The need for Health Technology Assessments
Health Technology Assessments (HTAs) are crucial for evaluating healthcare interventions, technologies, and treatments. They provide evidence for decision makers, guiding resource allocation and ensures cost-effective, clinically effective, and patient centred care. The HTA process promotes transparency and accountability, enhancing trust in decision-making processes in healthcare.
Following NICE guidelines
Following NICE guidelines is essential when preparing a submission. The guidelines ensure compliance with established standards and alignment with evidence-based practices; keeping them in mind during the process of writing a submission dramatically increases the likelihood of reimbursement. Wherever possible, submissions should use high quality evidence and adequate comparator(s). In rare conditions or where only limited data exists, it can be challenging to meet these expectations. We share some insights from our extensive experience in developing NICE submissions, with a particular focus on economic evaluations.
Aligning the economic evaluation with the decision problem
The key objective of the economic model is to address the decision problem. The model should be based on health states which accurately reflect real-world disease progression and functional status. Although there is no single correct approach to developing a robust economic model, it should be fully transparent in the assumptions it makes, and any uncertainties should be accounted for in the sensitivity analysis and described in the dossier. Ideally, the model structure should be simple, but still able to communicate the value of the product and capture all the effects and costs associated with the disease and treatment. Common pitfalls in economic modelling include unnecessary complexity or computational burden, failing to establish a clear link between risks and long-term clinical events, and failing to address uncertainty. Although model structures from previous assessments can be used to inform your submission, justification should not rely solely on this. An external assessment group (EAG) will scrutinise the credibility and implementation of the model and its results, so a transparent approach is key to communicating a robust argument to the Committee.
Addressing the right population and subgroups
The population presented in the submission should be consistent with the clinical evidence (e.g. trials) and the population targeted for treatment by the NHS. The question of ‘transportability’ of the data, i.e. whether similar effects are expected in the trial population and the NHS target population, should be explicitly addressed, which is particularly relevant when trials are carried out in a non-UK population. Similarly, when presenting evidence for subgroups, a clear justification is needed provided as to why and how they have been pulled out from the broader population. As such, individual subgroups should be presented because of their clinical relevance, rather than selecting subgroups that are more likely to be cost-effective. Recognising that the EAG may not possess detailed information for every subgroup, the submission should provide the necessary data and rationale to facilitate the evaluation process.
Considerations when applying survival data
In many models, short-term survival data must be extrapolated over a long-term time horizon using parametric models. The submission should carefully consider the extrapolation of survival data, especially if there are multiple good distribution fits. It is standard practice to assess the ‘statistical fit’ of the parametric distribution using the Akaike information criterion (AIC) or Bayes information criterion (BIC). Another statistical consideration is the assessment of ‘proportional hazards’, i.e., whether separate parametric models are appropriate for each arm of the trial, or whether a treatment covariate could be applied to estimate the difference between them. However, these methods do not necessarily reflect the most clinically meaningful or realistic outcomes. While the EAG and Committee expect to see a thorough assessment of the statistics underpinning survival extrapolations, additional evidence, such as registry data or expert elicitation or validation exercises to estimate long-term survival, can be extremely useful. These data sources do come with biases and limitation inherent, so they should be used with caution.
Addressing model uncertainties
The reliability of the results presented in the model and clinical dossier are dependent on the combined uncertainty in the submission, which is a key decision factor for the EAG and Committee. As a result, special consideration should be given to the probabilistic results of the model. Whilst deterministic results provide an average estimate of the product’s cost-effectiveness, probabilistic results reveal the non-linearity of the model and provide a more comprehensive understanding of the uncertainty present. To assess the accuracy of the model a sufficient number of probabilistic iterations must be conducted and compared to the deterministic incremental cost-effectiveness ratio (ICER). If a patient access scheme (PAS) price lowers the ICER below the willingness-to-pay threshold, the probabilistic ICER results must also fall below the threshold, or the EAG may raise concerns. Presenting a cost-effectiveness plane also clearly illustrates the trade-offs between costs and outcomes over a large number of iterations. Furthermore, for any uncertainties surrounding base-case inputs, additional scenarios should be added wherever possible.
At Initiate, we pride ourselves in being prepared to help guide our clients through changing times. We have experience in all major global markets and expertise across a wide range of disease areas. If you would like to discuss any market access or HEOR needs, get in touch at hello@initiateconsultancy.com.
