Medical innovation : using mechatronics engineering to reduce inequities in healthcare.

Abstract

Medical device innovation provides access to healthcare. Innovations come about be- cause of pressures, in particular financial pressures, and access to care. With increasing interoperability of devices, distinction is made between devices with specific interoper- ability (SIO) only able to communicate with a pre-determined range of other devices, and non-specific interoperability (NSIO). Devices with NSIO pose substantially greater potential benefits by allowing long-term system wide innovations.

Scales of innovation are discussed, where short-term innovations meet an immediate need, such as the inundation of intensive care units (ICUs) in the COVID-19 pandemic. Medium-term innovations see either incremental increase in efficiencies, or an increase in interoperability which enables subsequent innovation. Long-term innovations are disruptive, systemic changes, often enabled through the use of increasing interoperability. The uptake of innovation is often lacking, but through the use of a framework such as Tech-ISM the chance of adoption is increased. This framework sees establishment and fostering of close relationships with a range of end users, decision makers, and industry partners.

Diabetes technologies are presented as examples of innovation. Insulin pumps are an effective method of delivering insulin, and see considerable benefit in control. Widespread adoption of insulin pumps is posed through the development of an ultra-low cost (ULC) insulin pump, made possible by the separation of hardware and computation, and costing 12 × −20× less than currently-available devices, both for a traditional-style insulin pump, and also a novel spring-driven design.

Initial results show similar accuracy to current commercially-available insulin pumps, with a mean error of 0.64%, the same as the MiniMed™640G (Medtronic, Dublin, Ire- land) for 1 U boluses, and mean error of 0.06% for 10 U boluses. Basal windows of 1 hour are similarly accurate, with 100% within ±15%, 92% within ±10%, and 84% within ±5%, again very similar to the MiniMed™640G. The ULC insulin pump is a solution to the economic infeasibility of insulin pumps for the majority of New Zealanders.

System-wide adoption of insulin pumps would see considerable economic benefit for New Zealand, in particular with a patch pump. Several possible adoption scenarios are presented. Annually, direct savings associated with less insulin use and current public investment in insulin pumps is expected to total $6.6M - $25.3M, indirect savings from reduction of expensive complications are expected to save $2.5M - $25.5M, with direct costs of $0.8M - $25.7M. Projections are for a total overall system saving of $8.3M with no additional uptake of insulin pumps, but only replacing current insulin pumps with the ULC alternative, to $25.0M with widespread adoption. These figures do not account for additional savings made possible through future long-term development of smart, automated healthcare systems.

A continuous glucose monitor (CGM) is a device that estimates blood glucose (BG) every 1-5 minutes, replacing discrete, invasive self-monitored blood glucose (SMBG) measurements as required four to ten per day currently for approximately 40,000 - 60,000 New Zealanders with diabetes who administer insulin. Current CGM use is limited, but relatively unknown, due to no public funding, with expert estimates at 2-8% prevalence among individuals with type-one diabetes. A low-cost alternative is presented in the form of the blood optical biosensor CGM (BOB CGM) at an annual cost 10 × −20× less expensive than current devices. Initial, un-calibrated results show promise, with 91% of BG results deemed clinically accurate, and a further 8% sufficiently accurate to not cause treatment error. Fundamentally, cost savings arise from allowing access to otherwise inaccessible data, and thus turning the current data monopoly into a data market. Substantial economic benefit is seen from direct savings from current monitoring of diabetes disease progressions with SMBG and glycated haemoglobin (HbA1c), and also indirect savings from earlier identification of worsening diabetes control. Various adoption scenarios are presented, with overall annual economic savings of $1.9M - $25.1M.

Another medical innovation is presented in the form of the actuated, closed-loop, time- series inspiratory valve (ACTIV) dual ventilation system. This innovation is a short- term example, developed under pressure of inundation of the healthcare system due to the novel coronavirus disease (COVID-19). The basic operating premise is ventilatory effort from a single mechanical ventilator is delivered first to one patient, and subsequent to a valve switching state, to a second patient. The system is a solution that addresses valid concern for multiple ventilation from a consensus of oversight bodies for ICU treatment, in particular personalised therapy and monitoring, especially in the case of changing pathology. The system is designed to be low-cost, robust, portable, and readily manufactured in low-resource environments. Thus, it has an Arduino (Arduino, Massachusetts, USA) controller, and requires a 5.0 V power supply.

The system requires a flow and pressure sensor for detection of inspiration, and sub- sequent valve switching. A custom-made 3D-printed Venturi interfaced with simple electronics with an analogue 0.0 − 5.0 V output signal is presented. The sensor is validated against data from mechanical ventilation devices to be accurate over the range of 5 − 75L · min−1, with a Pearson Correlation ≥ 0.95 for flow and pressure, typically ≥ 0.97 in 5 S bins at fs = 50 Hz. Additional components are a 3-D printed pressure drop device in the form of the PANDAPeep Gen2 Inline valve, and off-the-shelf one-way valves, airway filters, and 22 mm⊘ tubing.

The switching ACTIV valve is another 3D-printed component, and uses a common HXT12K servo motor, or similar, for interoperability. The Arduino-based control system is a basic finite state machine (FSM) relying on low-pass filtered flow sensor data, implemented through a circular buffer, for state changes. These state changes, in combination with various necessary delays for safety, dictate the change of state of the ACTIV valve, and thus to which patient ventilation effort is delivered. An example of two considerably unbalance patients, with compliance C = 0.10L · cmH2O−1and C = 0.05L · cmH2O−1, being safely and efficiently balanced to achieve equal tidal volume is demonstrated.

Without innovations such as the diabetes technologies and ACTIV system, care will become increasingly rationed. Rationing of diabetes devices with high effectiveness, but also high cost, is already seen in the lack of public funding for CGM devices, and funding for only 8-10% of individuals with type-one diabetes to access insulin pumps, despite significant proven benefits of both of these devices. Since 2000, increases in direct out- of-pocket expenditure have grown an average of 4.3% per annum, compared to median wage growth of 3.2%, and inflation of 2.6%. These trends show that the rationing of healthcare is being seen in a reduction of access to publicly-funded services. Given an average annual wage increase of only 1.6% for the lowest 20th centile, individuals who are least well off are less able to afford the required personal expenditure to attain the same access the healthcare. Therefore, access to healthcare is seeing worsening equity of access.

With increasing demand for healthcare, and a taxation base stagnant at best, relying only on intrinsic changes, New Zealand faces significant taxation increases, or dras- tic reductions in healthcare services. The alternative is to increase the efficiency of healthcare delivery methods, using extrinsic, disruptive changes. These changes are only made possible through innovation informed by strong clinical insight, developing mechatronic devices with broad, non-specific interoperability, if not open-source design. This approach provides equitable access to care, and provides the necessary framework for automation of healthcare services, including diagnostics, prognostics, and person- alised care models under a one-method-fits all approach. This widespread technological innovation and adoption poses significant increase of access to care, combating current inequities.