We are unreservedly patient-centric in our approach … However, for that to work we also have to listen to Neurologists
I recently drafted a white paper to try and start a discussion …
This is what it says …
Neurologists need data
The NeuroPath Portal platform seeks to bring together relevant data on PwPs, to enable the Neurologist and other members of the care team to optimise the care path by …
- tracking PwP progression through a variety of biomarkers and derived metrics
- timing consultations more accurately
- adjusting medication based on patient reported outcome measures PROMs
- predicting events and planning timing of procedures like DBS
The overall objective is to improve Quality of Life and reduce the burden on the care team by means of home-based monitoring, To date the NeuroPath team has collected feedback from individual Neurologists and members of the care team.
The aim now is to bring together a number of different Neurologists into a forum to discuss possible new approaches based on data collected, and the most effective way to visualize this data and derive actionable insights. The way we aim to approach this, is to work backwards from postulated approaches and outcomes for individual scenarios to the data needed to achieve them.
Data comes in two varieties structured and unstructured. Structured data is familiar from patient records, medical history, prescriptions, test results, biometrics.
The data that has to date been harder to analyse in a consistent manner is ‘unstructured data’ – clinical observations of motor symptoms, facial expressions, voice and video recordings, unstructured notes etc.
Increasingly, Machine Learning approaches are making it possible to complement the clinical observations of an experienced Neurologist with biomarkers which are for the first time promising to make PD ‘measurable’.
However, data needs to be processed to turn it into information and information needs to be presented in a manner that makes it helpful to the professional.
A possible approach
Here is an example of a possible approach.
If we were able to obtain reliable pharmacokinetics for dopamine-based medication, we could combine the following data …
- On-Off state over time as observed by PwP
- PwP-reported symptom and side effect severity and prioritization (e.g. tolerable dyskinesia)
- Dopamine demand curve, based on PwP activities including exercise, diet, professional (e.g. speech)
- Dopamine supply curve based on monitoring of medication intake and changes in prescription
- DBS settings
- QoL analysis of a range of neuropathic pain experiences, based on a questionnaire
We could imagine graphical modification of an interactive dopamine demand curve combined with what-if exploration of the expected impact of a proposed change in medication.
The dopamine supply curve would look like this and could be supplied systematically by drug manufacturers.
Add to that some basic information about diet, biometric data from an exercise bike and motion video recordings of physiotherapy sensing progress with stability issues, with a biomarker analysis of voice, measuring progressive deterioration and we are starting to see a highly individual set of parameters with the potential of predicting outcomes.
Complement this with a more detailed medical history showing gene mutations GBA and LRRK2, an appendectomy, instances of diverticulitis and potential professional exposure to pesticides and research findings may start to become relevant to planning for a bowel procedure …
This is just an example of what might be possible and how to visualise this data in a layered manner.
A NeuroPath PD Visualisation Forum for Neurologists
It does not have to be that complex to be helpful ..
Just having the QoL responses of a PwP shown alongside those of his/her carer, can highlight meaningful differences to discuss …
Just as this vowel triangle could be a relevant biomarker to assess the effectiveness of speech therapy …
Or this ‘stickman’ could help assess stability issues and maybe act as early warning for the risk of falls …