The ADA recommends a change in therapy if a patient’s HbA1c target is not achieved after 3 months . Across 22 USA health care organizations, in bolus insulin-naïve patients with type 2 diabetes and HbA1c ≥ 8.0% (≥ 64 mmol/mol), therapeutic inertia (no new class of drug being prescribed or not achieving HbA1c < 8.0%) was observed in approximately 50% of patients after 6 months (ranging from 34% to 54% across organizations), more than one-quarter after 12 months, and > 10% of patients after 24 months (7–19% across organizations). These findings are of concern and broadly suggest that ADA guidelines are not being followed. Delaying the intensification of therapy during the early stages of type 2 diabetes will increase the risk of related complications in these patients, and the associated economic burden [7, 22, 23].
In agreement with previous reports in the literature, less therapeutic inertia was observed in patients receiving one or no OADs, while greater inertia was observed in those patients treated with multiple OADs or basal insulin at baseline. In a retrospective UK study of > 80,000 patients with type 2 diabetes, Khunti and colleagues showed a longer delay to intensify treatment with combination versus monotherapy regimens, specifically a 1.6-year and > 6.9-year delay in patients with HbA1c ≥ 8.0% (≥ 64 mmol/mol) on one or two OADs, respectively . In a similar study, among patients with type 2 diabetes treated with insulin and eligible for intensification (HbA1c ≥ 7.5%), only 30.9% were intensified, with a median time to intensification of 3.7 years . Our study generally corroborates the presence of therapeutic inertia in patients with type 2 diabetes. Therapeutic inertia was particularly prevalent for the first 6 months of suboptimal glycemic control but gradually decreased over the studied 24-month period, although a less prominent decrease was observed in patients using more complex regimens.
Studies have shown that therapeutic inertia was still present during the first 6 months in patients with type 2 diabetes starting basal insulin therapy, even when dose uptitration was considered . Therefore, although the inability to account for dose titration in the data was a limitation of our analysis, it is unlikely to entirely explain the lack of observable actions in patients with persistently uncontrolled type 2 diabetes over the studied 24 months. Those patients receiving basal insulin therapy at baseline are likely to have relatively advanced type 2 diabetes, and the fact that 35% had HbA1c levels that remained out of control with no new class of glucose-lowering medication prescribed after 1 year, and 17% after 2 years, is concerning. This suggests a need for further therapeutic intervention in addition to dose uptitration in this population.
After 24 months, an observable action was seen in approximately 90% of the studied patients; that is, they were prescribed a different treatment or they achieved HbA1c < 8.0% (< 64 mmol/mol). However, only 56% of the 28,000 patients followed ever achieved HbA1c < 8.0% over the 24-month follow-up (median time: 272 days). This finding indicates that, despite treatment intensification over 2 years, patients still have poor glycemic control, so there is an opportunity to further optimize the pharmacologic management of patients with type 2 diabetes. This analysis did not explore treatment patterns following intensification, but further research into the real-world application of type 2 diabetes treatment pathways, as recommended by the ADA , would provide insight into where changes in prescription behavior may be warranted.
The HbA1c target of 8.0% (64 mmol/mol) used in this analysis to indicate suboptimal glycemic control was the preferred HbA1c threshold chosen by the Together 2 Goal® campaign, as it captured glycemic control across a broad patient population. Other HbA1c targets may be used by health care organizations (including individualized targets), where it is expected that the rate of observed clinical inertia would be greater with more stringent HbA1c targets (e.g., < 7% [53 mmol/mol]) .
This was a longitudinal real-world analysis of a large integrated dataset from a diverse population of patients and different types and sizes of health care organizations across urban, suburban, and rural locations in 19 USA states. Unlike previous studies, which presented the average level of therapeutic inertia across the health systems investigated, the present analysis looked at the variation in the level of therapeutic inertia across and within each of the 22 individual AMGA health care organizations investigated, providing insight into where improvements in overcoming therapeutic inertia might be made. Our findings show that, while there is significant variation in therapeutic inertia across health care organizations, there is even greater variation within each organization, at the site level and among providers at those sites. This variation means that, in many cases, organizations have a practice site or provider performing at the highest level, from whom others could learn. In turn, each organization has a practice site or provider among the lowest performers, who could benefit from an intervention. This idea of identifying ‘high performers’ within an organization and sharing lessons learned may be key to overcoming therapeutic inertia [26, 27]. These conversations could also serve to identify the most prevalent underlying reasons for therapeutic inertia.
Other studies of inertia have provided some insight into possible causes. According to Khunti and Davies, reasons for therapeutic inertia include overestimating the care needed when intensifying treatment, using soft reasons to avoid intensification, and a lack of education and training regarding glycemic goal attainment . Some physicians may not have the confidence to intensify treatment by prescribing newer injectable therapies in patients with more advanced diabetes . Physicians may also be reluctant to intensify treatment due to a perception of patient nonadherence [18, 28]. This may explain the greater treatment inertia observed in patients treated with multiple OADs or basal insulin at baseline. Nonadherence to prescribed drug treatments is prevalent among people with diabetes . Patient adherence may be influenced by the complexity and side effects of the treatment, negative media coverage [30, 31], a fear of hypoglycemic episodes and insulin-associated weight gain , and poor patient–physician communication .
Importantly, health care system-related factors might also influence the prevalence of inertia; these might include a lack of adherence to clinical guidance, no active patient outreach, a lack of screening and referral programs, high medication costs, inadequate cardiovascular risk assessment and risk factor control, inadequate support staff capabilities, and poor internal communication [13, 18]. Addressing these factors at all levels (patient, physician, and health care system) may be key to overcoming clinical and/or therapeutic inertia. Evidence suggests a need for additional patient education around their disease, treatment options, and utilizing available resources, perhaps with more involvement from pharmacists and more opportunities for remote management.
Given the retrospective nature of this analysis, and some specific features of the dataset, there are some limitations to consider. The dataset was biased toward White and commercially insured patients, and toward integrated systems and multispecialty, rather than small independent practices. Furthermore, because AMGA members represent the leading health care providers in terms of quality performance and provision of value-based care, these findings may represent an underestimation of the extent of therapeutic inertia in the broader population. In addition, as data were collected for routine patient management, they were subject to missing values. Data indicating changes in medication dose or lifestyle modifications were not available; nor were pharmacy dispensing data. Therefore, among patients for whom there was no observable clinical action (therapeutic inertia as defined by this study), it is unclear whether their medication dose changed or lifestyle modifications were made. Regardless, many patients had persistently elevated HbA1c levels. Additionally, the study population may not be representative of the USA patient population nor extrapolate easily to other non-USA health care systems or to low- to middle-income countries.