Therapy Escalation Following an Elevated HbA_1c in Adults Aged 45 Years and Older Living With Diabetes in Australia: A Real-World Observational Analysis

Stepwise escalation of glucose-lowering therapy to more intensive regimens is an integral component of type 2 diabetes management (1). Early blood glucose control is associated with beneficial effects on long-term micro- and macrovascular outcomes (2). Understanding factors associated with time to therapy escalation following an elevated glycated hemoglobin (HbA_1c) may inform interventions to facilitate timely control. We used linked real-world data to examine time to therapy escalation following a recorded HbA_1c ≥8.0% (64 mmol/mol) and associated factors.

The EXamining ouTcomEs in chroNic Disease in the 45 and Up Study (EXTEND45 Study) is an Australian population-based linked data study in which participants of the 45 and Up Study have been linked to the Medicare Benefits Schedule (MBS), providing data on government-subsidized medical services, the Pharmaceutical Benefits Scheme (PBS), providing prescription claims data, and community laboratory databases (among others) (3). The 45 and Up Study comprises 267,153 individuals aged ≥45 years living in New South Wales (NSW), Australia, recruited between July 2006 and December 2009. Ethical approval for the EXTEND45 Study was obtained from the NSW Population & Health Services Research Ethics Committee (HREC/13/CIPHS/49).

In EXTEND45, 24,430 individuals with diabetes between 2006 and 2014 have been identified using multiple linked data sources (3). In the current study, therapy escalation was assessed during the 6 months following all linked HbA_1c results ≥8.0% occurring in non–treatment-naive, non–insulin-using individuals presumed to have type 2 diabetes. As a result, the unit of analysis was the elevated HbA_1c, with individuals able to contribute multiple results to the analysis. A threshold of 8.0% was used to reflect the typically higher HbA_1c targets recommended for older people (4). To ensure that all included HbA_1c results represented separate instances of an elevated HbA_1c in the same individual, those results ≥8.0% that occurred within 3 months after the end of the 6-month follow-up period were excluded.

Therapy escalation was determined using Anatomical Therapeutic Chemical classification codes to identify prescription claims for different glucose-lowering therapy regimens in the PBS data set. Regimens were categorized as monotherapy, dual therapy, triple/quadruple therapy, and temporary discontinuation (defined as a break of two or more standard coverage days). Escalation was defined as the addition of ≥1 oral glucose-lowering drug, a glucagon-like peptide 1 receptor agonist, or insulin to the individual’s existing regimen.

Explanatory variables included age, HbA_1c result, glucose-lowering therapy regimen, number of general practitioner (GP) visits, and number of consultant (any specialty) visits in the previous 90 days, assessed at the time of the elevated HbA_1c, as well as fixed demographic, socioeconomic, lifestyle, and clinical variables. To account for individuals contributing multiple HbA_1c results to the analysis, marginal Cox proportional hazards models were used, with variable selection performed using the Hosmer-Lemeshow method (5). For each included HbA_1c result, follow-up ended when therapy was escalated, the individual died, or after 6 months, whichever occurred first.

A total of 4,009 eligible HbA_1c results ≥8.0% among 2,456 individuals were identified, with 60% of individuals contributing a single HbA_1c result. The majority (59%) of included results occurred in female participants and 63.6% in participants aged 55–75 years. Forty-four percent of elevated HbA_1c results (n = 1,774) were met with a PBS dispensing indicating therapy escalation within 6 months (median time 45 days, interquartile range 14–101 days).

In a multivariable analysis, therapy escalation was more likely for HbA_1c results occurring in individuals who had a higher HbA_1c (P < 0.0001 for linearity), were receiving monotherapy at the time of the result, had visited their GP multiple times in the previous 90 days (P < 0.0001 for linearity), had visited a consultant physician in the previous 90 days, lived in Inner Regional Australia (areas classified by the Australian Bureau of Statistics as having some restrictions upon access to services due to geographic distance), or had self-reported anxiety on the 45 and Up Study baseline questionnaire (Fig. 1). Age was nonlinearly associated with therapy escalation (P < 0.0001 for nonlinearity). Therapy escalation was less likely for elevated results occurring in individuals receiving dual therapy or triple/quadruple therapy, or who had temporarily discontinued therapy (Fig. 1).

To our knowledge, this study is the first to examine therapy escalation and clinical and sociodemographic associations in an Australian population using real-world data. Through the inclusion of a range of longitudinal data sources, we have been able to incorporate time-varying risk factors, such as HbA_1c level and therapy regimen intensity, into our analysis, better reflecting the longitudinal nature of diabetes management. Moreover, we examined a range of existing therapy regimens, including temporary discontinuation, increasing the generalizability of our findings to all individuals with type 2 diabetes aged ≥45 years. However, we were not able to assess dose up-titration or examine individuals’ adherence to their existing regimens.

Measuring time to therapy escalation in response to an elevated HbA_1c is a useful indicator of quality of care of people with diabetes, and our findings can be used as a baseline for assessing future changes. The factors identified in this study can be used to identify potential interventions for promoting therapy escalation in patients with diabetes, but these should be investigated further. In addition, future studies should examine the clinical and patient-reported outcomes of therapy escalation.