Introduction
When treating type 2 diabetes (T2D), as with any chronic disease, the achievement of optimal therapeutic outcomes relies on both treatment persistence and treatment adherence. In real-world studies of patients with T2D, persistence with and adherence to antidiabetes therapies are defined and evaluated in many different ways. Treatment persistence is the continuation of therapy over time. It is often measured by the availability of the drug, expressed as the continuous filling of prescriptions (e.g. as the percentage of patients still obtaining the drug at 12 months) or alternatively as “the duration of time from initiation to discontinuation of therapy” [
1]. Treatment adherence is defined as “the extent to which a patient acts in accordance with the prescribed interval and dose of a dosing regimen” [
1]. This is based on the principle of conformity to a rule enacted by others. Understanding the rule is not necessary. In medical practice, this is understood as strict compliance with prescriptions in terms of schedules and dosages. Direct measurement of treatment adherence is difficult and is often expressed indirectly using refill records as either the medication possession ratio (MPR) or the proportion of days covered (PDC). Although the exact formula used to calculate MPR varies according to source, it may be defined as the sum of the days’ supply for all fills of a given drug in a particular time period divided by the number of days in the time period. Patients who routinely refill their medications early will have an inflated MPR, whereas population estimates of MPR may also be falsely elevated unless individual adherence is capped at 100%. PDC, in contrast, is a more conservative measure of refill record-based adherence [
2]. Instead of adding the days’ supplied in a given period, PDC considers only the days that are “covered”. The use of treatment adherence complements that of treatment persistence by including the notion of a drug amount necessary for the functioning of the treatment. Measuring treatment adherence from databases is complex, as compliance with a dosing schedule cannot be assessed; compliance with dosing is generally replaced with an assessment of the availability of a sufficient quantity of the drug to be able to comply with dosing. Adherence is a qualitative tool in its own right.
Pharmacological approaches to achieving glycaemic control in patients with T2D usually involve a sequential programme of treatment intensification from monotherapy with oral medication, dual and triple therapy with a combination of oral therapies or oral plus injectable agents, and combination injectable therapy [
3,
4]. Although persistence with and adherence to T2D medications vary according to the treatment modality, lack of persistence and adherence in general act as a substantial barrier to the achievement of glycaemic control. Persistence and adherence are affected by several factors, including the mode of administration, administration frequency/regimen complexity, and patient expectations. Regarding the mode of administration, persistence and adherence are generally better for oral medications compared with injected medications. However, persistence and adherence may vary even among different oral antidiabetes drugs (OADs) classes and for different insulin regimens, indicating that factors other than administration route are also important [
5‐
8]. In the case of insulin regimen, for example, a higher daily insulin dose and greater regimen complexity may contribute to poorer treatment adherence [
7,
8]. Persistence and adherence are also influenced by the frequency of administration, with fewer daily insulin injections and fixed-ratio combinations that reduce pill counts having beneficial effects on adherence [
5,
9]. Mashitani et al. [
10] showed that patients with T2D who take more injections are more likely to omit insulin injections. Mashitani et al. [
11] also reported that the risk of lower adherence compared with higher adherence increased as the number of injections increased (relative risk [RR] 0.98; 95% confidence interval [CI] 0.65–1.46). Fear of injections and adverse effects such as hypoglycaemia and weight gain have all been reported as reasons for nonadherence to antihyperglycaemic medications [
12‐
14]. They also vary according to patient expectations, such as the effects of medication on weight [
15]. Although the simplification of regimens associated with the use of premixed insulins can have a positive effect on adherence, other factors concerning patient satisfaction may come into play. Additionally, studies comparing premixed and basal insulin regimens have produced mixed results concerning the relative reduction of glycated haemoglobin (HbA1c), but most studies have shown an increased risk of hypoglycaemic events and weight gain with premixed insulin [
16,
17], which may have a negative impact on adherence and persistence.
The aims of this review article are to provide an overview of persistence with and adherence to different antidiabetes therapies in patients with T2D in the real-world setting, to examine factors contributing to poor treatment persistence and adherence, and to assess available data on the impact of poor treatment persistence and/or adherence on clinical and healthcare utilization. This review is based on previously conducted studies and does not contain any studies with human participants or animals performed by any of the authors.
Literature Search
A literature search was conducted in October 2015 using EMBASE and PubMed to identify sources to inform this review article. The literature search was limited to English language articles within the timeframe of November 2013 to October 2015. Although this timeframe excluded older articles such as those relating to adherence to sulfonylureas and, conversely, selected for modern treatment options such as glucagon-like peptide-1 receptor agonists (GLP-1 RA), the intention was to review topical literature relating to persistence and adherence in patients with T2D. However, when data were limited, the publication timeframe criterion was expanded. A reviewer screened the abstracts for inclusion in a full text review and then screened full text references to identify any additional publications with relevant information for inclusion in the targeted literature review, and hence in this review article. The literature search was complemented in February 2018 by internet searches performed by hand, which included literature published between November 2015 and December 2017.
Impact of Treatment Persistence on Clinical and Economic Outcomes
Lack of treatment persistence is associated with a lower likelihood of achieving the target HbA1c concentration and maintaining glycaemic control. In a large US study including > 56,000 veterans with T2D, patients reported as having good treatment persistence with OADs over the 12 months following treatment initiation were more likely than nonpersisting patients to achieve the HbA1c goal of ≤ 7.0% (RR 1.07; 95% CI 1.06–1.09) [
49].
A pooled analysis of data from 3 US-based retrospective studies in patients with T2D who had previously been treated with OADs and who were initiated on basal insulin analogues (insulin glargine or insulin detemir) evaluated the consequences of treatment persistence/nonpersistence over a 1-year follow-up period [
50]. Patients who were persistent with treatment compared with nonpersistent patients had significantly lower average HbA1c levels at 1 year (8.26% vs 8.46%, respectively;
P = 0.0260) and significantly greater average HbA1c reductions (− 1.15% vs − 0.86%, respectively;
P = 0.0078), with no significant difference in hypoglycaemic events. Higher insulin treatment persistence was also associated with lower healthcare utilization [
50]. Given these beneficial outcomes of treatment persistence, it does raise the question as to whether uncontrolled hyperglycaemia is part of a feedback loop involving not only persistence but also clinical inertia. Thus, intensifying treatment to improve glycaemic control might lead to improved persistence.
Treatment persistence in T2D is associated with decreased healthcare utilization and costs. In a US study, acute care (hospitalization and emergency department [ED]) costs were shown to be 9.6% higher among patients with T2D who discontinued insulin therapy (basal or mixture) early compared with patients who persisted with treatment beyond the first 90 days (
P < 0.001) [
51]. A German study reported that persisting on basal insulin glargine beyond the first 90 days, compared with switching insulins, was associated with significantly lower diabetes-related prescription costs (74EUR less) (
P = 0.0780) [
52].
Lack of treatment persistence with antidiabetes therapy is also linked to a higher likelihood of morbidity, mortality, and poorer health status [
53].
Impact of Treatment Adherence on Clinical and Economic Outcomes
Medication adherence has also been shown to be associated with improved glycaemic control in patients with T2D. A retrospective cohort study in the US showed that patients with T2D who were considered adherent to a GLP-1 RA (PDC ≥ 0.80) had a significantly greater reduction in HbA1c than nonadherent patients and were more likely to have a reduction in HbA1c of ≥ 1.0% [
54]. Conversely, poor adherence to antidiabetes therapy is associated with failure to achieve HbA1c targets in the real-world setting. Using the four-item Morisky Medication Adherence Scale (a validated instrument for which higher scores indicate poorer adherence), patients who self-reported that they forgot to take medications at baseline were shown to have an absolute HbA1c increase of 0.43% after 6 months of treatment with insulin and/or OADs [
55]. In another study, a one-point increase in the eight-item Morisky Medication Adherence Scale was found to be associated with a 0.21% increase in HbA1c in patients with T2D using basal insulin analogues [
56].
Poor medication adherence has been identified as a key reason why the effectiveness of antidiabetes therapy in the real-world setting is substantially less than the efficacy reported in randomized controlled clinical trials (RCTs) [
57]. In an analysis of real-world data from patients with T2D initiating treatment with a GLP-1 RA (
n = 221) or a DPP-4 inhibitor (
n = 652), significantly smaller reductions in HbA1c were reported at 1 year in the real-world setting than had been reported in RCTs after the same period for both GLP-1 RAs (− 0.52% vs − 1.30%, respectively;
P < 0.01) and DPP-4 inhibitors (− 0.51% vs − 0.68%, respectively;
P = 0.01) [
57]. The authors attributed approximately three-quarters of these gaps to patient treatment adherence rates, which were 29% with GLP-1 RAs and 37% with DPP-4 inhibitors in the real-world analysis, compared with 95% for both drug classes in the RCTs. The study underscored the urgent need to effectively improve treatment adherence among patients with T2D in the real world. Additionally, it alludes to the impact of the route of administration on treatment adherence.
Hyperglycaemia and long-term complications of T2D are associated with poor adherence to antidiabetes medications and have important clinical consequences for the patient [
58]. A narrative review of studies published in English and Spanish identified several studies linking poor adherence to T2D therapies with increased risks of morbidity and premature mortality [
58].
A large UK study of almost 16,000 patients with T2D who were treated with a combination of OADs and insulin found that after adjustment for confounding factors, medication nonadherence was independently associated with increased all-cause mortality (hazard ratio 1.579; 95% CI 1.167–2.135) [
35].
A systematic review by Banerji et al. [
59] investigating the impact of glycaemic control and medication treatment adherence on clinical and economic outcomes in patients with T2D confirmed the wealth of data linking poor medication treatment adherence to increased medical resource utilization and healthcare costs.
A 2005 analysis of the impact of medication treatment adherence in diabetes on healthcare costs, within a large population-based sample of patients aged < 65 years, estimated that a 20% increase in medication adherence, while increasing the average drug-utilization costs by 177 USD per patient per year, would result in a disease-related medical-cost reduction of 1251 USD per patient per year [
60]. This would result in a net saving of 1074 USD per patient per year [
60].
Real-world studies in the US have shown that total healthcare expenditures for patients with T2D who utilized insulin pens decreased with improvements in adherence. Significant decreases in ED visits (odds ratio [OR] 0.44; 95% CI 0.21–0.92;
P < 0.05) and physician visits (OR 0.39; 95% CI 0.24–0.64;
P < 0.05) have been reported after conversion from vial and syringe to pen administration of insulin therapy. These are associated with total mean all-cause treatment costs reductions of 1590 USD per patient per year [
61]. Additionally, a large study of 23,362 patients with T2D who used an insulin pen found that the average per patient per year healthcare expenditure was 9.4% lower for patients in the most adherent (MPR 0.81–1.00) compared with the least adherent (MPR 0.00–0.20) groups (23,839 USD vs 26,310 USD, respectively;
P = 0.007) [
62].
Other US analyses investigating the economic consequences of treatment nonadherence have shown increased resource utilization and healthcare costs associated with poor adherence. DiBonaventura et al. [
56] found that, for patients with T2D using basal insulin analogues, each one-point increase in treatment nonadherence on the eight-item Morisky Medication Adherence Scale was associated with a 4.6, 20.4, and 20.9% increase in the number of physician visits, ED visits, and hospitalizations, respectively. Encinosa et al. [
63] reported that, in non-elderly patients with T2D, an increase in treatment adherence to OADs from 50% to 100% resulted in a 23.3% reduction in the rate of hospitalization and a 46.2% reduction in ED visits, leading to cost savings of 866 USD per patient and a cost offset of 1.14 USD for every 1.00 USD spent on diabetic drugs.
Other studies have explored the potential impact of treatment adherence on diabetes complications. A retrospective database analysis of new OAD users found that good adherence (defined as MPR ≥ 0.8) was associated with significantly reduced risk of a new microvascular or macrovascular diabetes complication (adjusted hazard ratio 0.96; 95% CI 0.92–1.00;
P = 0.05) [
64]. Initial adherence appears to be important, with another retrospective cohort study observing that during the first 5 years of OAD treatment, those who were initially nonadherent to therapy were more likely to experience myocardial infarction, ischaemic stroke, or death [
65].
This review is limited by the inclusion of studies that the authors regard as being most pertinent to the central review objectives, identified within a relatively short timeframe. It is not a comprehensive review of the field, nor is it a systematic review. One consequent limitation is that no studies have been included concerning the use of long-acting insulin degludec. However, we know of no data suggesting any difference between insulin glargine 300 units/mL and insulin degludec regarding the quality of adherence to insulin therapy or the rate of persistence. Because reimbursement issues are very complex and differ widely according to the country and healthcare system, it has not been discussed here.
Conclusion
For patients with T2D, poor persistence with and adherence to antidiabetes medications can increase the risk of long-term complications, leading to poorer health status and an increase in healthcare resource utilization and costs. A clear unmet need remains in T2D for therapies that improve treatment persistence and adherence compared with currently available treatments, thereby positively impacting clinical and economic outcomes.
Several approaches to improving treatment persistence and adherence have been suggested, including: reducing treatment complexity (e.g. using fixed-dose combination therapy that decreases the frequency of administration [
24,
35], implantable therapies for drug delivery); developing medications with an improved safety profile (e.g. lower risk of weight gain, hypoglycaemia, gastrointestinal side effects [
20,
35]); enhancing educational initiatives; and improving communication (e.g. telemedicine approaches, including websites and electronic records) [
58]. In addition, the impact of treatment persistence and adherence on disease management must be stressed at the time of treatment initiation [
65].
Lack of treatment persistence and treatment nonadherence are highly prevalent among patients with T2D, and there is a global need to address the complexity of antidiabetes therapy and simplify treatment regimens. Fixed-ratio combinations offer a prospective solution to overcome barriers to injectable therapy and treatment intensification, and they would thereby increase treatment persistence and adherence due to less complex regimens requiring fewer daily injections.