Methods to estimate model parameters for economic evaluation and health technology appraisal on treatment sequences

Background

Healthcare interventions are often used in sequence: e.g. people may switch to second-line cancer treatment if their cancer progresses. People with diabetes or hypertension may add in other drugs if their blood sugar or blood pressure is not reduced sufficiently. People may also switch treatment due to side-effects or other reasons.

Which treatments patients received previously and what they might receive later in the sequence can have a substantial impact on both the efficacy and cost-effectiveness of treatment. Decision-analytical models of treatment sequences are often used in economic evaluations and appraisals by health technology assessment (HTA) organisations like NICE: e.g. for cancer drugs or schizophrenia.

However, there is generally very limited evidence to inform models of sequences: particularly on the impact of past treatment on efficacy and safety. Simplifying assumptions are often made (https://doi.org/10.1007/s40273-020-00980-w), which can bias economic evaluations and underestimate uncertainty; this could in turn lead to incorrect HTA recommendations.

research experience, research methods and skills training

This DPhil will examine:

• What evidence do we need on the impact of prior treatment on efficacy, safety, utilities and/or costs to inform models?
• What data and methods could be used?

After reviewing the literature, the student will develop and test methods that can be used to inform future economic evaluations and HTA on treatments used as part of a sequence. The student will choose between one and three of the following techniques for empirical analyses:

• Network meta-analysis (NMA)
• Re-analysis of data from randomised controlled trials with/without trial-based economic evaluation
• Decision-analytical models: exploring the impact of evidence sources on the cost-effectiveness of treatments used in sequences
• Elicitation of expert opinion
• Analysis of real-world evidence or routine data
• Individual patient data meta-analysis

Contact supervisors well before applying to indicate your preferred method(s) and any disease area(s) of particular interest.

The student will be supported to publish peer-reviewed papers during their DPhil and develop skills in conducting systematic reviews and/or scoping reviews of the literature, analytical techniques, research planning, data analysis and presentation skills.

FIELD WORK, SECONDMENTS, INDUSTRY PLACEMENTS AND TRAINING

Training in health economics, analytical methods and scientific writing will be provided as needed and attendance at seminars, workshops, conferences and courses will be encouraged.

PROSPECTIVE STUDENT

Candidates should have a Bachelor’s/Master's degree in health economics, economics, medical statistics or a closely-related discipline. Proficiency in R, Stata or Excel are desirable. Candidates should also have specific experience or training in at least one area related to their project proposal: decision-analytical modelling; evidence synthesis; regression; or randomised trials.