Simulation test: can medical devices pass?

Introduction

Simulation-based education and training can enhance healthcare professionals’ knowledge, skills and attitudes in a safe environment, without patient harm. Traditionally, simulators are used to train or measure procedural-based skills and teamwork behaviours. Studies have demonstrated that some skills trained using simulation are transferred to the real world; furthermore, inferences can be drawn between performance levels.¹

While improving individual and team performance, the overall philosophy of simulation is to develop safer healthcare for professionals and to increase patient safety. However, actual robust measures of safety gains directly attributable to simulation training are often difficult to measure. Simulation centres with their expertise and technologies could benefit patient safety in additional ways by investigating and empirical testing of the usability, suitability and safety of medical devices. Similarly, the human factors aspects and training needs required for the safe and effective use of medical devices can be elucidated within the simulation environment. Additionally, early preclinical testing of medical devices in the product cycle within the simulated environment offers the opportunity to avoid costly mistakes or poor designs from progressing further downstream in the development cycle.²

Many patients exhibit changes in respiratory rate (RR) in the hours leading to a cardiac arrest, and also in the early onset of sepsis and other heterogeneous medical conditions. Early detection of changes in vital signs may allow for early interventions to improve prognosis of seriously ill patients and to prevent admissions to intensive care. However, RR is frequently not recorded or infrequently measured; furthermore, the optimal frequency of vital signs measurements is largely unknown. Continuous monitoring with wearable devices offers the opportunity to identify earlier deterioration, identify subtle changes and alert the clinician to take action.³

The RespiraSense device comprises of three components: a reusable sensor that attaches to the patient, a lobe that attaches to the sensor and uses algorithms to remove artefact and noise, and also alerts care givers when certain thresholds in RR are reached; and a wireless tablet device that allows care givers to interface with the device, set thresholds, and monitor and retrieve information. The RespiraSense sensor measures chest and abdomen deflection during breathing to directly measure RR.

Project Aim

The ASSERT Centre, University College Cork, used a human patient simulator (HPS CAE) and intensive care ventilator (Hamilton G5) in preclinical testing of a continuous respiratory sensing monitor (RespiraSense PMD Solutions) to determine the device’s dynamic response measurements and stability over time following changes in ventilation.

Methods

The device was attached to the HPS in the appropriate position (figure 1). The HPS simulator was chosen for its high respiratory and ventilation functions and efficiency. In order to accurately control RR, a G5 Hamilton Medical intensive care ventilator was used to ventilate the manikin. The HPS was intubated with a size 8.5 mm endotracheal tube and ventilated on pressured controlled mandatory ventilation (CMV). Ventilation settings were pressure control of 15 cm H₂O and positive end expiratory pressure of 5 cm H₂O with an inspiratory:expiratory ratio of 1:2. These parameters were chosen as they are physiologically representative of a normal ventilated healthy patient. Normal ventilation and pressure values were achieved throughout the study.

Figure 1.

RespiraSense device and simulation test setup. HPS, human patient simulator; ICU, intensive care unit.

Results

The device’s dynamic response measurements were observed following changes in ventilation RR, ranging from 6 to 60 breaths/min. The breaths were increased in increments of 1, every 2 min. The device measurements and the ventilator’s CMV rate were plotted and analysed using the Bland-Altman method, revealing both measures within 95% limits of agreement for the difference of the means (±2 SD).

Second, stability over time was measured, where the ventilation rate was set for a period of 4 hours and was measured against the device’s performance. This was repeated three separate times, for 4-hour periods, of 10, 20 and 30 breaths/min. The device allows for historical and real-time plotted graphs of RRs, which were measured against the set CMV rate over the periods.

The device provided accurate and stable measurement of RR, including RR extremes in a controlled environment, using an HPS under mandatory controlled ventilation.

A limitation was that despite the high congruence between the CMV RR and that measured by the device, the manikin was not moving at any time, as a human would be. Therefore, any affect from artefact or noise would have been absent, which would not reflect most real-life uses of the device.

There were occasional connection problems that were identified between the sensor and the hardware device during the trial, which were communicated to the product engineers.

Discussion

With advances in technology, there are many more medical devices available than ever before. However, the environment in which they are designed, deployed, developed, certified and regulated in is increasing in complexity. The Global Harmonization Task Force on Medical Devices has been tasked with the harmonisation of regulations between jurisdictions, and more recently, the International Medical Device Regulators Forum has increased standards and changed guidelines and rules to ensure very high product standards and high levels of patient safety.⁴

The use of simulation centre expertise and technologies by engineers, clinicians and others in the medical device field may expedite product design, identify design errors, increase end-user satisfaction and reduce patient harm.² There seems to be a paucity in the literature of preclinical testing of medical devices in this manner, despite being strongly advocated in David Gaba’s 2004 paper.⁵

The traditional medical device’s ‘build and test’ product development cycle, which is essentially building, bench test, animal test and human test, is very expensive and time-consuming. There is a need for medical device companies to move away from this approach of product design. Medical device companies have also been slower than other industries in adapting computational modelling and simulation in their design strategy.⁴

Conclusion

Evaluating medical devices using simulation offers the opportunity to discover and correct design errors, which may not become apparent until later, during clinical trials or postmarket surveillance.