Book Navigation

We are sorry, but signing in with LinkedIn is not possible at this time.


Najoua Mlika-Cabanne Innovation Award 2019

Seventh Award

David Phillippo


David PhillippoDavid is early in his career, having only recently completed his MSc in Statistics at the University of Bristol in 2015, after which he took up his current position as Research Associate in Evidence Synthesis, working with the Multiparameter Evidence Synthesis group and the NICE Technical Support Unit. Within this time, he has been thinking innovatively around how we test confidence in recommendations based on network meta-analyses in a guideline development context.

Guideline developers need to assess the robustness of their recommendations to potential limitations in the evidence. However, current frameworks such as GRADE can only examine the quality of the NMA results, and do not take into account the decision context. Intuitively, low quality evidence that has negligible influence on the treatment recommendation should be of little concern, but more influential evidence should be scrutinised carefully and confidence in the robustness of the recommendation may be diminished. David has developed an algebraic approach to threshold analysis, which answers the question, “This evidence may be biased or of low quality, but how much would it have to change before the recommendation changes?” The result is a set of thresholds, within which changes to the evidence have no impact on the recommendation, but beyond the threshold a new recommendation (also output from the method) would be reached. Thresholds can also be obtained to investigate concerns about specific limitations in the evidence, for example a lack of blinding in certain studies. By combining these thresholds with judgements of the plausible magnitude and direction of possible biases or uncertainties in the evidence, the impact (if any) of plausible changes in the evidence is then easily determined, and decision makers can incorporate knowledge of robustness or sensitivity into their recommendations.

The threshold approach has been successfully used in the Specialist Neonatal Respiratory Care guideline on preterm ventilation (currently in development), and further applications in NICE guidelines are planned for later this year. David has developed and maintains a freely-available R package to perform threshold analysis using a simple interface.

Alongside the threshold work David is currently undertaking his PhD, focusing on methods for population-adjusted indirect comparisons and network meta-analyses. He is the lead author on a Technical Support Document published by the NICE Decision Support Unit that examines current population adjustment methods, their properties, assumptions, and limitations, and provides guidelines for their use in NICE Technology Appraisals. This is cited by manufacturers making submissions to NICE and by academic groups critiquing those submissions. He is now developing improved methods for population adjustment that will take network meta-analysis to an entirely new level.

David is an accomplished and exceptionally clear speaker and presenter. He has given invited talks and seminars on his work to academic and industry audiences at international conferences, universities, and companies, as well as NICE guideline development committees, working groups, and the NICE Technical Forum. This level of recognition and engagement is unusual for someone at this stage in their career. David’s work is not only innovative, but impactful.


February 2021
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
Page last updated: Oct 31, 2019
Back to top