participants reported having specific concerns before initiation and about a third of participants reported experiencing specific challenges during the first week of insulin use. Participants with different patterns of persistence exhibited statistically significant differences in patient characteristics, concerns and perceptions at insulin initiation, and experience after insulin initiation. Continuers reported fewer concerns before and after insulin initiation and fewer challenges during the first week of insulin use than interrupters and discontinuers. Understanding patient experiences can help providers to more effectively manage T2DM care with basal insulin analogs. Given a relatively high number of concerns before and after insulin initiation and challenges during the first week of insulin use across all treatment groups, the results of this study help to inform providers about the importance of addressing patient concerns and indicate that there is a clear opportunity to improve patients’ experiences with insulin treatment, including treatment persistence.

Transparency

Declaration of funding

Funding for this research was provided by Eli Lilly and Company and Boehringer Ingelheim.

Author contributions: M.P.-N., J.I.I., I.H., C.Z., D.C., S.Ka. and M.P. were involved in the conception and design of this study. L.S., S.Ki, and C.Z. were involved in the analysis of the data. All authors were involved in the interpretation of the data and in revising the manuscript critically for intellectual content, provided a final approval of the version to be published, and agreed to be accountable for all aspects of the work.

Declaration of financial/other relationships

M.P.-N., D.C., S.Ka., A.M.D. and I.H. have disclosed that they are employees of Eli Lilly and Company. J.I.I., L.S. and H.G.B. have disclosed that they are employees of Analysis Group Inc., which has received research funding from Eli Lilly and Company. S.Ki. and C.Z. have disclosed that they were employees of Analysis Group Inc. at the time of the study. M.P. has disclosed that he has received consultancy fees from Astra Zeneca, Calibra, Lilly, and Novo Nordisk, lecture fees from Lilly and Novo Nordisk, and has served on the Scientific Advisory Boards of Calibra, Lilly, and Novo Nordisk. The sponsors, Eli Lilly and Company and Boehringer Ingelheim, reviewed and provided comments on this article. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

CMRO peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgements

Medical writing assistance was provided by Ana Bozas PhD, a former employee of Analysis Group Inc.