Within the 18 months of the project, two WP were active in parallel: WP1 and WP2, respectively focused on business and technological feasibility validation.
Concerning WP1, the business feasibility study started with an analysis of the insulin delivery scenario, considering all the devices now available and different approaches to insulin infusion. Then, a list of stakeholders including patients, diabetologists and insulin delivery device manufacturers were selected and then contacted, following a user centred design approach. Starting from the information collected, it was possible to define two typical end-users’ customer journeys – discovering their pains and gains – and to map performances and limitations of current insulin delivery device, until to profile an ideal insulin pump. All the qualitative information gathered, along with what was found on competitors, were then translated into technical specifics for the R&D using the House of Quality analysis tool. Despite preliminary assumptions, the analysis revealed that current devices are very advanced for what concerns all the hardware components - especially patient’s data management software – but not so much in infusion performances. Regarding the intellectual property assessment, the PRISMA micro-pump was firstly analysed as project result using a specific Key Exploitable Result (KER) table. This result was then compared with all the pertinent IPs, with focus on tri-hormones closed-loop system, smart patch technology and algorithms regulating the drug administration. On the other hand, the preliminary Freedom to Operate (FTO) analysis carried out did not reveal any patent potentially in contrast with PRISMA technology.
A preliminary Health Technology Assessment (HTA) followed, analysing cost-effectiveness and cost-utility for patient and health systems deriving from an improvement in Continuous Subcutaneous Insulin Infusion (CSII), such as that promised by PRISMA micro-pump adoption. In this sense, both systematic reviews and on modelling-based results were considered.
The assessment confirmed the potentially key role of the pump’s performances, even in economic terms. A in depth market analysis was then carried out, collecting information on market share and trends of diabetes therapy’s market, listing also the most relevant insulin pump manufacturers and the specifics of their devices.
This paved the way to the definition of a Product development roadmap, in which development status is resumed and the best way to exploit the project result was suggested.
As for the WP2, the first task performed was gathering knowledge on insulin therapy and the products available. The goal was to identify the necessary figures of merit to define our device specifications. We investigated the insulin standard dosage and the bolus mechanism. Then we studied the feasibility of a three-hormone therapy, which quickly resulted in not being practicable. We compared several closed-loop systems already available for diabetes patients and identified the minimal technical requirements for our device. We evaluated the feasibility of microneedles patches as delivery vectors. The idea of a small-sized and painless injection device is attractive, but after interaction with suppliers, physicians and literature review discarded the idea. We came to know microneedles are not suitable for long contact with the skin and are considered unreliable for continued insulin administration. CeOx and Day One were in close contact during these tasks to share useful information. Once identified the minimum and market-ready specification of an insulin pump, we integrated the information with our electrostriction knowledge and designed a base model. Through simulation we evaluated the performances of our pump in terms of energy consumption and flow rate. As the results were not satisfying for a market-ready device, we performed a parametric study of the model to optimize the features of the pump. The performances achieved were above the requirements defined, so we proposed an optimized final model.
We assembled then two types of prototypes: a multi-membrane one, working on peristaltic mode, and a single membrane one. Both modes were tested in a real fluidic circuit, using an optical microscope to measure the flow rate. The peristaltic pump did not show pumping action, probably due to inconsistency in each membrane module. The single pump instead, achieved a satisfactory flow rate.
