STOCKHOLM – Insulin pump therapy is considerably more expensive than daily insulin injections for patients with type 2 diabetes, but the benefits it confers offset about 30% of the price difference.
A cost-modeling study showed that patients who use the pump gain almost another year of life expectancy and experience a delay in the onset of major complications, such as end-stage renal disease and amputation, Stephane Roze said at the annual meeting of the European Association for the Study of Diabetes.
Mr. Stephane Roze
Pump therapy also conferred a significantly improved quality of life over multiple daily injections, a benefit that carries its own economic value, said Mr. Roze of HEVA-HEOR, a health economics evaluation company in Lyon, France.
“Even though these patients are initiating the pump at an older age than patients with type 1 diabetes, the therapy still looks to be cost effective,” he said. “It’s a good investment for payers.”
Mr. Roze used the CORE Diabetes Model to examine the cost/benefit ratio of insulin pump therapy, compared with multiple daily injections in patients with poorly controlled type 2 diabetes who lived in the Netherlands. The CORE model comprises 14 sub-Markov models that run in parallel; it examines all of the micro- and macrovascular complications related to diabetes. Model input data were the cohort characteristics and 6-month clinical outcomes of the OpT2mise study.
The trial examined glucose variability in 331 patients with type 2 diabetes who were randomized to either pump therapy or daily insulin injections.
It is an excellent study for cost-modeling initiatives, Mr. Roze said. It’s multicenter and multinational, conducted in 36 institutions in the United States, Canada, Europe, Israel, and South Africa. The cohort is large and diverse. Importantly, all of the patients had their treatment optimized with daily insulin analogue injections before being randomized. That initial step lent credence to their baseline glucose control; despite optimization, they still had a mean hemoglobin A1cof 9%. The OpT2mise patients were a mean of 56 years old, with mean disease duration of 15 years.
At 6 months, the study showed that HbA1c had decreased significantly more in the pump group than in the injection group (1.1% vs. 0.4%). Patients using the pump also needed significantly lower daily insulin doses (97 unit/kg per day vs. 122 unit/kg per day).
These results were inputted into the CORE model, along with the costs of diabetes-related complications in the Netherlands and the costs of both pump therapy and daily insulin injections. For the first year of treatment, pump therapy more than twice as expensive as daily injections (4,407 euros vs. 1,597 euros). This difference remained when costs were extrapolated to year 2 and beyond (4,365 euros vs. 1,555 euros).
The model, however, determined that pump therapy has the potential to extend life expectancy by almost 1 year, compared with injection therapy, by reducing and/or delaying life-threatening diabetes complications. These clinical benefits translated into an additional 9.38 quality-adjusted life-years (QALY) for patients using the pump, compared with 8.95 QALY for patients using injections.
Over a lifetime of treatment (13 years), pump therapy would cost a total of 58,024 euros, compared with 20,237 euros for injection therapy – a difference of 37,787 euros. Compared with injection therapy, though, pump therapy would save almost 10,000 euros in the cost of renal complications; 625 euros in the cost of ulcer, amputation, and neuropathy complications; 152 euros in the cost of eye complications; and 410 euros in complications due to hypoglycemia. These savings reduced the total cost difference between the two to 27,051 euros, Mr. Roze said.
The benefits held over time, despite the paradox of survivorship, he said. “The longer a patient with chronic disease lives, the more opportunity they have to develop a complication.”
The incremental cost-effectiveness ratio was 62,295 euros/QALY – an amount that falls well within the Netherlands’ “willingness to pay” threshold, which is a measure of what payers or society are generally willing to accept as a good value for health care expenditures.
The newest data from OpT2mise, also released at EASD, bolster Mr. Roze’s modeling. In phase II of the study, patients who had been on injections crossed over to pump therapy. Within 6 months of crossover, these patients experienced a mean HbA1c decrease of 0.8%. By 12 months from study start, when all patients had been on pump therapy for at least 6 months, the mean HbA1chad dropped to 7.8% from baseline.
“We were reassured to see the 12-month data, which confirm the sustainability of the effect,” Mr. Roze said. “This is one of the main questions when you do health care modeling.”