ADHD Behavioural Clinical Data

The ADHD Behavioural Clinical Data project originated from the University of Dundee's Data Visualisation Crucible on 13 March 2014, and involves:

  • Dave Coghill and Shona Matthew (School of Medicine)
  • Gair Dunlop, Christopher Lim and (previously) Lilia Gomez Flores (Duncan of Jordanstone College of Art & Design)
  • Alistair Geddes (Geography in the School of the Environment)
  • Iain Murray (Computing, School of Science and Engineering)
  • Mark McGreehin and Deborah Chapman (students at Duncan of Jordanstone College of Art & Design)
  • Dimitrios Zandes (student at Computing, School of Science and Engineering)

Attention Deficit Hyperactivity Disorder (ADHD) is a complex developmental psychiatric disorder which affects around 5% of school-age children. While ADHD can be identified and diagnosed in early years, the associations between causes and outward manifestations are complex. Data on these associations are patchy and difficult to analyse and interpret, even following diagnosis. Messages from the data are consequently also difficult to communicate in effective and appropriate ways to ADHD sufferers and their families in relation to psychological and/or pharmacological therapies.

This project aims to tackle improvements in the collection, quality and visualisation of behavioural clinical data on ADHD. In turn, these changes will be used to enable improvements in self and clinical management of ADHD treatment and to improve public awareness and understanding of ADHD.

Work on the above changes is focussed on four strands, identified by the above interdisciplinary group including clinical specialists, experts in social digital design, digital film, computing and geographic information science. The four research strands are as follows:

  • Prototyping improved forms of visualisation of existing clinical behavioural data. This strand is based on the proven role for non-text graphic forms of visualisation in assisting with communication and understandings of large and/or complex data sets. Attention focusses specifically on developing visualisations which are patient/carer-centred, which are geared to assisting self/carer management of behaviour variations linked to current ADHD therapy regimes. Such visualisations must connect to and enliven the key messages from the data from the perspective of patients and their carers, and in addition they must be capable of presenting multidimensional information in concise ways. Work on this will combine expertise within the group in the clinical interpretation of existing data with design and cartographic expertise, in the latter case in the manipulation of visual variables for the display of multivariate data.
  • Prototyping an interactive film platform. Film is another powerful visual medium which in this instance can assist in exploring and understanding the lived experiences of ADHD behaviours. Conventional film viewing positions the viewer in a largely passive role, however in this case the intention is to build in tools supporting more active interactions in non-linear ways. Analyses of individual-level clinical data can be used to outline initial storyline pathways for this prototype, although any data/analysis drawn on will be anonymised. For similar reasons it is proposed that key actors are represented using a combination of puppetry with voiceovers which can retain the essence of personal experiences while ensuring ethical engagement of film viewers. As part of their final year degree project How It Felt, Mark and Deborah have developed some puppets and filmed them for ADHD awareness videos.
  • Spatial analysis of aggregate pharmacoepidemiological data. Among the best existing data on ADHD available for this work are the pre-existing Tayside Health Board records of pharmaceutical prescribing and on the collection of those prescriptions by/for the patients. These data are inherently geographical i.e. containing the locations of prescription and the pick-up pharmacy, and can be linked to the patients’ own home location. Currently, however, the spatial references in these data are under-utilised, and indeed some data exist only in paper rather than digital format at present. Digital conversion and linkage of records based on common geography is a core part of this envisaged strand of work. Individual patient records will be aggregated for appropriate spatial units – e.g. primary care trusts – in order to support an exploratory analysis of variations in the quantity and quality of prescribing as well as in actual uptake of prescriptions. In turn, this analysis has potential to assist assessment of cost effectiveness of existing treatments – e.g. if pharmaceutical spend is high, but is well targeted and with good evidence of uptake.
  • Initial specification for a wearable sensor. This is the most ambitious strand of work seeking no less than a step-change in spatial and temporal collection of individual patient data. It stems from growing attention more generally to the design and implementation of ‘biomapping sensors’ linking sensors of body states with time and location. Technology improvement, miniaturisation and portability, including Global Positioning Systems (GPS) receivers factors are all critical factors in such development. One example project to date involves exploring variations in individuals’ levels of stress in different urban environments (see http://www.youtube.com/watch?v=hHnh4FfQ2C8). In this case the focus is on developing a specification for a wearable sensor for measuring inner tension in child ADHD patients (inner tension is not necessarily visible from appearance). The sensor must meet a number of criteria: it must be accepted by the wearer, it must measure inner tension in a sufficiently accurate and precise way, it must simultaneously measure time and space (derived from GPS), must have sufficient storage capacity, be robust, unobtrusive and comfortable, and allow data to be downloaded easily. Addressing these challenges will require considerable planning, but they are not insurmountable. The longer-term goal would be to move from prototype to a working sensor which could be trialled for relatively short periods of time with a sample of patients, allowing a sample of data to be collected, analysed, and used to enhance existing visualisations as well as developing new forms of visualisation which can contribute to better self-regulation and quality of life.

Can't find what you're looking for? Contact Us