26 June, 2019

Disclosure

All authors declare no financial disclosure to make

Background and aims

Digital health interventions(DHI) can be used to promote

  • Smoking cessation
  • Healthy eating
  • Physical activity
  • Safer sex
  • Alchohol Consumption
  • Chronic disease outcomes
  • Improve accessability to treatments

Background and aims cont’d

Nature of DHI

  • Complex, multi-component, multiple aims
  • Components may include information, psychoeducation, decission support, behavioural change support, tools of self-management, etc

Limitations of DHI

  • Todate, the potential of DHI has not been fully realized. Example: imagine an iphone activity tracker app released today
  • Scope creeps/kitchen-sink syndromes vs RCT vs Publication of trials
  • Within this limitations, decission makers should make decissions
  • Researchers want to support the decission making by creating an actionable knowledge base

Aims

  1. Explore issues that arise in developing and accumulating DHI knowledge base
  2. Explore how this knowledge can be generated in atimely manner efficiently

Key Research questions in DHI

Key Research questions in DHI cont’d

Defining the problem

  1. Is there a clear health need which this DHI is intended to address?

  2. Is there a defined population who could benefit from this DHI?

Key Research questions in DHI cont’d

Defining the likely benefit of the DHI

  1. Is the DHI likely to reach this population, and if so, is the population likely to use it? 
 -effectiveness: controlled vs real world
 -principles that specify methods of buisness feasibility 
   . acceptability & usability
   . demand
   . implementation
   . practicability
   . adaptation
   . integration

Key Research questions in DHI cont’d

  1. Is there a credible causal explanation for the DHI to achieve the desired impact?
  • Example: DHI resulted in clinically significant reductions in HbA1c.
  • Is the change due to DHI???
  1. What are the key components of the DHI? Which ones impact on the predicted outcome, and how do they interact with each other?
  • Which components have effect on the outcome? Multiphase Optimisation Strategy (MOST) helps identify subsets of effective components.
  • Interaction effect? Sequential Multiple-assignment Randomized Trial (SMART).

Key Research questions in DHI cont’d

  1. What strategies should be used to support tailoring the DHI to participants over time?
  • micro-randomized trial? fully automated?

Key Research questions in DHI cont’d

  1. What is the likely direction and magnitude of the effect of the DHI or its components compared to a comparator?
  1. How confident are we about the magnitude of the effect of the DHI or its components compared to a comparator? 
- Addressing Q3-6 estimates direction and magnitude 
- If found feasible, cost-effective, etc, RCT may be required to determine efeect size 
- Substantial equivalence

Key Research questions in DHI cont’d

  1. Has the possibility of harm been adequately considered? And the likelihood of risks or adverse outcomes assessed? 
   - Although data is rare, DHI are not free from harm.
   - They may be designed for harmful outcomes.

Key Research questions in DHI cont’d

  • DHI may contain inaccurate info/misinformation/
  • Misinterpretation

Key Research questions in DHI cont’d

  • Opprtunity cost
  • Data breach

Key Research questions in DHI cont’d

  1. Has cost been adequately considered and measured? 
  - Considering sustainability and cost-effectiveness 
  - Long term maintainance costs
  - Economic evaluation
  1. What is the overall assessment of the utility of this intervention? And how confident are we in this overall assessment?
  1. Should research priorities change?
  1. Should clinical practice change?

Discussion and conclusions

  • Good digital health research requires multi-desciplinary collaborations
  • Technology: iterative methods, improvisation, multiple cycles of development & optimization
  • Medicine, behavioural science: RCT
  • Well-designed RCT remain important for evaluating DHI but only one part

Discussion cont’d

  • DHI requires a great deal of methodological research
  • What are open methodological questions?
  1. Enabling individual studies to generate more data
  • validation of short term proxy outcomes and when to use definitive outcomes(e.g. health status)
  • imroving methods
  • improving internal validity of RCTs without jeoparadising external validity
  • methods to reduce missing data
  • methods to sussussfully scale up and sustain

Discussion Cont’d

  1. Enabling more useful syntheses and comparison methods
  • contextual factors
  • target populations
  • specifications and classification of DHI
  • specification and determination of appropriate comparators
  • improved eporting: adherence to reporting guidlines CONSORT-eHealth, TIDieR
    • check EQUOTOR network

Questions