Yan Kestens - Managing large cohorts and collecting data on mobility and health behaviour
Managing large cohorts andcollecting data on mobility andhealth behaviour :Novel solutions and challengesYan KestensMontreal University, Social and Preventive MedicineMontreal Hospital University Research Center (CRCHUM)SPHERE Lab .orgParis, France21th May 2013
AimPresentation of tools/methods that facilitate thecollection of data in large cohorts (with a focus onspatial data)Three tools that have been pilot tested or implementedin existing cohort studies– Tool 1: Study Management Application– Tool 2: VERITAS interactive mapping questionnaire– Tool 3: Multisensor platform for real-time tracking
Managing a cohortLarge cohorts All kinds of challenges!RecruitmentParticipantsInterviewersQuestionnairesDevicesDatabaseAttritionResidentialmovesConfidentialityCoordinatorsParticipationrateBiasInvestigatorsData collectionTrackingData linkageGIS
Managing a cohortRecruitmentParticipantsInterviewersQuestionnairesDatabaseCoordinatorsInvestigatorsDatacollectionTrackingData linkage GISSamplingMeasurements HardwareAdministrativedatasourcesRecruitersFollow-upModelling
AttritionResidentialmoveConfidentialityParticipationrateBias Loss-to-follow-upComplianceManaging a cohortTransversal Challenges!
Tool 1: Cohort ManagementApplicationUse of a study management application to manage– People– Procedures– Questionnaires– Devices– Procedures
Study Management Application• A comprehensive application to manage cohorts• Facilitates the process• Keeps track of activities• Integrates questionnaires
Tool 2: Spatial data collection tool• Environmental determinants increasingly at stake,both as a cause of disease, social health inequalities,and as a target for intervention• Current shift to improve integration of daily mobilityand multiple exposures in epidemiological models
Collecting spatial information• Tools to collect location information include:– Residential history questionnaires – lifecourse– Travel surveys – often one day of detailed mobility– Activity space questionnaires – asking people’sregular destinations– Real-time tracking using GPS receivers
Spatial data used in health researchPlace ofresidence /Residentialhistory
Spatial data used in health researchRegular activityplaces
Spatial data used in health researchRoutes / GPStracks
Spatial data used in health researchPerceivedspaces
Spatial data used in health researchAttribute data:FrequencyAttachmentPerceptionTravel modeConstraintsWith whomEtc.
VERITAS, an online mappingquestionnaire• Uses an interactive map to collect spatial data• Can be administered or self-administered• Flexible and scalable• Allows to collect information on locations, routes,spaces and related qualitative assessment• Is linked to mapping and search APIs to facilitate theprocess and increase validity (Openstreetmap,Google Map, etc.)
VERITASA series of questions which can be answered on a mapthrough creation of:– A point location (marker)– A line (polyline)– An area (polygon)Map searching capabilities / streetviewfunctionalities can help identifying knownlocations/destinations.
VERITASExample: Where do you shop for food most often?
VERITAS RECORD• Illustration: VERITAS in the RECORD Study• RECORD Study: Large Paris area cohort on Cardiovascularhealth (n = 8,000)• Wave 1 in 2007-2008, Wave 2 in 2012-2014• VERITAS RECORD administered to some 4,800 participants asof today• 27 spatial questions including destinations for food shopping,sport activities, leisure, friends, family, etc.• Over 65,000 locations collected – Median of 14 locationscollected per participant• Median completion time of 20 minutes
VERITAS RECORD• Rich spatial information on regular destinations which canserve to identify multiple environmental exposures andinequalities• Spatial information transformed into spatial indicators to feedepidemiological models (Activity space size, maximumdistance, concentration etc.) (Camille Perchoux, Ph.D.candidate)• Interesting information to monitor spatial health inequalities,mobility behaviour and guide intervention in the distributionof resources/infrastructures
Tool 3: A multisensor platform for real-time tracking• Self-reported locations vs. objective measures• Multisensor device for tracking of:– Mobility (GPS, RFID)– Physical activity (Accelerometer)– Physiology (Various sensors)
SenseDoc Multisensor DeviceCAPTUREAccelerometerMarie-Lyse Bélanger, M.Sc. Student in kinesiologyAccelerometer validation using indirect calorimetryLab – 14 controlled exercises from sedentary to vigouros PAEleven adult subjectsCalculation of Vertical Magnitude Acceleration (VMAG)Testing of various bandpass filtersComparison with Actigraph GT3X performenceBest results obtained with Bandpass filter 0.1 Hz – 3.5 HzModelling of Energy Expenditure: Adj. R-square of .79Use of Vector Body Dynamic Acceleration (VEDBA)
SenseDoc Multisensor DeviceCAPTUREAcquisition serverBattery life
SenseDoc Multisensor DeviceCAPTUREData transmissionGPS Data sent over the air (cellphone network) every 30 minutesPossible alerts depending on- Location- Activity- TimeConnection to other sensors (2.4 GHz ANT+) Heart rate monitor,footpod, RFID tags, etc.
Issues in data processingPROCESSINGTransforming raw GPS data into meaningful and usefulinformation, combining with accelerometry- ‘Putting things into context’- Activity locations- Trips between locationsSPHERELAB GPSARCTOOLBOXwww.spherelab.org/tools
UsageUsing GPS/Accel to locate behaviour and assess exposureImprove the understanding of mechanisms linkingenvironments to health behaviours and profilesUse GPS to prompt recall and gain additional insightUse GPS to support qualitative studies (go-along, geo-ethnography, geo-tagged photos, environmentalperception, etc.)Use GPS/Accel data to assist clinical practice (mHealth)USAGE
UsageUSAGERECORD GPS Study, Paris191 participants wearing GPS & Accelerometer for 7 daysEstimates of:• Number of steps walked• Energy expenditure• Moderate to Vigorous physical activity• Sedentary timeAnalyses possible at the trip level and by travel mode
UsageUSAGEDuring wear time, transportation was responsible for:• 39% of steps walked• 32% of total energy expenditure• 33% of MVPA• 15% of sedentary time
UsageUSAGEGeographic variations in contribution of transport to physical activity
UsageUSAGEDifferences in PA compared to car driving, per 10 min of trip(n=4,984 trips with unique mode)
Spherelab GPS studiesRECORD-GPS Study, Paris, Basile ChaixBIXI bikesharing study, Montreal, Lise GauvinSte-Justine CIRCUIT Pediatric Intervention, Montréal, MélanieHendersonNovel Real-Time Measurement of Physical Activity Patterns in Type 2Diabetes and Hypertension through GPS Monitoring and Accelerometry,Kaberi DasguptaHealthy Aging in Urban Environments, Montreal, Paris, Luxembourg; YanKestens, Basile Chaix, Philippe Gerber
CURHA Project Develop an international platform and researchagenda to collect and analyse detailed data on dailymobility and health outcomes among older adultsliving in contrasted urban settings Use of novel methods to capture daily mobility tobetter understand interactions betweenenvironments, mobility and health
Objectives Provide evidence about how characteristics of urbanenvironments relate to active mobility and socialparticipation Disentangle the complex people-environmentinteractions that link urban local contexts healthyaging
Methods Trois cohortes, 450 participants par site: Montréal/Sherbrooke: Cohorte NuAge Paris: Cohorte RECORD Luxembourg: Nouvelle cohorte avec SHARE
MethodsVERITAS(Questionnaire onregular destinations)VERITAS(Questionnaire onregular destinations)Canada LuxembourgFranceExistingquestionnaires (ex:individual SES)Existingquestionnaires (ex:individual SES)Existingquestionnaires (ex:individual SES)Novel GPS/Accelerometry mobility protocolNovel qualitative assessment of place experienceExisting GIS Existing GIS Existing GISNOVEL PROCEDURES TO BESHARED AND APPLIED TOALL SETTINGS, DRAWINGON EXISTING EXPERTISE INDIFFERENT SETTINGSEXAMPLES OF EXISTINGCOMMON RESSOURCES INALL SETTINGS (Need forcross-validation of DB toensure comparability)VERITASQUESTIONNAIRE(Activity spaces)EXAMPLES OFEXPERTISE/TOOLS EXISTINGIN ONE SETTING TO BEEXTENDED TO OTHERSETTINGSMULTISENSORPLATFORMMULTISENSORPLATFORMMULTISENSOR DEVICEAND SERVERPLATFORMQUALITATIVEASSESSMENT OFMOBILITYQUALITATIVEASSESSMENT OFMOBILITYQUALITATIVEASSESSMENT OFMOBILITYTOOLS/PROCEDURES SHARING CONFIGURATIONSNovel spatio-temporal modelling
Data Mobility assessment will resort to:• Activity space questionnaires• Continuous 7-days monitoring of location• Physical activity using wearable sensors• Qualitative assessment of participants’ experiencesand meanings of his/her activity space, mobility, andhome territories.
Data Behavioural outcomes of focus:• Active living (including active transportation, walkingand sedentary behaviour)• Social participation• Spatial behaviour (activity space, modes oftransportation, relation to places) GIS for environmental exposure measures
Analyses Liens entre contextes urbains (SIG), mobilité, activitéphysique, participation sociale
Conclusion“Design for the young, and youexclude the old;design for the old and youinclude everyone”Bernard Isaacs, in G. Miller, G. Harris and I. Ferguson,“Mobility Under Attack”.