Training System
The present invention is concerned with systems for training medical practitioners.
Medical equipment becomes ever more complicated requiring careful training of its users whether they be in hospital, at home, in the workplace or in other environments. It is increasingly recognised that medical training needs to be formal and subjected to regular and 'rigorous scrutiny if the level of care given to patients is to remain satisfactory.
When a new piece of equipment is introduced into a care environment the burden of training typically falls upon either the manufacturer or the institution responsible for the care environment. Many attempts have been made to reduce the burden of training by developing computer or video tape based programmes, which explain the functions and use of the equipment. Some of these are automated and interactive so that trainees may assess their progress against a stock of questions or problems commonly encountered during use. Other methods of training include courses of lectures or demonstrations in situ by the • representatives of medical manufacturers, or trainers who are either internal or external to the organisations.
WO 01/15043 discloses a system for online medical data management and training which enables users to access information.
US-A-6, 074,213 discloses a simulator which uses telecommunications for training medical teams, the members of which are located in different places as they train together
using apparatus in co-ordinated fashion without input from a live instructor.
US-A-5, 882, 206 discloses a virtual surgery system or virtual testing system which provides a simulation or test based on image data .
US-A-5, 957, 699 discloses a remote computer assisted professionally supervised teaching system. A student user of a client computer system uses a teaching process to promote development of cognitive skills of the student and a supervisor uses a second client computer to remotely monitor the progress of the student.
None of the known training systems allow one to integrate training on dummies or simulators with training and/or procedures on patients. All the existing training systems fail to bridge the gap between dummies and/or simulations and actual patients .
Preferred embodiments of the present invention aim to allow the integration of training exercises on dummies or simulators, training exercises on actual patients and care of actual patients .
Preferred embodiments of the present invention comprise a training system which is based in bedside medical equipment and which is able to target those areas of knowledge and training which are specifically related to problems encountered during clinical use of the medical equipment. All existing training systems fail to bridge the gap between the use of equipment on dummies or via simulations, and its use on patients. Preferred embodiments of the invention aim to create a virtuous circle of care, training and continuing
education by allowing training to continue during clinical care.
Currently, patients undergoing intensive care are monitored by measurement of their physiological signs (e.g. blood pressure, levels of blood oxygen) without monitoring of the activity of life-saving equipment or its specific effects on the patient's condition.
Training in the use of equipment is often attempted for the first time on patients who are dangerously ill and unstable.
Preferred embodiments of the invention will help to ensure that staff are trained before they use the equipment on patients for the first time, and continue to develop their skills in the application of the equipment during patient care.
The bedside training system (preferably a computer (PC) based platform) will deliver the training, maintain a database of staffing skill levels and document the acquisition and maintenance of skills.
The computer-based platform will connect to sensors for use on models and on patients, as well as monitoring the activity of the equipment, itself, during training and during clinical care .
The circle of theoretical training, practical patient care, the application of medical knowledge and individual patient response will be closed by software which will bring simulations and real events closer together to allow training scenarios to be developed which are ever closer to reality.
The system may also develop a database of equipment usage in each medical or care institution for the purposes of audit, training and research. Analysis of this database will allow the likelihood of success of proposed treatments to be assessed before use. During use on patients deviations from the median response to therapy will action system alarms allowing deterioration to be identified early.
Preferred embodiments of the present invention will be described with reference to the attached figures, in which:
Figure 1 is a schematic diagram illustrating the various components of a medical training system embodying the invention;
Figures 2 & 3 are schematic diagrams illustrating the various components of a system for training carers in the use of a respirator for an infant;
Figure 4 is a flow diagram illustrating the different routines of a preferred embodiment of the invention;
Figure 5 is a flow diagram illustrating the user identification routine of figure 4;
Figure 6 is a flow diagram illustrating the session restart routine of figure 5
Figure 7 is flow diagram illustrating the display educational materials routine of figure 4;
Figure 8 is a flow diagram illustrating the manikin set-up and calibration routine of figure 4;
SUBSTITIF Ε SHEE" .E 26)
Figure 9 is a flow diagram illustrating the on-manikin training routine of figure 4;
Figure 10 is a flow diagram illustrating the patient test set- up and calibration routine of figure 4;
Figure 11 is a flow diagram illustrating the patient baseline recording routine of figure 4; and
Figure 12 is a flow diagram illustrating the patient or certification routine of figure 4.
Figure 13 is a flow diagram illustrating data collection during either the manikin routine of figure 9 or the patient routine of Figure 12;
Figure 14 is a flow diagram illustrating data collection following either the manikin routine of figure 9 or the patient routine of Figure 12;
Figure 15 is a flow diagram illustrating a process for evaluating data subsets collected in the routines of Figures 13 and 14; and
Figure 16 is a flow diagram illustrating a process for evaluating the data collected from the different data subsets such as task duration and recovery time; forces applied to patent (or manikin) ; and the variation in values from those recorded during the baseline period.
Preferred embodiments of the invention relate to a system for medical training in the use of specific items of equipment in hospitals, at home, or in other environments. It is designed to facilitate and enhance training in the use of complex
medical equipment by allowing the effects of simulated and real use of the equipment by a trainee to be monitored and documented continuously from the moment of first contact with the equipment to the trainee's first applications to patients and until the trainee is fully trained. Equipment function and each patient's progress may then be monitored continuously through the computer or PC-based system
The system consists of hardware, software and a database of training materials. Considering these components in turn.
Hardware components
These may include: (a) - sensors inside, or for use solely on, a physical simulation such as a training doll or manikin
(b) - sensors for external attachment to a training doll and/or a patient
(c) - a PC-based platform for display, user interfaces (d) - interface and cabling for connection between sensors or other patient-connected equipment (e.g. ventilators, vital signs monitors) and the PC-based training platform (e) - printer for certification of trainees
Software components
These may include:
(a) - interface graphics
(b) - audiovisual material presented appropriate to trainee' s level of training and individual requirements
(c) - examinations of trainees' knowledge before each new level of training is attempted
(d) - prompting for application of sensor array to physical simulation
(e) - comparison of multiple attempts at placement/use/configuration of equipment
(f) - prompting for application to the first patient under clinical trainer's supervision (g) - comparison with simulation and with further clinical use by the same and other trainees
(h) - local, institutional database of trainees, equipment, patterns of use, patterns of response by patients, skill-mix of carers (i) - central patient, demographic and equipment usage database
Equipment-based components
These may include:
(a) - components designed to give an output to the PC-based hardware/software relating to functions and conditions within the medical equipment (e.g. the duty cycle of a mechanical ventilator) (b) - outputs might include: alarm status, output level, display/interface status, duty cycle, power consumption, oxygen content of gases
(c) - these outputs may be used to monitor the patterns of use by trainees and will be collated by the PC-based software (d) - longer term monitoring of these outputs may also be used to determine the effects of therapy in relationship to e.g. pressure applied to a patient's airway, or the amount of oxygen being required at a given stage in an illness (e) - these outputs may then be compared with other institutions/carers/trainees
Training materials
These may include:
(a) - stored, audiovisual e.g. still pictures, x-ray images, etc
(b) - database on bedside PC.
The inter-relationships between some hardware, software, equipment and training materials components is shown schematically in Figure 1 for an embodiment of the invention.
Patient and manikin sensors 1, 2 (which may be the same) are provided for fixing to a patient 3 or a manikin 4 on whom or which a carer or medical practitioner is being trained, certified or monitored. The manikin and/or patient are subjected to one or more pre-defined actions or procedures by the trainee, trainer or supervisor, and outputs are generated by sensors during this or these procedures. There may be other non-patient and non-manikin sensors, such as environmental temperature, ambient oxygen, and noise level whose output is monitored by the system.
A hardware interface 6 (i.e. cabling, servers etc) allows the sensor outputs to be supplied to the PC or computer 7 having the training software. The training software then manipulates or processes the sensor outputs to produce prompts for the appropriate trainee or trainer to take certain actions.
The local database on the local PC or computer stores single and aggregate scores achieved by individual trainees. Data may be transferred to and from a central database 8 from and to the local database 7. The central database forms a central knowledge source accessible by other trainers or trainees and allows one to build up data for a larger population than is
possible locally. Collecting and storing and interrogating data from a larger population allow one to better monitor uses of the equipment and better determine what are optimal methods of use and/or training.
All the software components may be built into a particular piece of equipment (e.g. a mechanical ventilator), or may reside within a purpose-built device comprising a PC platform, user interfaces, and hardware interfaces . Thus, training and its effects on patients, may be carried out at the bedside and may continue to be monitored during longer term use of the medical equipment for which training is sought.
None of the existing automated methods of training address adequately the problem of relating the patterns of use of equipment during the training programme and its effects on patient physiology, comfort or well-being. Once a new user of equipment has been trained it is then difficult to achieve continuous monitoring of their performance in the use of the equipment and the effects of those patterns of usage on patient care.
Preferred embodiments of this invention are designed to improve the training of carers using equipment for the first time by:
1) allowing a direct transfer of skills acquired during simulated training sessions to on-patient procedures:
2) monitoring the effects of training on an individual's use of equipment; and 3) monitoring the effects of patterns of usage of the equipment by the same and different individuals on the health of the patient.
During the initial training period in the use of new equipment the ^learning curve' is quite steep and during this time patient care can be adversely affected unless constant supervision of the new trainee is undertaken. The preferred embodiment of the invention described with reference to figures 1 to 16 will allow carers to learn more before their first attempt at the use of the equipment on a real patient. It should also shorten the overall period of training and may reduce the number of patients subjected to use of the equipment during the carer's learning curve, particularly the initial steep portion of the learning curve where the carer or medical practitioner is more likely to misuse the equipment.
The system described should give an early warning of treatment failure or inappropriate use following introduction of the new equipment and it should ensure critical assessment and reassessment of therapy both in individuals and in groups of patients on whom the equipment is used. The system should help trainees by giving them confidence in their ability to deal with different categories of patients, different clinical conditions and a specific piece of equipment before their first use of the new device on a real patient.
The system will also help to ensure that the abilities of trainees are documented for the purposes of continuing professional development or continuing medical education and for demonstration that they are up-to-date in their training in the use of the device both for the purposes of employment and clinical governance. This is essential to good patient care and may reduce the costs of medical liability insurance.
The institutional and equipment company trainers should be helped because a standard platform and approach to training can be adopted. Direct feedback on the use of the device may
be given when trainers are not in attendance and guidelines for use of the device may be incorporated into the training algorithms. The equipment may also be set to alarm when use by the trainee or response by the patient is outside certain limits. Documentation of the status of trainees within the institution, scheduling of their workload and managing staff mix within institutions will be facilitated.
Equipment manufacturers should be helped by allowing redesign to occur according to the problems and benefits seen in real use, and by allowing new products to be introduced which address the specific clinical problems identified. There should be a reduced requirement for sales staff to train clinical specialists. The monitoring of the maintenance of devices and the likely ordering requirements for disposable elements of the equipment should be facilitated.
One example of a practical or preferred embodiment of the invention is shown in figures 2 and 3. These figures show a training system whose primary purpose is to improve the delivery of care provided in neonatal intensive care units and other centres using respirators to aid the breathing of infants. The EME Infant Flow™ (Trade Mark of Electro Medical Equipment Ltd) product family includes examples of such respirators. This range of products applies a continuous positive airway pressure (CPAP) to the airway of an infant or child by means of equipment which regulates air and oxygen flow through a nosepiece or mask which must be carefully attached to the patient's head and face.
The range of Infant Flow™ products comprises: (a) a flow driver 9, which regulates the flow of air and oxygen from medical gas pipelines and/or cylinders to generate a controlled flow of gas with operator-prescribed oxygen
concentration; (b) gas and electrical supply connection means; (c) a patient breathing circuit, which delivers said controlled flow of gas to a patient attachment; and (d) a range of patient attachments comprising a pressure generator and either nasal mask 10 or nasal prongs, which must be carefully attached to the patient's head and face such as to deliver a controlled flow of gas at a controlled and continuous positive pressure to the patient's airway (CPAP - continuous positive airway pressure) .
It will readily be appreciated that whilst the bulk of the discussion which follows is concerned with training on the use of a respirator for a prematurely born infant, embodiments of the invention may be used to train carers in the use of any piece of equipment for use on a patient.
Referring to figure 2 which illustrates the system set up for manikin based training, the system includes a manikin 4 having sensors 2 for monitoring the manipulation of the manikin during training. The sensors may include linear acceleration, torsion and load sensors. The outputs from these sensors pass to a computer 7 having the training software, an associated display 11 and input devices. Sensors 12 attached to the respirator nasal attachment and the EME flow driver also provide their output to the training software.
The training software compares the sensed outputs to stored outputs in a local and/or central databases and prompts the trainee or trainer accordingly.
Referring to figure 3, the system is used to monitor procedures carried out on a patient 3 using the same apparatus as used for procedures carried out on a manikin except that rather than sensors integral to the manikin, equivalent
12
SUBSTITUTESI-
external sensors 1 are attached to the patient. As well as acceleration, torque and load sensors, outputs from physiological sensors monitoring physiological parameters of the patient such as, e.g., heart rate, oxygen saturation may be monitored and/or stored. The combination of physiological patient outputs, trainee and equipment outputs allows one to better monitor and determine optimal, sub-optimal or possibly dangerous routines, actions or practices.
The preferred embodiment of the invention shown in figures 2 to 16 and described below is intended to improve the quality of training provided to care-givers through enhanced feedback regarding the application of medical equipment such as, e.g. the Infant Flow™ products to training dolls and individual patients.
The system comprises several separate components. These include a manikin 4 (see figure 2) or training dummy, manikin and patient sensor array (or arrays) , software and integration hardware and training resources .
For training in the use of products or equipment aimed at caring for a preterm infant such as the Infant Flow™ family of products, the manikin 4 simulates the physical properties of the preterm (i.e. prematurely born infant) skull and face. It incorporates sensors to determine specific stresses applied to the head, and electronic circuitry to encode these in a suitable form for transmission to the processor. For training in the use of other medical equipment, appropriate manikins representing the likely patients' characteristics would be used.
The manikin of the preferred embodiments is intended to be used as a training aid for teaching the correct handling of
the pre-term infant. It is used in the classroom situation to allow incorporation of feedback regarding the student carer's skills into the training programme. The manikin provides feedback in the form of a digital signal to an analysis system on a processor such as a personal computer.
The manikin and patient sensor array affixes to the respirator patient interface (e.g. nasal airway pressure generator, affixing straps, and bonnet) and provides in an electronic and/or visual format information regarding the stresses (mechanical and physiological) imposed on the manikin or, later, the patient (see figure 3) .
The sensor array is intended to be fastened, or possibly form a modification to, the respirator patient interface, such that feedback may be provided regarding the physiological effects and stresses imposed during application of the interface, firstly to the manikin, and secondly to the patient. The sensor array (or arrays) may provide feedback both via digital signalling to computer software, and possibly via direct indication.
Software is hosted on a Windows PC or similar platform, which can integrate the information from the manikin and patient sensor array into the provision of training tailored to the requirements and skill level of each individual trainee.
The software comprises the framework for a flexible and tailored multi-media training programme, incorporating feedback regarding the trend of improvement in each individual care-giver's skills derived from the automated data capture and from incorporated questionnaires and problem-based testing.
Multi-media training resources may be combined with the other elements of the system to provide the framework for a cohesive and comprehensive training programme, automatically tailored to the requirements of the individual student care-giver, and capable of configuration to the requirements of the institution at which the student or trainee is based, to allow conformance to local protocols of care while retaining the basis in clinical evidence.
For the embodiment concerned with training in use of a respirator for a pre-term infant (e.g. the Infant Flow™ family of products) , the manikin is a reasonable simulation of a new-born infant of between 26 and 40 weeks gestation. The manikin incorporates sensors for monitoring how the manikin is being manipulated during training. The manikin sensors may include:
Linear acceleration sensors:
Sensors capable of determining linear acceleration forces on the head in any direction: the range and level of sensitivity appropriate to normal and abnormal clinical handling.
Torsion sensors:
Sensors capable of determining rotational accelerative forces applied to the head of the manikin around any axis: the range and level of sensitivity appropriate to normal and abnormal clinical handling to be determined;
Load sensors:
The linear force applied at specific points of attachment of the CPAP circuit will be transduced at sites which are of clinically relevant to the handling of the infant and the application of the equipment. These sensors will be calibrated so as to allow
excessive forces to be both documented and , later, avoided as training advances.
The manikin also includes the electronic circuitry required to integrate and encode the information from the various sensors into a form suitable for transmission to the processors.
The manikin and patient sensor array is designed to affix to, or be a modification to, the various patient attachments that are provided as part of the Infant Flow™ system. The sensor array can be used on the manikin skull and face and on patients themselves.
The manikin and patient sensor array may incorporate elements that will transduce load applied to the nose, tension in the fixing straps, and acceleration forces to the head.
Load applied to the nose may be monitored by providing a sensor capable of determining the linear force applied to the nose by the nasal prongs and/or mask which hold the respirator on the patient used to create a loose seal in or around the nostrils and to deliver the CPAP or continuous positive airway pressure .
Tension in the fixing straps may be monitored by providing tension transducers capable of determining belt tension in the webbing fixing straps, and providing an electrical indication to the integration electronics. A visual indication (e.g. a light-emitting diode) will be provided of excessive tension in the fixing straps.
Acceleration forces to the head may be monitored by providing acceleration transducers fixed into a bonnet for use on
16
SUBSTITUTESI-
manikins and/or patients in order to help staff to avoid excessive shaking or other movements of the skull and brain.
The software components of the system may run on a Windows PC, or optionally a hand-held PC, and capture information from the manikin and patient sensor array.
The software integrates with training resources to effect the following functions:
(1) Identification of trainee;
(2) Identification of clinical skill to be attempted;
(3) Instructions for application of patient attachment;
(4) Illustration of correct application; (5) Illustration of common errors of application;
(6) Feedback from guided application to manikin to assess proficiency;
(7) Internal comparison with data representing other attempts, and other trainees, to provide proficiency score;
(8) Controlled repetition until sufficiently proficient to attempt application to patient;
(9) Guided application to patient to incorporate feedback and assessment; (10) Continued monitoring of patient to assess long-term stability of application;
(11) Integration with theoretical learning and examination; and
(12) Issuing of certificates to denote achieved learning targets .
The server will provide a central repository for the resources used by the software. This includes:
(1) Storage of illustrative materials including photographs, video sequences, and figures;
(2) Storage of references to literature and other external resources such as 3r<^ party web resources; (3) Storage of a secure database of registered trainees and trainers; (4) Storage of anonymised data representing previous application of the system, both to the manikin and in clinical practice; and (5) Storage of training scripts, questionnaires, programmes and similar resources used by the software.
Training resources may be provided to cover the application of the equipment (e.g. respirators such as EME' s Infant Flow™ product line) . These may include, instructions for use, illustrations, photographs, video clips, worksheets and questionnaires designed for on-screen use, and representations of correct and incorrect application of the equipment.
An example of how the system of figures 1 to 3 may be used to train an individual carer will now be described with reference to figures 4 to 16.
Figures 4 to 16 are flow charts illustrating a program implemented in a processor or a computer according to an embodiment of the present invention. In particular, a computer program corresponding to the flow charts of one or more of figures 3 to 16 may be implemented by the computer or processor shown in figures 1 to 3. Further additional embodiments of the present invention described or referred to elsewhere in this document may be implemented which are not particularly illustrated in the computer flow charts of figures 4 to 16.
18
SUBSTITUTESI-
Figure 4 is a flow chart summarising the routines which together form a preferred program. Figures 5 to 16 are flow charts illustrating details of (and the sub-routines which make up) each of the routines of figure 4.
The programs of the flow chart of figure 4 start at step 410. In step 410, the prospective user is identified before the session is started in step 420. When a session is started, the user may either start an entirely new session or continue on a previous uncompleted session where they left off (or at some other selected stage in the flow chart) .
The system will be described for the situation where the user starts an entirely new session which is worked through to the conclusion of the program. In practice, it is unlikely that a user will complete the program at a single attempt. It should therefore be borne in mind that in practice a user may terminate or pause, and re-start sessions at the various points indicated in figure 4.
Once step 420 is completed and a session is started, the program proceeds to step 430 where educational materials are displayed.
Once the user has successfully completed the education routine and passed the necessary test he or she then proceeds to the manikin set-up routine of step 440 in which the manikin, sensor array and medical device (e.g. continuous positive airway pressure respirator) are prepared and calibrated. Once the manikin, sensor array and medical device have been properly set-up, the program proceeds to the training manikin step 450 in which a training exercise is carried out on the manikin and the outputs of the various sensors monitored to determine the competence of the trainee.
When the trainee is considered to have satisfactorily completed the "on manikin training" routine, he or she is then invited to proceed to complete his or her training on a patient. Prior to the actual use of the medical device and sensor array on the patient, the medical device and sensor array are checked and calibrated in step 460 and a baseline or at rest recording of relevant patient clinical signs is made in step 470. The baseline recording is made so as to have a means of monitoring the patient during the training procedure. If during training on the patient, the measured clinical signs depart sufficiently from the baseline or at rest recording an alarm may sound to warn the trainee and his or her trainer or supervisor that remedial action may be necessary.
The trainer or supervisor of a trainee may alter the manikin' s response characteristics so as to better test a trainee. Selection by the trainer of different apparent responses by the manikin to simulate different clinical responses by patients with known conditions or complications of therapy e.g. increased fragility of skin and nasal structures in very premature babies of less than 26 weeks gestation shortly after delivery. Random selection of a number of different responses may be selected during the course of the training session after the trainer has told the trainee to be especially careful in performing e.g. attachment of the hat to the manikin's head, e.g. by changing the thresholds for ideal tension achieved in the fixing straps in step 1330 - see figure 13.
During step 480 (the "on patient training" routine) the sensor array and medical device used for the "on-manikin" training are re-used (after any necessary re-calibration) to monitor use of the device and its effect on the patient. The monitored
or sensed data is compared to defined acceptable standards and if these are met, the trainee's session is considered completed and an appropriate certificate issued in step 490.
The various routines of figure 4 will now be described in more detail in connection with the flow diagrams of figures 5 to 16.
Figure 5 illustrates the user identification routine of step 410.
The programs of the flow chart of figure 4 start at step 510. In step 510, the user is interrogated as to whether or not it is their first session. If it is their first session, the program proceeds to step 520 in which the user is interrogated by the system so that the system may determine the user' s identification and their role in the provision of medical care. This information is used to create a new history record in step 530 so that the system may track and monitor the identified user. If it is not the user's first session, they are asked to identify themselves and when they have done so, the system loads their training history (the recorded history of what they did in. their previous sessions on the system)
(step 540) .
If the user is working on their first session, after creating the new history record in step 530, they proceed to step A of figure 7. If the user is re-starting a previous session and this is not their first session, the system restarts by displaying their training history in step 610 (see figure 6) and then asking them whether they wish to continue where they left off in step 620. If they do not wish to continue where they left off, they are invited in step 630 to select the step at which they wish to restart. If they do wish to continue
where they left off, the program takes them to this point which will have been remembered by the system (see figure 6) .
If the user indicated in step 610 that it was their first session or in step 630 that they wish to start at step A, the system starts by displaying the stored training objectives in step 710 (see figure 7) . Once the training objectives have been noted by the trainee, the system displays instructional materials in step 720 for the trainee to digest and learn. Once the trainee has indicated to the system that they have reviewed and learnt the displayed instructional materials, the system questions or tests the trainee on their knowledge of the displayed instructional materials. This is done by requesting that the trainee complete a questionnaire in step 730 and then calculating a score in step 740. If the score passes a predefined sufficiency or pass mark, the system or program proceeds to step B whereas if the test score is below the pass mark, the educational materials are repeated and the program returns to step 710 for the trainee to redo steps 710 to 740 until an acceptable score is achieved.
Once an acceptable score has been achieved and the program has proceeded to step B (see figures 7 and 8) , the system then takes the trainee through the procedures necessary to set-up the medical device with its associated sensors, sensor array and manikin for an on-manikin training exercise. For the preferred embodiment of the invention concerned with training a nurse or other medical professional in the use of an infant respirator such as the Infant Flow™ continuous positive airway pressure respiratory aid systems, the system prompts the trainee to remove the sensor array from its packaging (step 810) , connect it to the interface (step 820) and then connect the driver to the interface (step 830) before then checking and calibrating the sensors (step 840) . Then in step
850, the trainee and/or trainer are prompted to test the sensor operation which is done by manipulation of the manikin, medical device and sensor array and cross-checking of the displayed sensed parameters. Once this check has been satisfactorily completed, the user informs the system that checks have been completed in a satisfactory manner in step 860 so that the system may proceed to the on-manikin testing step which starts with step C (see figures 8 and 9) .
For the preferred embodiment of the invention concerned with training and use of as infant flow respiratory aid, the practical tests on both the manikin and patient monitor both the time taken to complete the, for example, mounting of the respirator and the, for example, forces exerted on the manikin, patient, sensor array and/or respirator during the mounting process. This is necessary because during the mounting of the respirator, the infant will be without the oxygen supply and therefore prone to injury if this goes on for too long. On the other hand if the mounting process is rough or involves some shaking of the infant, other injury may be caused to the infant by the mounting process itself. It is therefore essential to try to install or mount the device as quickly as possible with the minimum force applied to the infant. It is the aim of the system to help train users or medical professionals to achieve this.
In step 910 (see figure 9) , the trainee and trainer are prompted to indicate whether they are ready to start a timed attempt. If either the trainee or trainer believe they are not, the system takes them back through the set-up and calibration steps 810 to 860 of figure 8 or alternatively back to the educational steps or sub-routines 710 to 740 of figure 7.
If the trainee and trainer are satisfied that a timed attempt on the manikin is now appropriate, the system proceeds to step 920 in which the trainee attempts to carry out the medical procedure (e.g. mount the infant respirator) and the system collects data from the various sensors and monitors the time taken until the trainee confirms in step 930 that they have completed the mounting.
Once the mounting is complete, the system then compares the date collected from the sensors to pre-defined bench marks and calculates a so-called learning score in step 940 (see figures 9, 12-16) . This learning score is displayed and compared with bench mark learning scores (step 950) . If the score is within a pre-defined tolerance of the target (e.g. 10%) (step 960), the trainee is invited to proceed to step D (see figures 9 and 10) . If the score is not satisfactory, the trainee is prompted to try another timed attempt and returned to step 910.
The system may be programmed to require the trainee to achieve a certain number of successful or satisfactory manikin mountings before he or she is allowed to proceed to the step D. It may, for example, be necessary to carry out the tests satisfactorily on, say, five consecutive occasions or on say, six out of seven occasions. The appropriate pass score will be defined by a supervising or programming medical professional to reflect medical needs and necessities.
Once the on-manikin training has been completed to an acceptable level, the trainee may proceed to attempt mounting the medical device on a real patient. The determination that the trainee can proceed to work on a patient may involve clearance by the system or program on its own using the bench mark scores as described above or, more probably, also require
that the trainer or supervisor separately confirm that the trainee's on-manikin training has been successfully completed In other words the system certification or validation may need to be complemented by the trainer' s certification or validation.
Once the on-manikin training has been satisfactorily completed, the system programs lead the trainee into the routines concerned with on-patient live training (see figure 10) . In steps 1010 to 1060, the medical device and sensor array are set-up and calibrated in a manner analogous to that described above for the manikin set-up sub-routines 810 to 860 of figure 8.
Before proceeding from the completion of the manikin training to the on-patient training, the patient is monitored both to check that they are fit for the procedure to be carried out and to establish a base line recording of appropriate clinical parameters for comparing against measurements of these clinical parameters during on-patient training. The base line recording routine illustrated in figure 11 can take place before, after or in parallel with the set-up calibration routine of figure 10.
In step 1110, the system prompts the trainee and/or his trainer to determine whether or not the patient is ready for undisturbed base line recording. If not, the trainee or trainer are prompted to wait until the patient is ready or to take whatever action is appropriate to ready the patient. If the patient is ready, data is collected in step 1120 over a defined period of time for use as the base line recording. For the preferred embodiments of the system which are used for training medical professionals in the use of the infant respirators, the base line data might include, some or all of
the following: heart rate as transduced either from electrocardiography or from digital pulse (and which provides evidence of physiological stress) ; oxygen saturation in the digital bloodstream as transduced e.g. using a pulse oxi eter (and which provides evidence of efficacy of ventilation) ; respiratory effort as transduced e.g. using a pressure sensor applied to the abdomen caudal to the diaphragm; and temperature of core and peripheral parts of the body (which are indicative of protective circulatory mechanisms that act when the cardiorespiratory system is compromised) .
The base line recording is then used or viewed by the system or trainer to determine whether the patient appears to be fit and able to withstand the training exercise (step 1130) . If not, the trainee and trainer are prompted to wait until the patient is ready or ready the patient. If the base line recording data indicates that the patient is fit for the practical task or training, the program proceeds to step F
(see figures 11 and 12) which is the initiation of the on- patient training sub-routines .
In step 1210, the trainee is prompted by the system to confirm that they are ready to start a timed attempt at using the medical device which in a preferred embodiment is an infant respirator such as an Infant Flow ™ respiration aid. If they are answer no, they will turn to step B where they are taken through the manikin exercise again. Alternatively, they can return to any other stage in the programs that is deemed appropriate. For example, if the reason that they are not ready to start a timed attempt has nothing to do with lack of confidence (for example it transpires the patient may not be ready) it may be more appropriate to take them through to step D or E.
If the trainee indicates that they are ready for a timed attempt, they proceed to step 1220 where a timer starts and they attempt to use the medical device on the patient. During this step, the information picked up by the sensors at the sensor array medical device is stored. Once the trainee has confirmed completing the exercise in step 1230, the system calculates a learning score in step 1240 (see also figures 13- 16) . This is then displayed in step 1250 where the learning score is then compared with bench mark scores indicating various levels of achievement and/or satisfaction.
If the achieved score is within a predefined tolerance of a target score (say 10%) (see step 1260), a certificate and certification is produced (step 1270) confirming that the medical professional has now been trained up to the appropriate level. The system may include a requirement that a certain number of satisfactory training exercises have to be completed before certification and it may also require that a parallel certification by the trainer is made at the same time so that a successful certified or validated candidate may need to complete a number of consecutive training exercises and to satisfy the trainer.
If the score achieved is not within the pre-defined tolerance of the target, the trainer is then prompted in step 1280 to assess whether a patient is fit for a repeated attempt by the trainee. If not, the trainer themselves will carry out the procedure before the system takes the trainee back to step B.
If the patient is deemed fit for a repeated attempt, the system takes them back to step F so that they may have another attempt at carrying out the training exercise on the active patient.
The routines of steps 940,950,1240 and 1250 described above with reference to figures 9 and 12 for determining whether a trainee has reached an acceptable or competent level will now be described in more detail with reference to figures 13 to 16.
Referring to figure 13, in step 1300 the system receives an indication that the trainee or carer has started the relevant procedure on a patient or a manikin. In step 1310 a timer starts. In step 1320 the system then retrieves and stores data representing the outputs from the different transducers or sensors (whether mechanical or physiological) . In step 1330 the system then calculates the mechanical parameters which are picked up by the sensors (e.g. the strap tension) and in step 1340 the system calculates and stores how far sensed physiological data deviates from a pre-defined base line. Step 1350, the system calculates and stores the deviation of the measured airway pressure from a baseline or desirable airway pressure. In step 1360, an observer (which may be either the trainer or the trainee) enters their assessment of the task or procedure carried out.
An assessment in step 1360 by an experienced clinical observer can often add much.
Not all tasks performed during patient care are amenable to measurement by sensors but may be observed by an experienced medical trainer. Embodiments of the invention may therefore provide an option for a trainer to input an assessment or scoring of aspects of a trainee's attempt.
Assessment by the trainee themselves may be of use in creating a self-critical and realistic approach to evaluation of one's own performance .
When the system is informed that the task is completed (step 1370) , the timer is stopped and in step 1380 the measurement of time taken to perform the task is stored together with the data derived from steps 1320, 1330, 1340, 1350 and 1360.
The routine shown in figure 13 is concerned with the collection of the data produced by the system. Figure 14 illustrates the routines representing performances and outputs from the various sensors during patient recovery from the procedure. When monitoring the use of medical equipment it is often important to monitor how a patient recovers from a procedure as well as how the patient actually responds during the procedure itself.
The routine of figure 14 monitors how long it takes for the sensed physiological data to return the baseline levels pre- procedure determined in the routine shown in figure 11. The sensor outputs are monitored during and after the procedure.
When an indication is given in step 1410 that a task has been completed, a recovery timer is started in step 1420 and the physiological data collected (step 1430) by the various sensors is retrieved and stored. This is compared (either continuously or at pre-determined intervals) with the baseline physiological data in step 1440. When the sensed physiological data returns to the baseline values or within an allowed deviation of these, the recovery timer is stopped and the recovery time stored.
Figure 15 illustrates a routine for determining whether the measured transducer or sensor outputs correspond to an acceptable or unacceptable procedure. This routine is used to calculate the "learning score" of steps 940 and 1240 of
figures 9 and 12 respectively. The routine of figure 15 is applied to each of the sensor or transducer output data sets.
The system includes a pre-determined or pre-calculated target parameter Z. For a system used to train medical practitioners in the use of an infant respirator, the target parameters might be those set out below in table 1.
The target level Z will, in a preferred embodiment, be set to an initial- default value defined by the medical practitioner but could be modified over time to represent data from a preselected proportion of the best attempts recorded by the system. For example, the target Z may represent the range of the best 25% of competent attempts recorded by the system.
Referring to steps 1510 and 1520 of figure 15, if the measured and calculated sensor output or equivalent mechanical physiological measurement corresponds to the target parameter Z, a score component is set to 3 in step 1520.
If the measured value is not within the parameter target, the system then checks in step 1530 as to whether the deviation is within a first threshold range X. This might be 50% of competent attempts recorded respectively. If the deviation is within this range, the score component is set to 2 in step 1540. If it is not, the system then proceeds to step 1550 which determines whether the deviation is within a larger threshold value Y which might be, for example, 75% of competent attempts. If the measured values are within this larger threshold value Y, the score component is set to 1 in step 1560. If the deviation is not within this larger threshold value, the score component is set to 0 in step 1570. Once the score component has been set, the system proceeds to the next step 1580. The system may of course include more or
less ranges or different threshold values. This scoring system would of course then be amended in a corresponding manner .
Not all tasks performed during patient care are amenable to measurement by sensors but may be observed by an experienced clinical trainer. The trainer may incorporate the completion of specific tasks in the learning score calculation in order to conform with their own institutions guidelines. e.g. correct selection of different sizes of attachments for varying sizes of babies, documentation in nursing records, .explanation to babies' parents.
Figure 16 is a summary figure which illustrates how the system collects and manipulates information relating to four different important sets of data derived from someone being trained or certified on a medical device, for example a respirator such as the Infant Flow™ device. Timing, force or mechanical, physiological and airway pressure data is derived by the system in steps 1610, 1620, 1630 and 1640 respectively. This data is fed into routines or modules which calculate the score components in the manner shown in figure 15 for both the parameters derived during the task and those derived during recovery.
These different score components are then weighted to produce a weighted sum which forms the learning score produced by the system and which is the output measuring the competence, level of competence or lack of competence of the trainer or user of the system.
Figures 13 to 16 illustrate the program steps for calculating a score related to the ease and proficiency with which a trainee completes a set of tasks as part of the training
programme. The score is a weighted sum of the differences between measured values and benchmark values .
The weighted sum may be of components related to duration (of whole and/or part of procedure) , variance and maxima (of tension, pressure, head bounce) . All of these components, when the device is applied to the real patient, may adversely affect the patient during application of the respirator interface to the airway. In this case the task consists of applying the respirator interface, hat, etc to the patient's face and skull without causing physical damage or physiological deterioration during the handling episode.
The weighting starts by showing best scores for most experienced users, but then varies according to running average effects of physiological disturbance of babies during training iterations on the patient.
As described above and shown in figures 9 and 12 each component of the weighted output may be assessed on both dummy and patients. The weighting will be set according to clinical judgements of what is a competent level or variation from the ideal . -
For example, each component may have the following values assigned: Value "1": not achieved/completed
Value "2" achieved within 30% ideal value Value "3" achieved within 10% ideal value
Each of the ten components of the score is then multiplied by the weighting factor after iterations involving patients in their groups according to gestation age, postnatal age, birthweight, illness category.
32
Table 1
An "ideal trainer" would have a maximum score of 128 (50 from the last two components, which should be compulsorily passed at 80% before the trainee is allowed near the first patient i.e. the trainer has the last say). The optimal score would change following feedback adjustments which would change the values of the weighting factor.
33
SUBSTITUTE Sr £ 26
It might be that a score of 75% of this (96) would allow progression to the first patient.
A preferred embodiment of the invention has been described with particular reference to a respirator for a fragile infant such as a pre-term or prematurely born infant. It will readily be appreciated that the invention can be embodied in training systems for any medical device and i.e. particularly useful for training on medical devices which are more easily misused or if misused have potentially serious adverse consequences .
The preferred embodiment has been described with a large number of optional features. Embodiments of the invention need not include all the features and this disclosure therefore explicitly encompasses systems which include only some of the preferred routines described with reference to the various figures.