The Diabetes Treatment Lottery

The best and most affordable treatment for type 2 diabetes

Since the introduction of thiazolidinediones in the 1990s, the idea of the significance of mechanism of action as a core element of pharmacotherapy in diabetes has been growing in importance in the discussion of population health and policy.  The development of physiologic insulin regimens and incretin agents further pushed the conversation away from therapies, which increased undirected insulin. These developments came to a crescendo when targeted cardiovascular trials like EMPA-REG and LEADER demonstrated rapid cardiovascular benefit where more general A1c-oriented trials failed.

As clinical opinion shifted in the diabetes community, policy came into question.  Leaders in the community, like Ralph Defronzo, suggested that drugs like sulfonylureas had little place in the treatment of diabetes and should be removed from clinical use.  Others like David Nathan reminded the community that large trials like UKPDS used sulfonylurea therapy in treatment regimens to generate microvascular benefit. This is further amplified by a 50-to-150–fold difference in price between the inexpensive sulfonylurea therapy and agents like SGLT-2 and GLP-1 agonists.

The concept of cost in chronic disease is complex. Short-term cost of therapy is relatively easy to assess, though factors like efficacy and adherence have many patient-specific factors. Cost of disease burden is considerably more difficult as both short-term complications, such as hypoglycemia, and long-term cost, such as cardiovascular disease, must be considered. In order begin this kind of analysis for a complex population, simulated decision making needs to be explored.

The GlucosePATH decision support software was designed to help patients and providers make practical scientific and economic health decision at the bedside. The clinically validated software takes data from randomized clinical trials, patient’s past medical history, and clinical preferences through an algorithm that can rank over a million therapeutic combinations. Advances in processor speed and parallel processing now allow decision support to be adapted to simulated population management.

The team at GlucosePATH is beginning a complex analysis of optimized decision making for several treatment paradigms in multiple economic scenarios. Populations will be optimized for our current pharmaceutical environment as a control, an environment without sulfonylureas and similar drugs, and finally, an environment where both sulfonylurea and non-physiologic insulin regimens are eliminated.

For this analysis, certain base values must be established in terms of A1c, chronic disease state benefits and risks, and cost. Given the current state of recommendations and evidence, A1c is used as a primary goal. In order to account for variability in the base population and the 3-month look forward of the software, a goal of 7.5% was used.  Results were optimized with no minimum score in other areas to provide a baseline. If the patient’s new baseline is worse than the baseline of the control, there is an attempt to further maximize total score by increasing the maximum available cost.

While results are early, it appears as if the effect of the elimination of overinsulinizing agents is variable depending on focus of cost. Patient cost and A1c outcome differences were negligible with some of the shifts in prescribing moving toward less-used agents like bile acid sequestrants and dopamine agonists. The most striking changes were observed in a hybrid model measuring payer and patient cost. While elimination of sulfonylurea as an option did not change HbA1c, cost increased by 11%.  Elimination of both sulfonylurea and non-physiologic insulin therapy increased A1c by 0.26% with a cost increase of 14%.

These findings illustrate the challenge that faces providers in the era of risk contracts and population health. While traditional commercial models allow for near ideal implementation of modern diabetes strategies, challenges develop when coinsurance or shared-risk models are implemented. As such, it will be important to be able to measure benefit in economic terms in addition to laboratory markers.

Comment:  Dr. Stanley Schwartz – “Justification of the use of agents that avoid hypoglycemia, minimize risks of hyperinsulinism (that increases risk of ASVD, cancer ,dementia) will be globally more valuable, even  with ‘borderline economic value assessments,’ as one takes wonderful decision support software such as Glucopath and combined with ‘omics’ data, apply ‘on-the-fly,’ eg: implementing ‘dynamic decision support’, at the point of care, allowing one to fulfill the promise of Precision Medicine.”

If you are interested in collaborating on a population health simulation on diabetes therapies, please contact us. Email: publisher@diabetesincontrol.com and request more information.

Dr. Bradley Eilerman is an endocrinologist in Covington, Kentucky. He received his medical degree from University of Kentucky College of Medicine and has been in practice between 20+ years.  Glucose Path is decision support software for physicians treating patients with Type 2 Diabetes. It helps physicians deal with the complexity of finding the optimal set of medicines, from among the millions of possible combinations. Our population analytics tools can help payers achieve better outcomes at substantially lower costs.