Articles & stories

Artificial intelligence allows detailed diabetes market research

The proprietary Credit Suisse Healthcare Database allows us to follow the therapeutic journeys through various diseases.

What the Credit Suisse Healthcare Database tells us about the diabetes market 

The Credit Suisse Healthcare Database, covering one-third of the US population, contains anonymized medical and prescription claims (delivered in both office/hospital settings and those given in outpatient settings). It makes it possible to create representative patient journeys in diabetes treatment. It tracks persistence on a therapy, compliance with a therapy, common treatment cocktails, and transitions onto and off specific drugs. 
We initiated our use of the Credit Suisse Healthcare Database by looking at the type 2 diabetes market, as this is an important growth area and one where there is much debate over the relative strengths different classes of drugs.
This data is most useful when looking at longer-term trends and may give us an idea of patient satisfaction with a diabetes treatment and help define diabetes market penetration, both of which are important for modelling sales.

Complexity of diabetes treatment

Diabetes is a chronic disease that causes the inability to manage blood sugar levels appropriately. Type 2 diabetes (when a patient produces insulin in response to the presence of blood sugar but the body becomes resistant to the insulin produced) makes up about 95% of the occurrences of the disease. According to the WHO, more than 400 million adults worldwide have diabetes today, and this figure is expected to increase to 700 million by 2024 (+51%). With the complexity and extensive duration of the treatment it is quite difficult to track trends, draw conclusions, estimate costs, or plan on improvements to the therapy of so many affected by the disease. However, the Credit Suisse Healthcare Database allows us to track patients over time and thus assess treatment choices, persistence of treatment, the use of drug cocktails, and likely follow-on choices.

Although several effective therapies are available today, many patients' management of the disease deteriorates over time, and they need to add more therapies or switch to new therapies to control glycemia appropriately. With time, we have seen new classes of drugs come to the market which offer both better blood-glucose control and better management of weight and other comorbidities of the disease.

Type 2 diabetes treatment journey

When diagnosed with type 2 diabetes, patients are prescribed diet and exercise. When that fails, patients are first prescribed metformin. Then they move to alternative and oral antidiabetic medicines (various drug classes, SU (sulfonyl ureas), DPP-4 (Dipeptidyl-peptidase 4 inhibitors), SGTL2 (Sodium-glucose cotransporter 2), and most recently the first oral GLP-1 (Glucagon-like peptide-1)). These newer drugs may be given either alone or more commonly in combination with metformin. Diabetes is controlled – but not cured – by this medication and as it progresses and patients fail to control their blood sugar levels, they typically move to more intensive therapy, including the option of an injectable GLP-1 therapy.

When all the above treatments have failed, patients are given insulin as the last-line option. Insulin is unfavorable for patients as the therapy causes weight gain. In contrast, the GLP-1 class lowers weight with increasingly potent treatments available that lower both blood sugar significantly and weight. GLP-1s also benefit from lower side effects – specifically, these do not cause hypoglycemia, which is a potentially serious side effect of insulin treatments.
The diabetes market is characterized by very high levels of competition, which can be seen in the growing difference between gross and net sales for key drugs.

Credit Suisse Healthcare Database key insights on diabetes

Thanks to advanced big data analysis, we were able to reach the following conclusions:

  • Increasing use of metformin in younger cohorts in the US suggests earlier intervention in the disease, or a greater incidence of type 2 diabetes in younger patients which will be sustained as this cohort ages. 
  • With more patients starting initial treatment at an earlier age, we expect patients to need more intensive treatment before the age of 65, when many patients transition to Medicare plans that still typically look to use a higher proportion of generic versus branded drugs.
  • Comorbidities appear unlikely to affect the uptake of branded diabetes drugs. Surprisingly, we saw no higher incidence of patients taking heart disease medication in our diabetes cohort than in an age-matched sample of non-diabetic patients. 
  • Most patients take more than one drug for their diabetes, which adds to costs.
  • Rates of abandonment after only one claim are high. In the real world setting we see a 50% drop-off in treatment in 13-16 months, despite the chronic nature of the disease.
  • The database highlights the growing use of Glucagon-like peptide-1 (GLP-1) drugs as well as Sodium-glucose cotransporter 2 (SGLT-2) versus Dipeptidyl-peptidase 4 inhibitors (DPP-4s).