Emergency Monitoring and Prevention (EMERGE) Results EU Specific Targeted Research Project (STREP) 6. Call of IST-Programme 6. Framework Programme Funded by IST-2005-2.6.2 Project No. 045056 Dr. Thomas Kleinberger thomas.kleinberger@iese.fraunhofer.de Phone: +49 631 6800 2288
Motivation 44% of Emergency Medical Services (EMS) system resources are dedicated to patients older than 65 years of age (statistical report of the town of Kaiserslautern 2005) 24,5% of elderly people are living alone only 3% of the affected people in emergency situations have a PERS at hand From these 40% used PERS, PERS indicated 37% 60% did not use PERS, PERS indicated 67% In total, the PERS was used only in 1.3% of all cases to report the incident Automatic detection and alarming of emergencies and potentially arising emergencies would be a great benefit Slide 2
The goals of the project? Goal: Live longer in the preferred environment (at home) Early automatic detection of acute emergencies resp. potential announcing critical health status Analyze typical behavior based on vital data and activities Reason on functional health status Unobtrusive approach Unobtrusive and Ambient Sensors measure Activities Vital Data No cameras, micros, body-mounted sensors (just one bracelet on the wrist) Detect and assess situations Adapt individually to assisted person Base: Behavior parameters and medical parameters Adapt to individual environment and possible sensor set Process data in the home environment (no transmission of sensor data to outside) Detect critical situations acute situations (fall, helplessness) long-term behavior deviations Stepwise Assistance Proactive Integration of relatives, caregivers and opt. EMS Support persons do not replace Slide 3
Sensors Vital Data Electronic weight scale Electronic blood pressure device Pulse and skin temperature with wrist device (watch) Heart frequency and breathing frequency with bed mat Activities (Sensors in the environment) Light switches, blind switches (home automation) Position tracking of persons at areas of interest (movement sensors) Usage of furniture, doors, windows (contact sensors) Object usage (toilet, shower, faucet) Device usage (power measurement) Fall detection: Wrist device (watch) Not usable in practice turned out to be Continuous position tracking of persons (intelligent carpet, with tags) In general: Sensors in the environment, mountable afterwards Battery/batteryless operation Wireless data transmission Slide 4
Functional Health Status Funct. Health Status a perfect known restrictions b h t = f t (p 1, p 2,...., p n ) normal Dementia Depression c critical Fall Helplessness Time Slide 5
Situation Recognition Activities that are geriatrically relevant for determining functional health status Allow prediction of critical behavior deviations from perspective of emergency medicine Individuality of persons Behavior Environment Characteristics Ambient unobtrusive sensors No cameras No microphones No body mounted sensors Slide 6
Model: HCM HCM describes Acute critical situations Pulse Breathing rate Blood pressure Weight Motionlessness Fall Long-term deviations Vital data ADLs Sleep (day, night) Personal Hygiene Toilet Usage (day, night) Prep. Meals General Mobility Slide 7
Example An example ADL assessment Daily measure (orange) Personalized region of normality (red) Calculated dynamically Upper limit Lower limit Self adapting Assessment for various time periods Day -> Week Week -> Month Month -> Half year In total ca. 18 assessments Slide 8
What can be detected? Acute emergencies Helplessness (e.g. after fall) No activity (motionlessness) Critical vital parameters Manual alarm Long-term deviations Untypical behavior Changed behavior Changes movement pattern Vital data Vital data Pulse Blood pressure Weight Breathing frequency Movement Distance/speed Room changes ADLs Sleep Personal hygiene Preparation of meals Toilet usage Analysis Trends Deviation from norm Sensor data of Vital data sensors Environmental sensors Activity sensors (Position tracking) Situationen HCM Parameter Sensordaten Slide 9
Assistance, Applications For the assisted person For caregivers Slide 10
Test and Evaluations - Assisted Living Lab Realistic simulation of environment Integration of solution parts together with research and industry partners Test of individual technology (prototypes) Preliminary studies for field trials Evaluation of dependability, accuracy, usability, reaction time, security, acceptance, ethics Slide 11
Evaluations - Measurable Objectives View User Validation of solutions according to evaluation criteria from the quality model in different perspectives Professional Technology Main Aspects Accuracy of situation detection Accuracy of emergency situation detection Efficiency increase when handling emergencies Increase of social interaction Impact of assistance functions Usability Increase of Quality of Life Meet specifications Meet user requirements Dependability of system Evaluation tools: Statistical: Empirical Evaluation (e.g.. dependability, accuracy, ) Perception: Questionaires(e.g. user requirements, impact, advantage, ) Analysis: System analyis (e.g. meet specifications) Slide 12
Evaluation Process Overview Slide 13
Evaluations in different categories (1) Formative evaluations of all prototypical components during the research and development phase Summative evaluations of technical devices/sensors developed in EMERGE Wrist Device for vital data measurement Wrist Device for fall detection Sleep monitor for measuring vital data in bed Position tracking systems Ubisense, Cricket, BATSY Summative evaluations of SW components for situation detection Detection of Activities of Daily Living (ADL) Calculation of Activity Score Detection of Motionlessness Detection of short-term and long-term behaviour deviation with Trend Analysis and the Human Capability Model Slide 14
Evaluations in different categories (2) Summative evaluations in field trials, including acceptance, usability and ethical evaluations Short field trial in the O Nestor Home Care Center in Athens/Greece Long field trial in the Seniorenresidenz of WKK in Kaiserslautern/Germany Summative evaluations with the emergency medical service (EMS) Information and communication flow with the EMS in Kaiserslautern/Germany Usability Walkthroughs of Applications With experts Usability Evaluations of Applications With assisted persons With caregivers System Acceptance With assisted persons and focus groups Slide 15
Evaluation results: ADL detection Scenario-based evaluation in AAL-Lab 220 scenarios / 15 test subjects Detection of ADLs: Toilet Usage Personal Hygiene Preparation of Meals Slide 16
Field trial in Kaiserslautern, Germany Two single apartments 67-year old female, widowed, active, diabetes, depression, social activities 90-year old female, good mental condition, surgeries every 3 months, hearing constraints, few social activities Applications for assisted persons and caregivers Installation in September 2009 Second EMERGE Prototype Comprehensive ADL-Detection Applications for assisted persons and caregivers Operation: 4 Months Slide 17
Usability 3 kinds of Usability Evaluations Usability walkthrough with experts 3 usability engineers, 1 requirements engineer Usability evaluation of applications with assisted persons 3 test persons (2 female (age 70-80), 1 male (age 60-70)) partly living alone at home, partly married experienced with emergency situations at home (helplessness after fall) 2 have no experience with computers, 1 is familiar Usability evaluation of applications with caregivers 2 test persons (female, age 40-50) Tech. experience: Familiar with computers and internet Job experience: 20-23 years Slide 18
Usability Research Objectives Q1: Are there any usability problems while using the HOME UI prototype? Q1a: Have the improvement suggestions from the usability walkthrough with experts been successful? Q2: Does the EMERGE system address requirements (from the viewpoint of assisted persons) elicited within the scope of the requirements engineering phase? Q2a: Is there any functionality missing? Slide 19
Usability Tasks Task 1: Reminder Blood Pressure Measurement: Find next date Task 2: Reminder Blood Pressure Measurement: Measure immediately Task 3: Raise alarm manually Task 4: Leave apartment Task 5: Visitors are coming Task 6: Access Health Status Reports Task 7: Cancel alarm Slide 20
Usability Results (1) Goals Overall rating of the goal fulfillment by assisted persons was very good. Again, it was mentioned that communication possibility with recipient of an alarm in an emergency situation would be calming. Slide 21
Usability Results (2) Suitability for the task Attribute avoid lack of information / functionality reached partial agreement, as there were some additional functionalities proposed that could be included in the emerge@home prototype, such as reminder functionality for daily activities (such as appointments). Furthermore it was suggested to integrate a component that enables social contact with other elderly persons or a security function to alarm in case of burglary (to increase the feeling of security) Slide 22
Usability Results (3) Self-descriptiveness, Controllability and Suitability for Individualization Affected by the font size ( Could be a little larger ) and that some medical expressions as included in the health reports are not understood by the test persons. Navigation support has been rated somewhere between partial and barely agreement as navigation was difficult to judge by the test persons. Slide 23
Usability Results (4) Suitability for Learning and Acceptance Suitability of learning of the system has been rated very well. The test persons also indicated that they would like to use the system. When answering the question whether the interface would fit into their living environment the test persons they were in doubt whether they would need such a device in every room. Slide 24
Privacy and Ethical Issues Ethical Advisory Board (together with Project Management) Intention: Handle ethical issues within the project (better than ethical review from outside) Advise the project board on every ethical issue arising in the project Provide list of potential ethical issues in the project, concerning end-users and involvement of physicians privacy, data protection, sensitive data about health Suggest procedures to handle ethical issues and avoid ethical implications Allow end-users to participate or resign Prohibit misuse or publication of sensitive private data Detailed information of end-users with informed consent Submit Data Security and Ethics Application for all trials involving end users (at data security officer and medical council of the local state, where field trials takes place) Ethics Handbook in accordance with the opinion of the European group on ethics in science and new technologies to the European Commission (especially No 17, 4th February 2003) and the statement by the Ethics Committee of the American Psychological Association Organize ethical review, if required Evaluate results of handling ethical procedures (checklist) Requirements phase: Privacy and anonymity of data collected in questionnaires Evaluation phase: Privacy and anonymity of data collected during evaluation (data security approval) Slide 25
Acceptance Quantitative acceptance measurement in focus-groups using TEAS questionnaire (Test for the Evaluation of the Acceptance of ambient assistive Systems) Categories Performance Expectancy: feeling of security caused by the system and belief in functioning Effort Expectancy: Amount of effort and costs that the system caused Self-Efficacy: Ability to use, control and understand interaction and user interfaces Design Aspects: Optical acceptability of sensors and installations Social Influence: Relevance for other people Facilitating Conditions: Getting used to the technology, getting along with new situations Anxiety of Social Isolation: Anxiety of being left alone because of the emergency monitoring system Anxiety regarding Security Aspects: Anxiety of burglary, foreigners in the own apartment, misuse of data Anxiety regarding Privacy Aspects: Fear of system damaging privacy, observation, intimate or personal information seen by other people Intention: Intention to use such a kind of emergency monitoring system Slide 26
Acceptance Results (1) Person involved in Field Trial: Mrs. B (female) Rating: Likert-Scale: 1 (does not match at all) 5 (does match completely) Slide 27
Acceptance Results (2) Person involved in Field Trial: Mr. K (male) Rating: Likert-Scale: 1 (does not match at all) 5 (does match completely) Slide 28
Acceptance Results (3) Focus group 1: 7 participants, >60 years (62-83), living alone at home Focus group 2: 4 participants, >60 years (63-74), living alone at home Focus group 3: 10 residents of Seniorresidence, >60 years (except 1), living in home care center Rating: Likert-Scale: 1 (does not match at all) 5 (does match completely) Slide 29
Acceptance: Main Findings Feeling safe and secure is important for all potential end users. The sensitization of end users who experienced a situation of helplessness already or know cases in their social context is higher than the sensitization of people that do not imagine they could be helpless some day. The level of sensitization does not depend on the age of end user. Potential end users, who highly rely on social care and protection, are not so interested in a technological system like EMERGE. Their main goal is to get help fast and this help should be given by people they do trust in. The system idea to detect critical situations and to provide assistance is highly accepted, if the system works properly, does not cost too much, is unobtrusive and easy to maintain. End users are aware of ethical issues like potential privacy and security problems. If the system supports a self-determined life, these potential risks are judged as acceptable. None of the interviewed potential end users fears to get isolated by using the EMERGE system. Slide 30
Thank you for your attention Contact Dr. Thomas Kleinberger thomas.kleinberger@iese.fraunhofer.de +49 631 6800 2288 Slide 31